Landscapeand UrbanPlanning, 14 (1987) 321-330 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
THE IMPACTS OF POLICY: A FRAMEWORK AND AN APPLICATION
P. WATHERN’ and S.N. YOUNG 2 University College of Wales, Aberystwyth (Gt. Britain)
I.W. BROWN and D.A. ROBERTS Clwyd County Council, Mold (0. Britain) (Accepted for publication
12 August 1986)
Wathern, P., Young, S.N., Brown, I. W. and Roberts, D.A., 1987. Assessing the impacts of policy : a framework and an application. Landscape Urban Plann. 14: 321-330. The paper describes a framework for policy appraisal and its application to a retrospective assessment of the EC LFA policy. Appraisal is shown to be dependent upon the ability not only to predict the effects of a policy during its formulation, but also to assess its impacts after implementation. Duringpolicyformulation it is important to describe the baselinesituation and to identify potential impacts, conflicts with other policies, and appropriate monitoring schemes.
‘Present address: International Institute for Environment and Society, Potsdamer Strasse - 58, D- 1000 Berlin 30, Germany. ‘Present address: Anglian Water Authority, Colchester, Essex, Ct. Britain. 0169-2046/87/$03.50
0 1987 Elsevier Science Publishers B.V.
Methods previously used in project EIA, namely matrices andjlow diagrams, appear to have an application in policy appraisal. After implementation, appraisal should follow a strict sequence of legislative, procedural and substantive review. In assessing policy impacts, environmental change and the causal factors responsible for change must be identified and dissociatedfrom the effects of other policies. The main impacts of the LFA policy in the U.K. have been on upland vegetation and farm amalgamation, but indirect effects on upland landscapes, recreation, wildlife and water resources have also probably occurred. In the U.K., application of the policy has not realised its objectives.
major development proposals, popularly referred to as environmental impact analysis (EIA), has become a commonplace component of environmental management. It has become widely adopted throughout the world, mainly because of the ease with which it can be incorporated into existing procedures and
practice. In contrast, only limited advances have as yet been made in the assessment of plans and policies (Wood, in press). Recent experience suggests, however, that project-based EIA is a simplistic solution to the problem of integrating environmental considerations into decision-making. Many of the issues associated with large development proposals are pre-empted by the policies which underlie them. Thus, for example, the planning inquiry into the Windscale reprocessing plant become bound up in matters of national energy policy which the U.K. land-use planning system is ill-equipped to address (Clark et al., 198 1). This was not an isolated incident. For instance, the protracted Sizewell “B” PWR inquiry re-enacted similar arguments to those heard previously over Windscale, while proposals for a new coal mine close to Coventry (Anon., 1985) is likely to throw U.K. Green Belt policy once again into the development arena. There are also important implications for policy formulation in the cumulative effects of individual development proposals. These incremental decisions may push policy imperceptibly in a particular direction. For example, location policy with respect to North Sea oil production platform fabrication yards, embodied in the National Planning Guidelines (Scottish Development Department, 1974)) ensures opportunities for development in certain shallow-water areas close to urban centres and the more controversial deep-water sites in several West Coast sea lochs. This “policy” merely normalised the existing situation whereby steel platforms were being built in the former sites and concrete platforms in the latter. Thus, the policy post-dated the development proposals which had the greatest impact upon the coastal environment and, in essence, was contingent upon ad hoc decisions made concerning the first few developments. A rational approach to impact analysis in land-use planning therefore requires the sepa-
ration of issues related to projects (so-called “project appraisal”) from those issues which can only be addressed at a policy level. O’Riordan and Sewell (198 1) refer to the latter process as “policy review”, although the term “policy appraisal” seems more appropriate. as review is but one of the processes involved. “Policy appraisal”, therefore, is used throughout this paper. This paper outlines a conceptual framework for the assessment of policies. The framework is discussed within the context of a retrospective review of the application of the European Communities’ (EC) policy on less favoured areas (LFA) to show how it might operate in practice. Particular attention is focussed on the implications of this policy for upland land-use change in the U.K. and its effects on semi-natural vegetation. A POLICY APPRAISAL
Policy formulation and implementation is a complex, iterative process represented in diagrammatic form by Fig. 1. The activities involved in the process can be divided into two groups, namely technical activities (represented by the rectangular boxes in Fig. 1) and the remaining political and administrative procedures. In this paper, special attention is focussed on the technical aspects involved not only in predicting the probable impacts of the LFA policy, but also in assessing its actual effects. However, some implications of the administrative procedures adopted in implementing policy are also reviewed. THE LESS-FAVOURED-AREAS
The LFA policy, devised to sustain the populations and landscapes of sparsely populated areas designated “less favoured”, has two main elements. First, the EC directive on the continuation of mountain and hill farming contains
PERCEPTION OF PROBLEM
Fig. 1. A framework for policy appraisal.
