Livestock production and the environment: some impacts of growth and trade liberalisation.
Livestock production has been increasing at a rapid pace, with a consequent increase in the associated environmental problems in many regions of the world. In New Zealand, for instance, there are concerns about the apparent growing environmental impacts of livestock production. (2) Some worry that these environmental problems will continue to worsen over time and that international trade may further exacerbate the problems. In this paper we investigate the linkages between growth, trade and the environment for livestock to the year 2005, including Uruguay Round trade reform.
With income growth, urbanisation and the modernisation of marketing infrastructures, food consumption patterns are switching from an emphasis on traditional foods (cereals and root crops) to non-traditional cereals (e.g. wheat-based foods) along with processed and high-protein foods such as animal products (Huang and Bouis 1996, Rae 1998, Delgado et al. 1999). In Asia for example, cereals still provide the bulk of calorie intakes but rapid economic development is encouraging shifts from these foods to higher-value and higher-protein foods such as those derived from livestock. This typically involves a switch in the domestic utilisation of grains from human consumption to feeding of livestock.
For a variety of reasons some countries, including several in Asia and Europe, have a comparative disadvantage in livestock production. Government assistance, including protection from trade, has been used to encourage domestic livestock production to help meet the growing demand. Such assistance has in some cases led to more intensive livestock farming systems and to a shift from backyard to commercial production units. The development of intensive livestock systems has caused concerns over effluent disposal -- animal densities per hectare of cropland are much higher in many Asian countries than in the USA, for example (Taha 1992). Environmental degradation such as water and atmospheric pollution from increased livestock production is increasing the private and social marginal costs of livestock production in these and other regions.
While much progress has been made in modelling the global consequences of agricultural trade policy reform, rather less work has been done in modelling the consequences of such policy reforms on the natural environment (Leuck et al. 1995, Anderson and Strutt 1996, OECD 2000). This is understandable given the complex interactions between farm production and the environment and the dearth of available data on those relationships. Yet it is important that progress be made if we are interested in as full a picture as possible about the welfare effects of trade reform. Increased growth and changes in the global location of farm production resulting from trade reform may have an ambiguous effect on global environmental damage (Anderson and Strutt 1996). But the concerns of environmentalists may have the potential to derail or stall trade negotiations. Improved information on anticipated environmental impacts will help us to address such concerns, while still facilitating trade liberalisation and the benefits it brings.
The first objective of this paper is to determine the impact of some of the global demand and supply developments on the location of livestock production and hence changes in the regional severity of environmental damage from livestock pollution. We attempt this through modelling a projection of the world economy from our 1995 base dataset through to 2005. This benchmark projection allows for no changes in policies, including no new policies that may target reductions in livestock environmental damage over that 10-year period. In this sense, it is a worst-case scenario. The paper's second objective is to determine the effects of trade reform on environmental pollution due to livestock production; in particular we model implementation of the Uruguay Round. The Uruguay Round (UR) made important progress on agricultural trade reform, but with barriers to agricultural trade several times as high as barriers to trade in manufacturing, much remains to be done in the next round of multilateral trade negotiations (Anderson, Hoekman and Strutt 2001). To assist the progress of negotiations in the next WTO Round, we need to improve our understanding of the anticipated environmental effects, particularly for agriculture, which has so often been a contentious issue in negotiations.
We begin by outlining the nature of environmental degradation from livestock and the environmental data we will use. In the second section of the paper, we explain our simulation of growth to 2005, both with and without implementation of the Uruguay Round. We then move to a discussion of the environmental results of our modelling, including decomposition into the scale, composition and technology effects. We end with some tentative conclusions.
2. Environmental Degradation from Livestock
Pollution from livestock farming can affect air quality, surface water and groundwater. Livestock produce around 13 billion tonnes of effluent annually and while a large part of this is recycled, such waste can pose enormous environmental problems. Animal effluent can be an environmental hazard due to its high concentration of nitrate, phosphate, potassium and ammonia. For example the global pig and poultry industries produce 6.9 million tons of nitrogen per year, equivalent to 7% of total inorganic nitrogen fertiliser produced in the world (Delgado et al. 1999). Animal feeds can contain heavy metals such as copper and zinc as growth stimulants. Their addition to the soil can pose human and animal health risks. Decomposition of effluent can release these elements directly into surface waters or they can be leached through soil to ground water sources. This threatens the quality of drinking water and may cause damage to aquatic and wetland ecosystems. In surface water, nutrients may cause excessive growth of algae and fish kills. High levels of nitrates in drinking water can cause methemoglobinemia which can be fatal to infants.
