|Title||Agriculture and the environment: applied general equilibrium policy analyses for the Netherlands|
|Source||Agricultural University. Promotor(en): A.J. Oskam; J.H.M. Peerlings. - S.l. : S.n. - ISBN 9789058082824 - 198|
Agricultural Economics and Rural Policy
|Publication type||Dissertation, internally prepared|
|Keyword(s)||landbouw - milieu - landbouwbeleid - milieubeleid - evenwichtstheorie - nederland - econometrische modellen - economische analyse - agriculture - environment - agricultural policy - environmental policy - equilibrium theory - econometric models - economic analysis - netherlands|
|Categories||Agriculture (General) / Environmental Policy / Environmental Economics|
There is a growing awareness of actual and potential threats to the natural environment in the form of the exhaustion of natural resources, the pollution of air, land and water resources, and the deterioration of bio-diversity. As in most industrialised countries, the concern for maintaining or improving environmental quality has taken a firm place on the policy agenda in the Netherlands. Hence, for policy makers and interest groups, it is important to understand the nature of different environmental problems, the linkages between the economy and the environment, and the economic and environmental consequences of government intervention.
The Dutch economy, agriculture and environment are highly interrelated. Agriculture, industries that are directly related to agriculture (agribusiness) and international trade in agricultural and food products form a substantial part of Dutch economic activity. Moreover, agricultural production causes a number of specific environmental problems, primarily related to the use of industrial inputs like fertiliser and pesticides. In addition, agriculture also contributes to some general environmental problems like the greenhouse effect, acidification and eutrophication.
Three relevant categories of policies can be distinguished that stress the changing policy environment of agriculture and the linkages that exist between the economy, the environment and agriculture:
In addition, the importance of environmental policies relatively to other policies in agriculture is increasing. Hence, there is scope for empirical analysis of Dutch agriculture and agribusiness, in order to unravel the qualitative and quantitative relation between the environment and economic activity.
The purpose of this thesis is to quantify the economy-wide environmental and economic effects of agricultural and environmental policies and the interactions between these policies, in the Netherlands. Some of the most important policy issues are dealt with in this thesis. Policy simulations are:
The basic tool used in this thesis is a static, single-country applied general equilibrium (AGE) model for the Dutch economy, in which environmental relations are incorporated explicitly. Given the linkages described and the economy-wide and trade effects that can be expected from agricultural and environmental policies, using an AGE for a small open economy is appropriate. Moreover, the availability of new environmental data at a very disaggregated level for the Netherlands makes it possible, and from a scientific point of view interesting, to link environmental data to economic activity in an AGE model. Finally, an AGE model provides useful information on several variables that are relevant for policy makers and interest groups.
Chapter 2 presents and discusses the AGE model and data used. Since in the different policy simulations different modifications of the model are used, the description of the model is not exhaustive. Modifications of the model, used in the different policy simulations, are dealt with in the concerning chapters. A complete description of the basic model is presented in appendices. The chapter also deals with the economic and environmental data used. Data obtained from own calculations (e.g., detailed environmental data and disaggregation of agricultural data) are summarised in appendices.
In Chapter 3, the effects on the Dutch economy of a reduction in intensive livestock production are analysed. Such a reduction is a possible solution to environmental problems linked with the excess supply of minerals to the environment. A decrease in intensive livestock production to achieve a phosphate loss of 30 kg/ha (policy goal in 2002) will decrease income from pig and poultry farming by 2.6 and 1.0 per cent, respectively. If pig production alone is reduced, the income from pig farming will decrease by 4.8 per cent. The lower production in pig and poultry farming affects the production and income of the compound feed, pig and poultry meat industries more seriously than the livestock industries because of the absence of quota rents as part of income. The effects on trade are that net exports of livestock and net imports of feedstuffs decrease. Moreover, in all cases, the exchange rate appreciates, which indicates that the trade position of the Netherlands would deteriorate because of the livestock reduction. In the case of a permitted phosphate loss of 30 kg/ha when only pig production is reduced, welfare decreases by 800 mln 1990 guilders which is only 0.15 per cent of national income. This welfare reduction would be offset by environmental improvements that are not included in the welfare measure.
The simulations give a good insight into the economic effects of a stricter mineral policy. It clearly shows that the introduction of an environmental policy that is specific for agriculture entails economy wide effects, revealing the linkages that exist between agriculture and the rest of the economy. The results form the background to discussions on the advantages and disadvantages of reducing Dutch livestock production and on the design of policies in other countries that deal with the same environmental problems. An important policy implication is the fact that industries related to the livestock industries (compound feed, pig and poultry meat industries) are affected more seriously than the livestock industries. This result is mainly due to the compensating effect of the quota rents for current farmers. However, the value of this quota (production rights) forms an entry barrier and has a negative effect on the structure of intensive livestock farming.