provisions for subsidising agricultural practices within less favoured areas. In the U.K. this takes the form of headage payments for grazing animals under the hill livestock compensatory allowances (HLCA) scheme. Secondly, the farm modernisation and “new agricultural structures” provisions permit member states to grant aid improvements on farms. The two sets of provisions are linked in that, for example, higher rates of grant for capital projects are available within the LFA. Thus, the key elements of this policy are headage payments and grant-aided capital investment. ASSESSING THE POTENTIAL OF POLICIES
Determining the probable impacts of a policy prior to implementation involves the suite of activities identified in Fig. 1. Impact identification, particularly determining the probable magnitude of impacts, is notoriously difficult, with wide discrepancies between predicted and actual impacts (Bisset and Tomlinson, in press). A number of methods for impact identification in project appraisal have been
developed (see for example Clark et al., 1980; Bisset, 1984; Wathern, 1984a,b). Of these, matrices and flow diagrams appear to have the greatest potential for policy appraisal, although most remain untried in practical situations (Wathern et al., 1983). Identifying impacts
The availability of headage payments and capital grants encourages farmers to increase stocking rates and to raise the productivity of their holdings. Upland vegetation has an inherently low productivity, which can be raised somewhat by the use of lime and fertilizers. Generally, however, the only effective way of raising productivity sufficiently to meet the demands of large numbers of grazing animals is to plough up this vegetation and re-seed with agricultural grasses. With these changes, therefore, there are likely to be five main types of impact associated with the LFA policy. First, there should be a beneficial effect upon the socio-economic base of upland areas, with rural depopulation arrested, or at least retarded. The main adverse impact will be upon the loss of
upland vegetation. This is discussed in more detail below. There may be direct adverse effects upon upland landscapes with farm road construction, and upon archaeological sites as a result of ploughing. In addition, there are also likely to be secondary impacts associated with the loss of upland semi-natural vegetation which affects wildlife, scenic, water and recreational resources. For example, there is likely to be a decline in visual quality with the loss of upland semi-natural vegetational mosaics. The availability of certain models of upland grazing ecosystems allows impacts upon vegetation to be explored in more detail by way of illustration. Perkins (1978) has quantified the inter-dependencies in an upland grassland ecosystem (Fig. 2). Energy-flow diagrams have been used previously in project appraisal, for
example Gilliland and Risser (1977). Such models show not only the direction, but also the magnitude, of an interaction. A component interaction matrix accommodates the cyclic features of this system (Fig. 3) and clearly shows the interdependence of sheep and the vegetation (Wathern, 1984a). The work of Perkins (1978) also indicates that the spatial heterogeneity in grazing and defaecation intensity reinforces the interactions between these two compartments of the model. Any change in either the intensity or distribution of grazing animals will therefore have marked effects upon the vegetation. Yalden (198 1) has shown that the incidence of dwarf shrubs, particularly Calluna heath, is affected by sheep density. When this climbs much above two sheep per hectare, Calluna is
STANDING DE A 0 PLANT MATERIAL
593 1 LITTER
Fig. 2. Energy-flow diagram of an upland grazing ecosystem (modified
after Perkins, 1978).
0.3 0.6 0.3
SHEEP SHEEP DUNG
Fig. 3. Component
for an upland grazing ecosystem (modifed after Wath-
gradually replaced by grassland. The Institute of Terrestrial Ecology study of upland vegetation (ITE, 1978) contains a schema showing the fate of semi-natural upland vegetation under different management regimes (Fig. 4). On the basis of this schema, the direction of a change resulting from differences in numbers of grazing animals alone can be predicted. Acidic grassland is only a density-dependent stable situation. With increasing sheep density, these grassland systems become overgrazed, with a consequent loss of vegetation cover, bringing the risk of soil erosion and raised sediment loads in water courses. On blanket peats damage occurs at quite low density, when peats may be mobilised or oxidised, increasing the problems of colour in water destined for supply. The threshold densities at
which these changes are initiated depend upon local circumstances of soils and vegetation. It should be noted, however, that the number of sheep frequently occurring in the uplands - up to ca. 14- 15 animals at some times of the year in Wales (Wathern et al., 1986) - far exceeds the threshold. Thus even without ploughing and re-seeding there are likely to be dramatic changes in upland vegetation. Lolium perenne-based agricultural swards have a higher productivity than semi-natural vegetation in upland areas, particularly when treated with lime and fertilizers. Although they regress rapidly and have to be re-sown at frequent intervals, these swards enable farmers in upland areas to raise productivity signilicantly. Fencing is an essential part of this improvement process, as it allows stock to be
q CALLUNA L ERICA
Fig. 4. Effects of management
on upland heath vegetation
controlled and the distribution of grazing animals to be regulated. Fences, however, may have an adverse effect upon recreation as they change the open character of moorland landscapes and prevent the free movement of walkers. Baseline descriptions Collection of baseline data is important for two reasons. First, impacts can only be determined by comparing the situation before and after implementation. Secondly, the identification of impacts is intimately associated with a description of the environmental characteristics of an area. It is unlikely that a broad policy will affect all areas equally. Baseline analysis, for exam-
after ITE, 1978).
