Domestic Objectives in Agricultural Policies and Liberalisation

Daniel Zulauf ¹>>
^{Federal Office for Agriculture,
Switzerland}>

Introduction
1. The purpose of this short presentation is first to make a few comments on the excellent paper of Mr. Martin Scheele from the EC Commission and, second, to give some information on how domestic agricultural policy objectives may be defined. The short analysis focuses on only one of the domestic functions of local agriculture: the valuation and efficient production or provision of the environmental public good "cultural landscape".

Comments on the Paper presented by Mr. Martin Scheele
2. I agree with all the comments on the setting of environmental targets and reference levels. So especially:

• Land-use property rights must be clearly defined
• Reference levels and targets must be defined and attached to the property rights in order to allow to distribute costs between farmers and society
• Reference levels and targets must be kept separate in order to allocate costs between farmers and society, the environmental targets reflecting societal preferences for a desired allocation of resources
• "Good agricultural practice" corresponds to the reference level in the EU

3. There are a few points which might be worth to be further discussed:

• What are "good agricultural practices": what level of provision of positive and negative externalities are implied?
• You seem to imply the same effects from the application of the Polluter Pays Principle (PPP) and ecological payments. Isn't it the case with the PPP that marginal land would fall out of production and would therefore reduce the basis of pollution?
• Are society's preferences adequately reflected in the outcome of the political process, as you seem to imply?
• Isn't it inefficient to calculate premia granted under agri-environmental measures on the calculation of income losses or costs incurred, at least from a theoretical marginal benefit thinking point of view (since farmers would not need to change production in order to receive payments and consequently get the whole producer rent without doing or refraining from doing something additional)?

Domestic Objectives and Joint-Effects from Agricultural Production - a Question of Efficiency
4. High agricultural price support has several "joint-effects" at the national level. After the first main objective which was and still is to have a certain amount of efficiently produced national agricultural supplies, the main secondary effects were/are (often these functions are summarised under the term "multifunctionality"):

• Landscape management: esthetical aspects or prevention of natural disaster
• Positive ecological consequences (in comparison with a reference situation without agriculture): agricultural ²>> biodiversity for example
• Negative ecological consequences (in comparison with a reference situation without agriculture): water pollution for example
• Contribution to short run and long run food security and relating to this: maintaining the fertility of the soil for future generations
• Function of recycling all the nutrients which are a consequence of the food consumption (corresponds to the amount to be distributed on the agricultural soil as organic fertiliser). Or to put it another way: the output of the human food consumption imposes a certain production of food and fibre which cannot be reduced
• In some countries or regions: a welcome contribution to employment in rural areas (direct and induced employment)
• In many countries: contribution to certain cultural values

5. It is now clear that it was not efficient to get the positive "joint-effects" with price support because of spill-over effects: for example costly surpluses and negative externalities were generated. The main questions in many countries may be the following: "how can we achieve the goals in relation to landscape management (including in touristic regions), to the prevention of erosion and snow avalanches by forest plantation and open landscape in mountain regions, to food security (in short and longer terms) and (in some cases) to rural employment?".

Visualisation of possible Trade-Offs and Synergies
The following graph is a visualisation of the possible synergies and trade-offs in countries which are not competitive on world markets and which give great importance to non-trade concerns or domestic functions of their agriculture.

Source: B. Lehmann, 1998 (unpublished). Presented at the OECD Workshop on Emerging Trade Issues in Oct. 1998 in Paris

Domestic Objectives pursued, in particular, by less favoured Regions (as compared with CAIRNS Countries)
6. Positive secondary effects (to what extent they are coupled to the production of food and fibre is an issue for discussion) of agriculture were not perceived as being scarce until the 70ies, and nobody questioned agriculture being best at achieving these goals. The goals were only formulated when, on the one hand, trade liberalisation started to make its first inroads in agriculture and, on the other hand, agricultural activities provided less and less positive secondary effects (and more and more negative secondary effects). Also concerns were raised that, as a consequence of lack of competitiveness, agricultural activity may disappear on marginal land and with it the positive secondary effects. Furthermore, local and international tourism was more and more influencing the setting of domestic objectives (or targets) like the one of having beautiful cultural landscapes. In other words, some non-agricultural sectors were putting a value on the provision of cultural landscapes (usually provided by agriculture).