Chapter 4 deals with a general environmental policy that also has consequences for individual agricultural industries. In 1996, the Dutch government implemented an energy tax on fossil fuels for heating and electricity by households and 'small' energy users (small-user energy tax). The revenues of the energy tax are used to lower the pre-existing distortionary taxes related to labour. The research in this chapter shows the detailed environmental and economic effects of this Dutch unilateral environmental tax reform. Special attention is paid to the double-dividend argument that the introduction of a small environmental tax reform not only improves the environment (first dividend) but might also raise non-environmental welfare, due to an improvement in the efficiency of the tax structure (second dividend). The effects of the small-user energy tax are compared with a general energy tax, while also different tax recycling mechanisms are considered.
The simulations in this chapter show that the small-user energy tax (25 per cent for gas, 15 per cent for electricity, 25 per cent for coal and 20 per cent for other fuels for heating) causes a CO 2 reduction of 3.5 per cent while total emissions of greenhouse gases are reduced by 3.1 per cent. By recycling revenues of the small-user energy tax, employment increases by 0.10 per cent and existing tax distortions decrease, resulting in a higher national welfare of 0.06 per cent. The second best welfare improvement occurs due to the redistribution of existing tax distortions from labour to capital. When the tax base is broadened to all energy users and exemptions are ignored, welfare decreases by 0.02 per cent and the exchange rate increases by 0.25 per cent. This illustrates that in the case of a general energy tax, international competitiveness of the large energy-using industries deteriorates. Within agriculture, horticulture under glass is the most affected industry although the effects are small. Sensitivity analyses of the results show that the positive welfare effects of a small-user energy tax only apply at low tax rates. At higher tax rates, the negative distortionary effects of the introduction of a small-user energy tax dominate the positive effect of redistributing existing distortions from labour to capital. At a CO 2 reduction higher than 25 per cent, welfare costs of a small-user energy tax even become higher than welfare costs of a general energy tax, which is due to a broader tax base of the general tax.
The results show that it is rational to exempt large users from an energy tax to avoid loss of international competitiveness. Only at high reduction levels might it be more efficient to tax large energy users as well, since then an increased tax base proves to be less distorting. Under the restrictions of the model used, a second dividend can be achieved by the introduction of a small-user energy tax. At low tax rates, a welfare improvement is even possible when the revenues of a small-user energy tax are recycled in a lump sum fashion. These typical second-best results occur due to an inefficient initial distribution of the tax burden. From a policy perspective the question remains, however, whether introducing an energy tax is the appropriate tool to reduce distortions caused by other taxes.
The Dutch government has developed environmental policy targets, specified in terms of environmental indicators that measure phenomena like the greenhouse effect, acidification, eutrophication, and waste accumulation. Typically, each policy target entails a reduction in emissions that cause the environmental problem measured by the indicator. Chapter 5 analyses the environmental and economic effects of restricting these indicators, using a system of emission permits for the Netherlands. Indicators are linked to inputs, aggregate output, consumer goods and aggregate consumption at a very detailed level. Agriculture is an important contributor to these environmental indicators. The analysis focuses on the different effects of restricting single environmental indicators, the effects of restricting different environmental indicators simultaneously and the tradeability of emission permits.
The results in this chapter show large differences in welfare losses as result of restricting different environmental indicators, which can be explained by the extent to which inputs, aggregate output, consumer goods and aggregate consumption can be substituted. In the case of waste emissions and to a lesser extent of eutrophication, where emissions are related to aggregate output and aggregate consumption, substitution is hardly possible and a reduction of emissions will therefore be very costly. In the case of acidification and greenhouse gas emissions, however, a reduction can be achieved by substitution of zero or low emission commodities for high emission commodities, which entails relatively low costs. Moreover, in the latter case, emissions are widely distributed over all industries and consumers, which, especially in the case of tradeable emission permits, offers scope for an efficient allocation of the emission reduction. These results emphasise the need for a very detailed emission matrix at a disaggregated level as applied in this chapter. The simulations also show that environmental policies might interact, when different environmental indicators are related to the same economic variables. When two or more environmental policy goals are set simultaneously, individual restrictions are less restrictive and hence shadow prices of restrictions will be lower. In addition, the welfare loss of an additional environmental restriction is relatively small. Finally, the simulations in this chapter show the potential benefits of a system of tradeable permits over a system of non-tradeable permits. When permits are tradeable, permit prices for 1 kg CO 2 equivalent (greenhouse effect), 1 mole H + (acidification), 1 kg N equivalent (eutrophication) and 1 kg waste (waste accumulation) at 10 per cent reduction of the concerning emissions are 0.04, 0.18, 1.52 and 3.37 guilders (1993) respectively. These are lower than the average shadow prices in the case of non-tradeability (0.13, 1.03, 21.43 and 9.41 respectively). The difference in welfare loss between non-tradeable and tradeable permits is largest in the case of eutrophication (5476 vs. 1060 million guilders) which is due to the large differences in eutrophication emission coefficients between agents.