ple, will indicate those geographical areas where individual impacts are likely to be manifest and where there are likely to be considerable regional variations. With respect to upland agriculture, Sinclair (1983) has identified two types of area which respond to change in different ways. It is important that baseline data are collected from both types of area and they should cover each of the main types of potential impact identified above. To allow for these differences when studying the LFA policy, Wathern et al. (1985a) selected two contrasting areas in Clwyd, north Wales, where impacts could be monitored. The initial distribution of land use and semi-natural vegetation was determined from a study of O.S. maps which more or less coincide with the onset of the post-war policy on upland agricul-
ture. No maps were available showing the position immediately prior to the implementation of the policy. As EC policy represents merely the extension of U.K. domestic provisions, see below, this is less of a constraint on analysis than at first it may appear. One of the key issues in both project and policy appraisal is the management of uncertainty (ERL, 1985). Two factors influence the level of data required in order to describe adequately the baseline situation. These are the stability of the system under consideration and the inherent variability of the parameters measured. Identifying monitoring schemes
Baseline studies should be regarded as the first phase in a stream of data-generation which runs throughout policy appraisal. It is important to identify appropriate monitoring schemes to ensure that adequate data are produced so that impacts can be assessed. With respect to the LFA policy, data on upland land use and vegetation change over time are required to indicate the effect of the policy on the physical environment. In order to explain these changes, however, detailed information on changes in agricultural practice over the same period is also required, so that cause-effect relationships can be identified. Determining policy conflicts
Increasingly, policies emanating from different sectors are seen to be in conflict. Thus, the policies which provide capital grants for farm improvement and headage payments to encourage agricultural production in many situations conflict with the provisions for landscape and wildlife conservation embodied in various aspects of environmental legislation. Similarly, grant aid for fencing encourages farmers to fence upland areas, impeding the use of these for recreation, one objective of multiple-use policy.
ASSESSING THE EFFECTS OF POLICIES
Many EC directives, being in effect supranational policy, leave detailed implementation to member states. In considering the impact of a policy in achieving change, it is therefore important to determine the extent to which this reflects a feature of the policy, or is a result of the way it has been implemented. Only when appropriate legislation and implementation measures are in place is a policy likely to effect a substantive environmental change. Reviewing the effects of a policy will be most eflicient when each of these factors is considered in turn. Policy review, therefore, should follow three stages involving legislative, procedural and substantive review. Legislative review
Special hill farm subsidies were introduced in the U.K. during the Second World War. The Hill Farming Act, 1946, formalised these provisions and was designed to increase the amount of land available for agriculture, principally livestock production, by assisting the reclamation of hill land (Rogers et al., 1985). This Act has been the basis of all subsequent post-war measures for upland farming. Thus, no new legislation was required to implement the EC directive, as domestic legislation was already in place. Procedural review
The LFA policy has been grafted onto existing U.K. practice with respect to hill farming and a narrow definition of the provisions has been adopted (Wathern et al., 1986). The area initially declared as “less favoured” coincided exactly with that previously defining eligibility for hill land subsidies under domestic policy. The provisions of the directive related to the need to protect the environment were largely ignored. Thus, the LFA brought about no fun-
damental change in U.K. policy; it was merely absorbed into a sequence of national policy spanning the past 40 years. In the U.K., therefore, any attempt to review the LFA policy per se becomes also a de facto appraisal of preexisting U.K. policy on upland agriculture. Substantive review