7. The main reasons for lack of competitiveness include: natural disadvantage, farm structures (also as a consequence of agricultural policy and tradition), technical inefficiency (deviation from best practices), the level of the factor prices (incl. opportunity costs) and also different environmental and animal welfare standards and regulations.

8. For efficiency reasons, domestic objectives of policies in relation to the rural area need to be defined by the local population in order to reflect their preferences. This may be done at the national or at an adequate regional level. It should be up to the local population to choose the kind of management of the natural resources they wish to have defining property rights and attached to them: reference levels.

9. In several publications on environmental economics or resource economics (as the one briefly presented at the end of this presentation, based on a recent PhD by Anna Roschewitz), we can observe a willingness to pay for cultural landscape conservation. Some studies also show a willingness to pay for more natural elements in agricultural landscape. In other words: people are willing to pay for some domestic functions of agriculture and they would also pay for more nature.

Landscape Management by Agriculture in less favoured Regions: an Example for further Discussion
10. Lehmann et al. made a short analysis for Swiss conditions for the achievement of the single goal of landscape management (by agriculture). The concept of the analysis was designed as follow:

• The agricultural land (about 1 million hectares) of Switzerland should be exploited in an efficient lowest cost way by agriculture (to what extent this would be done by "agriculture" depends on how competitive it is compared to e.g. landscape gardeners)
• The cheapest form of exploitation is grassland use (meadows, pasture)
• The grassland is used by sheep and cattle (mother cows)
• The optimisation model could choose between the two types of animals, different types of "practices" and intensities (labour and capital per ha)
• The prices for the products were actual world market prices (FAO data base)
• The prices for the factors were current actual prices including opportunity prices for labour and capital (price level of Switzerland)
• One unit of production (farm) is responsible for 40 ha or 100 acres of cultural land
• Landscape management with animals in alpine regions requires fodder conservation during the summer months; the sales to the market can by this way be continuous during the year
• The objective function is a cost minimising function

11. Results of the optimisation: Landscape management by a "minimal" agriculture:

Source: B. Lehmann, 1998 (unpublished). Presented at the OECD Workshop on Emerging Trade Issues in Oct. 1998 in Paris

12. Interpretation of the results:

• As stated above, the analysis contains only one of the domestic functions of local agriculture. It is obvious that the achievement of a different target (e.g. a more diverse landscape) and of more objectives (e.g. higher employment in agriculture) would cost more to the local, regional and national taxpayer.
• The results clearly show that price support is not efficient and not adequate for the purpose of landscape management (if we admit that an extensive land use permits to fulfil the target of landscape management and the maintenance of the soil fertility).
• The theoretical amounts for targeted landscaping payments required may appear to be high. It may be very expensive to provide cultural landscape functions. They will depend on the type of landscape chosen. They may have to be increased in the case of free trade in order to get a similar landscape. Without direct payments for land use, there may not even be minimal agricultural activity (if prices did not cover the variable costs).
• The analysis shows that the non trade functions of agriculture do affect trade. We have in any case (limited) trade distortion through the supply of agricultural products as a joint product of cultural landscape.
• A consequent issue for discussion is the definition of a tolerable supply which cannot be called trade distortion (definition of green-box requirements including a provision of de minimis).

13. The challenge for the next WTO Round will be to take these important aspects into account.

The monetary value of cultural landscape: the example of the "Zürcher Weinland"
14. Summary of the PhD by A. Roschewitz:

There is no doubt that environmental goods, such as agricultural landscapes, have an inherent value. In economic terms, this value depends on the individual benefits emanating from the environmental goods, whereby it is possible to differentiate between the various value components – use value and non-use value. The selection of the welfare measure is a decisive factor for the economic evaluation of environmental goods. On the individual level, the compensating and the equivalent variations are suitable measures and their aggregation permits on overall Direct and indirect methods are available for the monetary evaluation of environmental goods, whereby they are distinguished by their specific strengths and weaknesses. The Travel Cost Method and Hedonic Regression Method are both indirect methods while the Contingent Valuation Method an Individual Welfare Function Method number among the direct methods. The indirect methods are based on observed preferences of individuals, i.e. on effective market data. On the other hand, direct methods aim at the acquisition of revealed preferences in a hypothetical decision situation within the scope of a survey.