From a policy perspective, the simulations in this chapter give insight into the potential effects of achieving different environmental policy goals. Since both direct and indirect effects are taken into account in the AGE framework used, the links between environmental problems and economic activity are placed in a broad perspective. The simulation results show that the economic impact of an emission reduction depends largely on substitution possibilities. Since these possibilities are often limited, especially when emissions are related to output, there is a potential pay-off to increasing the search for low-emission technologies. Moreover, confirming the results obtained in earlier studies, the gain of a tradeable emission permit system over a non-tradeable system shows the need for a market-based approach when emissions have to be reduced. Finally, since restrictions on different environmental indicators might interact, there is clearly scope for policy co-ordination when multiple environmental policy goals are to be met.
Chapter 6 focuses on the environmental and economic effects of an agricultural policy change. It analyses the effects of an increase in milk quota in the Netherlands when nitrogen (N) emissions in agriculture are restricted. This policy simulation is an example of an agricultural policy change that entails environmental effects. In addition, it clearly shows the linkages between agricultural industries. The AGE model applied in this chapter is written in mixed-complementarity format (AGE-MC model), in which dairy farming is represented by a series of different Leontief technologies. Each technology is characterised by a different emission-input-output mix. Consequently, technology switches make it feasible to reduce emissions without necessarily reducing output, which would be the case if emissions were related to output in a well-behaved neoclassical production technology.
The results show that as milk quota rights become less scarce, the value of milk quota reaches zero. Since N emissions in agriculture are restricted, a higher production in dairy farming will lead to a positive and increasing shadow price of N emissions. At the point where milk quota is no longer restrictive, the shadow price is 0.99 guilders (1993) per kg N. A welfare gain can be reached by increasing milk quota while keeping N emissions at the same level. Under such a policy change, inactive N-extensive technologies in dairy farming become active and (partly) replace N-intensive technologies, while output in other agricultural industries decreases. The latter shows that policy measures taken in one industry may indirectly (through the market for N emissions) entail effects in other industries.
The simulations in Chapter 6 show that results are sensitive to the specification of technology in dairy farming. The AGE-MC approach, using multiple Leontief technologies, seems to be more flexible than using the single CES technology. If the AGE-MC approach is adopted, results depend on the specification of the alternative (both existing and latent) technologies. Especially latent technologies are difficult to specify because of a lack of information. However, if this information is available the AGE-MC approach is a useful tool for policy analysis in cases where technology switches can be expected as a result of policy changes.
The policy simulations in this thesis clearly reveal the economy-wide environmental and economic effects of agricultural and environmental policies and the interactions between these policies, in the Netherlands. However, the results should be interpreted with care for several reasons. First, since real policies are usually too complicated to be tackled in an economic model, there is always the chance of a certain degree of policy mis-specification. For example, the presence of energy covenants (in horticulture) or seasonal manure application norms are difficult to deal with in an AGE model. Second, it is worth mentioning that policies could be subject to large changes during the period in which applied policy research can be completed. Policies that first look premature, may eventually be implemented and finally turn out to be replaced or supplemented by other policies. Finally, the results are conditional on the model and data characteristics; for example, functional forms, specification of agents and commodities, and the static nature of the model. Therefore, for some of the critical assumptions (factor mobility, trade, and labour supply) sensitivity analyses were performed.
Considering the remarks and conclusions in the preceding chapters, several suggestions for future research are coming to the fore. First, in order to get more insight into the interaction between agricultural and environmental policies, there are still some policy simulations left to deal with, like a simulation on pesticides policy and policy simulations related to the CAP reform. Second, since a drawback of the AGE model is that it is not econometrically estimated, maximum entropy econometrics (an estimation techniques for small samples) in combination with frequently published SAMs could be used in the future to (partially) estimate AGE models. Third, an interesting area of research might be to incorporate micro-econometric simulation models into AGE models. Many issues in environmental economics require both detailed insights at the level of the decision-making units (e.g., individual farms) and the consequences of such decisions for the environment and the economy as a whole. Micro-econometric simulation models, on the one hand, provide detailed insight at the level of the farm (sometimes sector) and incorporate technological differences between farms. AGE models, on the other hand, consider the linkages with the rest of the economy but are less detailed. Theoretically, a link is possible given that both types of model are based on micro-economic theory. Finally, it may be interesting in further research to consider regional differences in agriculture, using regional Social Accounting Matrices (SAMs). The appearance and functioning of rural areas is receiving increasing attention because of issues like rural employment, nature production and countryside maintenance and conservation. Since agriculture contributes to rural activity and largely determines the appearance of the countryside, regional differentiation is appropriate.