Substantive review is likely to be the most expensive and time-consuming aspect involved in assessing the impacts of a policy. Data on environmental quality have to be collected and analysed to see if any changes can be attributed directly to the policy. The activities involved are shown in Fig. 1. The objectives of such a substantive review are to determine the quality of the environment before and after implementation, and to isolate the effects of a particular policy. There are three stages involved. First, changes in environmental quality must be identified. Secondly, these changes must be linked to their causal factors. Finally the causal factors, in turn, must be linked to the influence of a policy. With the LFA policy it is possible to unravel some of the linkages. In the Clwyd case-study areas, the extent of semi-natural vegetation decreased markedly in the post-war period. Agricultural change was responsible for much of the demise, although increased afforestation was also a feature, particularly on the peatland in the west of the county (Wathern et al., 1985a). In this situation, it is easy to dissociate the losses which are attributable to forestry or agricultural policy. The data were obtained from O.S. maps which made it impossible to separate the effects of the LFA policy from previous U.K. domestic policy. Brown et al. (1985), in a parallel study of land-use policy, showed that aerial photographs were a more satisfactory source of data which enabled the different phases of an evolving policy to be evaluated. However, agricultural statistics for the areas also provide data on the LFA which allow the
causal factors responsible for changing environmental quality to be identified. These sources revealed that an acceleration in the rate of change in upland areas coincided with the introduction of the LFA provisions. The extent of semi-natural vegetation within the two study-areas was inversely related to rises in sheep numbers and total grant payments. A differential payment associated with different breeds of sheep included in the LFA measures provided a good indicator of policy-induced change, as the composition of flocks was directly related to the size of the differential (Wathern et al., 1985a). This source also gave details of land holdings within the study areas. Over the period of interest, the number of farms and the number of part-time farmers declined, while the size of farms increased. These data reveal continued farm amalgamation and, presumably, a related decline in the farming population. Thus, in summary, the policy which was intended to protect the landscape and sustain rural populations in sparsely populated areas designated “less favoured” appears to have been totally ineffective in realising these objectives. The policy has been implemented in such a way that a self-reinforcing cycle of intensification of agriculture has been initiated. This is in marked contrast, for example, to the situation in The Netherlands (Bennett, 1984). The main impacts in the U.K. have been a loss of semi-natural vegetation and the continued demise of the small hill farmer. The impacts of this intensification on other aspects of the upland environment such as wildlife, scenic resources and surface water bodies have not been analysed in detail, but are likely to be significant. It is known that upland archaeological sites have been affected (Wathern et al., 1987a). Only a qualitative analysis of the impact of the LFA policy has been possible because of the characteristics of the data. In other situations a more quantitative approach has been employed. The major difficulty in these cases
is to dissociate real from random change, a problem common to all aspects of environmental management. Although various statistical techniques have been used to appraise the impact of individual policies (see, for example, Jost and Sartorius, 1979; Roy and Pellerin, 1982; Wathern et al., 1985b, 1987b), no consensus on a standard analytical method has emerged.
methods for assessing the environmental impact of Community policies - a study of Wales”. The work was conducted under the aegis of the Prince of Wales Committee. This paper does not necessarily reflect the views of the Prince of Wales Committee, Clwyd County Council or the Commission with respect to the research, nor necessarily anticipate the future position of the Commission with respect to policy appraisal.
CONCLUSIONS REFERENCES Adopting a system of policy appraisal would represent a significant advance in environmental management. It would provide an opportunity to anticipate, and hopefully ameliorate, the adverse environmental effects of a policy at an early stage in its formulation. Although it should be seen as an inevitable part of the trend towards anticipatory, rather than purely reactive, environmental mangement, the realisation of policy appraisal as a routine practical process is far in the future. The major deficiencies lie in the lack of a predictive capability, some form of EIA for policies, and in detecting policy effects after implementation. Although Lee and Wood (1980) remark that the suitability of existing EIA methods for policy appraisal has yet to be demonstrated, there are at least sufficient parallels between the two processes to offer encouragement that these or related methods might be applicable. This study of the LFA policy shows that matrices and flow diagrams may have some practical application in policy appraisal. Statistical techniques and qualitative methods have been used successfully in the identification of policy impacts. Application of the framework described in this paper could do much to help make policy appraisal a practical reality. ACKNOWLEDGEMENTS This paper is based upon research undertaken for the Commission of the European Communities, entitled “The evaluation of