Livestock farming also results in emissions of ammonia and (in the case of ruminant animals) methane gases into the air. Livestock and effluent management contribute about 16% to global annual production of methane (Delgado et al. 1999). Methane is a potent greenhouse gas, and in some countries is a major contributor to the greenhouse effect. Land application and the storage of livestock waste are also important sources of ammonia emissions. The release of ammonia into the atmosphere contributes to acid rain and therefore to the acidification of soils and water and damage to crops and forests. Livestock's contribution to global climate change has been estimated at between 5% and 10% (de Haan et al. 1997, Steinfeld et al. 1997).
2.1. Modelling Livestock Pollution
The difference between nitrogen output and input is the surplus of nitrogen remaining on the farm during the production process. It is this surplus that may cause environmental damage through emissions to the soil, water and air. Inputs would include the purchase of fertilisers, organic manure, feed and (young) animals. They also include nitrogen supplied from the environment, such as N-fixation. Outputs would include the nitrogen content of products sold or otherwise disposed of by the farm, such as animals and animal products, crop products and manure. We use gross nitrogen production from livestock effluent as a proxy for the nitrogen surplus for various livestock production systems (3). From reviews of the scientific literature and personal communications, data were assembled on N-production from dairy, beef, sheep, pig and poultry effluent production in the USA, Europe, South Korea and New Zealand. These data were then used to develop proxies for other regions.
Data for the USA were obtained from CARD (1997). (4) Estimates were given of the weight of total waste generated as if a single animal lived on the farm for an entire year. They therefore estimate annual waste production per "animal space". Data were available for finishing hogs and breeding sows, for milk cows as well as for a further five categories of cattle, and for layers and broilers. Average data were obtained for pigs and cattle by using animal inventories for the various livestock categories as weights.
N-production data per animal-year for the European Union have been well researched and documented relative to other regions. (5) Estimates of N-production per animal year for dairy cows, sheep and poultry were taken directly from Brouwer et al. (1994) for the Netherlands, Denmark, and the rest of the EU. The same source was used for cattle and pigs, except that the data are averaged over the various cattle and pig classifications.
Some New Zealand data are available and, given the extensive nature of livestock farming in New Zealand, the data may also be useful for analysing other extensively farmed regions. Lambert et al. (1982) and Quin (1982) give data for sheep farming systems, and Ledgard et al. (1997) compare nitrogen surpluses for dairy farming in New Zealand, England and the Netherlands (the N-surplus per cow for the New Zealand system was only 25% of that in the Netherlands). These data are also discussed in Goh and Williams (1999). In the absence of New Zealand-specific data for N-production or N-surplus from pigs and poultry production, we apply USA coefficients to New Zealand. South Korean data were obtained for cattle, dairy cows, pigs and poultry (6).
The above data were in turn used to develop proxies for other regions considered in this study. Estimates for the EU (excluding Netherlands and Denmark) were applied to Japan, perhaps not unreasonably due to the intensive nature of livestock production in Northeast Asia. USA data are assumed to apply in Canada. The New Zealand data are applied to Australia and all remaining regions, on the assumption that their livestock systems are more similar to the extensive systems of New Zealand than the more intensive systems of livestock production found in the EU or the USA.
The above, combined with data on animal inventories (FAOSTAT) for our 1995 reference period, allowed estimates of total N-production (N) due to livestock effluent to be assembled:
[N.sub.r] = [[Sigma].sub.i] [a.sub.ir] [n.sub.ir]
where [a.sub.ir] is the total inventory of livestock type i in country or region r, and [n.sub.ir] is nitrogen waste production per year per animal of type i in region r.
Since any increase in total N production relative to a given area of cropland is likely to increase the N surplus, we measure total N production from farm animal waste relative to the area of agricultural land and use it as a proxy for the nitrogen surplus:
[NL.sub.r] = [N.sub.r] / [A.sub.r]
where [A.sub.r] is the combined area of cropland and permanent pasture in region r.
We combine the above nitrogen waste coefficients with agricultural land data (FAOSTAT). The nitrogen-to-land ratio (NL) is highest for the Netherlands, with over 350 kg of nitrogen produced from livestock waste per hectare of agricultural land. This is almost three times the output level for Denmark, Japan and Korea, where livestock pollution problems are also recognised as serious. The level of N-waste output relative to the agricultural land area for the remainder of the EU is not much higher than that produced in New Zealand (40 - 50 kg/ha). This measure of livestock pollution is lowest for China, North and South America, Southeast Asia and Australia. (7)
Losses due to such pollution might also be expected to increase with the size of the human population. Generally, countries with the highest levels of N-waste per agricultural hectare are also those with the highest human population densities. This is especially true in the case of the Netherlands. North and South America and Australasia have the lowest population densities and among the lowest N-waste levels per hectare.