The Contingent Valuation Method is based on the creation of a hypothetical contingent market from which the individual's willingness to pay for certain environmental conditions can be levied. This is carried out within the scope of a verbal, written or telephone survey whereby the good to be evaluated is described precisely and the form of financing, e.g. taxes or donations, is explained. The actual evaluation questions can be formulated either as open or closed questions in dichotomous or iterative form. The Contingent Valuation Method is susceptible to difficulties arising from the hypothetical nature of the survey scenario and the associated possibility of strategic behaviour. In addition, when using closed evaluation questions there is a risk of starting point bias, i.e. the willingness to pay is influenced by the sums prescribed in the evaluation question.

The purpose of the empirical section of this paper is to use the Contingent Valuation Method to establish the magnitude and determinants for the economic value of agricultural landscape taking the Zurich wine-growing region as an example. To this end, a computer aided telephone interview survey (CATI) was conducted in which 816 people living in the Zurich wine-growing region itself and in the neighbouring city of Winterthur were surveyed. They were questioned, on the one hand, about their maximum willingness to pay to protect the landscape in its present state and, on the other hand, to implement additional improvements in the landscape of this region. This was realised in closed, iterative form using a referendum. The compensating variation was taken as the welfare measure. Those questioned had to vote on a local protection and improvement programme and, consequently, an associated rise in taxes. Four different tax amounts were specified in order to verify the influence exerted by the starting point on willingness to pay. Depending on their acceptance or rejection of the initial amount, those questioned could further increase or reduce their willingness to pay on the basis of a triple bounded dichotomous choice. The final question was open and allowed for a possible revision of the amount specified. Furthermore, those questioned could attribute their willingness to pay to various reasons representing the value components use value, existence value, option value and bequest value. In addition, the willingness to perform voluntary work in favour of the landscape was also investigated.

Both the survey concept and the aggregation of the individual willingness to pay are based on a procedure which leads to under-evaluation of the average appreciation rather than an over-evaluation. Therefore, 75% of those questioned expressed a favourable willingness to pay for the protection of the landscape. The individual willingness to pay amounts to an average of CHF 360 per annum. Fifty per cent are prepared to pay over CHF 240 (median) per annum. Sixty per cent of those questioned are willing to make a financial contribution of the improvement of the landscape amounting to an average of CHF 200 per annum (supplementary to protection). Half of those questioned stated a willingness to pay in excess of CHF 72 (median).

In the case of the protection programme, roughly 20% of those questioned made use of the possibility to revise their willingness to pay, while about 10% took advantage of this possibility with reference to the improvement programme. This led to an overall reduction of the willingness to pay amounting to 8% and 4% respectively. The answers concerning the individual value components and voluntary work do not stand up to critical examination. Therefore, the determination of the use and non-use value of the landscape has been omitted and voluntary work, expressed in monetary terms, is not used as a comparative value indicator for willingness to pay, nor is it applied to the determination of overall appreciation.

The question of determination of willingness to pay is answered in a multistage procedure. The principle willingness to pay is dealt with first of all, i.e. the positive or negative payment decision regardless of the amount of the financial contribution. Logit Models reveal that principle willingness to pay is determined in an economically plausible manner by various socio-economic and landscape-orientated variables. Therefore, the probability of a positive payment decision increases with rising income, but sinks with advancing age as well as the estimation that the Zurich wine-growing region is not really suitable for leisure activities and relaxation.

Finally, we investigate which determinants define the magnitude of the willingness to pay. To this end, the influence of the individual variables is examined using a monocausal analysis and the multiple influence of all the variables is subject to a regression analysis. It is apparent that, in general, the same factors determine the magnitude of the willingness to pay both with regard to protection and improvement of the landscape. Among the socio-economic variables, income and donations made to environmental protection exert a positive influence, while age has a negative influence on the extent of appreciation. In addition, two landscape-orientated factors have a significant effect. The magnitude of the willingness to pay declines when the promotion of meadows with flowering vegetation and hedges as well as the enlargement of areas set aside as nature reserves is felt to be unimportant. On the other hand, neither the regional origins of those questioned, nor the sum specified in the referendum question exert any influence on the magnitude of the willingness to pay. Therefore, there is no starting point bias.