Anonymous, 1985. Environmental hurdles for new NCB mine. ENDS, 125: 4. Bennett, G., 1984. Application of the Less Favoured Areas Directive in The Netherlands. Instituut fur Europaische Umwehpolitik, Bonn. Bisset, R., 1984. Methods for assessing direct impacts. In: B.D. Clark, A. Gilad, R. Bisset and P. Tomlinson (Editors). Perspectives on Environmental Impact Assessment. Rheidel, Dordrecht, pp. 195-2 12. Bisset, R. and Tomlinson, P. Monitoring and auditing in EIA. In: P. Wathern (Editor), Theory and Practice in Environmental Impact Assessment. George Allen and Unwin, Hemel Hempstead, in press. Brown, I.W., Wathern, P., Roberts, D.A. and Young, S.N., 1985. Monitoring sand dune erosion on the Clwyd coast, North Wales. Landscape Res., lO( 3): 14-l 7. Clark, B.D., Bisset, R. and Wathern, P., 1980. Environmental Impact Assessment. Mansell, London. Clark, B.D.. Bisset, R. and Wathern, P., 1981. The British experience. In: T. O’Riordan and W.R.D. Sewell (Editors), Project Appraisal and Policy Review, Wiley, Chichester, pp. 125-l 53. ERL (Environmental Resources Ltd), 1985. Handling Uncertainty in Environmental Impact Assessment. Ministerie van Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer, ‘s-Gravenhage. Gilliland, M.W. and Risser, P.G., 1977. The use of systems diagrams for environmental impact assessment: procedures and an application. Ecol. Modelling, 3: 183-209. ITE (Institute of Terrestrial Ecology), 1978. Upland Land Use in England and Wales. Countryside Commission. Cheltenham. Jost, D. and Sartorius, R., 1979. Improved air quality due to lead in petrol regulation. Atmos. Pollut., 13: 1463-1465. Lee, N. and Wood, C., 1980. Methods of Environmental Impact Assessment for Use in Project Appraisal and Physical Planning. Department of Town and Country Planning, Paper No. 7, Manchester University, Manchester. O’Riordan, T. and Sewell, W.R.D., (Editors), 198 1. Project Appraisal and Policy Review. Wiley, Chichester. Perkins, D.F., 1978. Snowdonia grasslands. In: D.W. Heal and D.F. Perkins (Editors), Production Ecology of British
330 Moors and Montane Grasslands. Springer, Berlin, pp. 289-295. Rogers, A., Blunsden, J. and Curry, N. (Editors), 1985. The Countryside Handbook. Croom Helm, London. Roy, R. and Pellerin, J., 1982. On long-term air quality trends and intervention analysis. Atmos. Environ., 16: 161-169. Scottish Development Department, 1974. Coastal Planning Guidelines. Scottish Development Department, Edinburgh. Sinclair, G., 1983. The Future of the Uplands. Environmental Information Services, Narbeth, Dyfed. Wathern, P., 1984a. Ecological modelling in impact analysis. In: M.R. Roberts and R.D. Roberts (Editors), Planning and Ecology. Chapman and Hall, London, pp. 80-98. Wathern, P., 1984b. Methods for assessing indirect impacts. In: B.D. Clark, A. Gilad, R. Bisset and P. Tomlinson (Editors), Perspectives on Environmental Impact Assessment. Rheidel, Dordrecht, pp. 2 13-23 1. Wathern, P., Brown, I.W., Roberts, D.A. and Young, S.N., 1983. An Approach to Policy Appraisal. Report to the Directorate General for the Environment, Consumer Protection and Nuclear Safety, Commission of the European Communities. Wathern, P.. Brown, I.W., Roberts, D.A. and Young, S.N., 1985a. Monitoring Policy Impact in the Uplands and the Coastal Zone. Report to the Directorate General for the
Environment, Consumer Protection and Nuclear Safety. Commission of the European Communities. Wathern, P., Brown, I.W., Roberts, D.A. and Young, S.N., 1985b. Practical Considerations in Policy Monitoring. Report to the Directorate General for the Environment, Consumer Protection and Nuclear Safety, Commission of the European Communities. Wathern, P., Young, S.N., Brown, I.W. and Roberts, D.A.. 1986. The EEC less favoured areas directive: implementation and impact upon upland land use in the U.K. Land Use Policy, 3: 205-2 12. Wathern, P., Young, S.N.. Brown, I.W. and Roberts. D.A., 1987a. Upland policy and archaeological resources in North Wales. Land Use Policy, 4: 342-346. Wathern, P., Young, S.N., Brown, I.W. and Roberts, D.A., 1987b. Use of cumulative sum analysis to assess policy impacts on the River Ebbw, Wales. J. Environ. Manage., 23: 139-148. Wood, C. EIA in plan making. In: P. Wathern. (Editor), Theory and Practice in Environmental Impact Assessment. George Allen and Unwin, Hemel Hempstead, in press. Yalden, D., 198 1.Sheep densities on moorland - a literature review. In: J. Phillips, D. Yalden and J. Tallis (Editors), Peak District Moorland Erosion Study. Phase 1 Report. Peak District Planning Board, Bakewell, pp. 125-l 3 I.