Compared with other Contingent Valuation Studies, this present investigation is characterised on the one hand, by a high forecast success with regard to willingness to pay in principle, and on the other hand a relatively large share of the overall variance in the magnitude of the willingness to pay can be explained. On the whole, the results of the model permit the conclusion that the answers given by those questioned are valid and economically plausible.

Taken conservatively, the projection of the individual willingness to pay onto the population of the Zurich wine-growing region and the city of Winterthur shows that, in its present state, the landscape of this region provides an annual benefit of roughly CHF 9 million and that improvement would result in an additional benefit of about CHF 3 milllion. Applied to the agricultural landscape, this represents a value of about CHF 600 per hectare for protection and an additional CHF 180 for improvement. These results are to be understood as the scale of the absolute and area-orientated value of the agricultural landscape, whereby by definition, the appreciation of further segments of the population living outside the area under survey is not taken into consideration. Furthermore, it must be borne in mind that these benefits refer to a specific, regionally limited landscape and are not applicable to any other landscape or region. In addition, no settlement claims for the provision of agricultural landscape can be derived from the monetary value of agricultural landscape established here since entitlement to such claims depends on respective shortage situations.

On the whole, the use of the Contingent Valuation Method in this present paper for the determination of the monetary value of agricultural landscape proved satisfactory. The empirical results reveal that the population benefits to a significant extent from the landscape of the Zurich wine-growing region and that improvements to this landscape would result in additional benefits.

15. Some additional remarks on the study by A. Roschewitz:

• The landscape of the "Zürcher Weinland" is very diverse with a lot of semi-natural spaces between crops, meadows and smaller forests. The region is suitable for cycling and walking but hardly for any other leisure activities
• The monetary values are higher than the ones observed in similar studies in Germany, France or Austria but they are comparable to studies on the valuation of forests in Switzerland
• There are no indications on which components of the cultural landscape have a higher value than others (knowing them would be a condition to improving the value of the landscape)

16. Some possible policy implications derived from comparing the two studies:

• For cost-benefit analysis it is necessary to know something about the monetary value of different landscapes as well as the minimal costs to provide them. The question, whether a certain level of payments (transfers) to farmers for cultural landscaping is efficient can only be answered in comparing the two.
• The value (willingness to pay) per hectare (here US $ 400) could be compared to the costs of providing efficient landscaping (was about US $ 2'500 in the Swiss case under world market conditions, but could be lower if farms would manage more than 40 hectares or if production functions change i.e. changing factor shares). The gap could be bridged by taking into account possible positive joint effects to landscaping. In any case, landscape would only be provided to the point where marginal costs equal marginal benefits.
• A comparison of these figures can only be done with highest caution since they do not refer to the same region and were not done at the same time. Interregional as well as intertemporal stability are therefore not given.
• The difference between actual PSE (Producer Support Estimate) and the costs of providing efficient landscaping could be explained by the fact that minimal cost landscaping provides additional positive externalities or/and by political economy theory (mainly influence of farmers' lobby groups).

17. The presentation tried to show a rather unorthodox way on how domestic agricultural policy objectives may be defined for cultural landscape in comparing social benefits (calculated by e.g. the contingent valuation method) to private costs in order to reach the most efficient point where marginal (private) costs equal marginal (social) benefits.

¹>Swiss Federal Office of Agriculture, International Affairs Section. This presentation draws mainly from a paper presented by B. Lehmann (Swiss Federal Institute of Technology, Zürich) at the OECD Workshop on Emerging Trade Issues, Paris 1998 as well as a PhD by A. Roschewitz (Zürich, 1999). The views expressed in this paper do not necessarily reflect the position of the Swiss Government Back to text
²>There seems to be a low potential of being able to „legitimise" environmental payments targeted at providing biodiversity since agricultural habitats are less diverse than natural or even semi-natural habitats. Back to text.