3. Global Projections: Growth to 2005 and Implementation of the Uruguay Round
3.1. Methodology and experimental design
We use the Global Trade Analysis Project (GTAP) applied general equilibrium model (Hertel 1997) to project national and regional production, consumption and trade flows. This is a multi-region model built on a complete set of economic accounts and detailed inter-industry linkages for each of the economies represented. The GTAP production system distinguishes sectors by their intensities in five primary production factors: land (agricultural sectors only), natural resources (extractive sectors only), capital, and skilled and unskilled labour. Producers choose inputs that minimise production costs subject to separable, constant returns to scale technologies. Market clearing conditions equate supply with demand for each factor of production. In trade, products are differentiated by country of origin, allowing bilateral trade to be modelled, and bilateral international transport margins are incorporated and supplied by a global transport sector. The model is solved using GEMPACK (Harrison and Pearson 1996). In light of our interest in livestock production, we have also modified the standard GTAP model, introducing a constant elasticity of substitution among the various feedstuffs used in livestock and milk production (Rae and Hertel 2000).
The version 4 GTAP database used here is benchmarked to 1995 and offers a disaggregation of livestock production into ruminants and non-ruminants. We aggregate this database from its full 50 sectors by 45 regions up to 10 regions and 14 commodities. This aids computation and enables us to highlight sectors and regions of particular interest. Livestock farming is represented by three aggregates: beef cattle (i.e. ruminant livestock), other livestock (i.e. non-ruminants) and raw milk production. These farming sectors provide inputs to the beef processing (ruminant meat), other meat (non-ruminant meat) and dairy product industries in each region.
3.2. The base projection scenario: growth to 2005
Following the innovative work initiated by Hertel et al. (1996) in using the GTAP model for projections, we shock a small number of exogenous macroeconomic variables to simulate the effects of growth in the global economy over time. These exogenous shocks drive projections of intersectoral and international economic impacts. We make some modifications to the standard version of GTAP for simulating the projections presented here. For example increasing the Armington trade elasticities was shown to lead to more accurate simulation results in a backcasting exercise with GTAP (Gehlhar 1997). With this in mind, and following previous work on projecting with GTAP, the Armington elasticities are increased to double their standard GTAP values. Since rigidities in trade flows are expected to diminish over time, this should better facilitate the large exogenous growth shocks involved with projecting long into the future.
To project changes in the global economy, we make a number of assumptions about economic and factor growth rates. The assumed growth rates for the period 1995-2005 are given in Appendix Table 1. It is possible to close the model with either gross domestic product (GDP) or factor productivity as exogenous targets for each region. In this work, we endogenise GDP, but compare our projections with World Bank forecasted values of GDP in the final columns of Appendix Table 1. We assume an average rate of non-farm productivity growth (9) in the OECD economies (except Korea) of 0.75% per year. For non-OECD economies and Korea, we assume a somewhat higher rate of productivity growth of 1.25% per year, except for China where we assume an annual productivity growth rate of 1.75%. Therefore in China, Southeast Asia and Korea, productivity growth rates are expected to remain quite high, but somewhat lower than implied by the period prior to the Asian crisis in an effort to capture the effect of the crisis. Empirical evidence suggests that agriculture has a higher total factor productivity growth rate than other sectors (Martin and Mitra 1996). We follow Strutt and Anderson (2000) in assuming that agricultural productivity increases at a rate of 0.6 percent per annum above the rate of growth of productivity in the non-farm sector.
Forecasts for population, investment (capital stock), and labour force are based on forecasts from the World Bank. Projected changes in skilled labour are based on expected increases in the stock of tertiary educated labour and are taken from Ahuja and Filmer (1995) for developing countries. Projections for the OECD countries are based on inputs developed by the World Bank (1998). The stock of farmland in each region is held constant.
As a check on the plausibility of these assumptions, we compare our baseline cumulative GDP growth projections to those forecast by the World Bank (see the final two columns of Appendix Table 1). Together with projected growth rates in capital, skilled and unskilled labour, our productivity assumptions yield overall GDP growth rates for these countries which are in the main rather similar to those forecast by the World Bank (World Bank, 1998). We capture the continuing rapid growth in China, and a return to positive but still low growth in Japan. Otherwise our projected growth rates are a little below those of the World Bank for China, Korea and the EU, and a little above in the case of Southeast Asia.
In the solution projected for 2005, differences in the relative rates of factor accumulation interact with the various sectoral factor intensities and the income elasticities to drive changes in the sectoral composition of output over time. Globally, the largest output expansions in livestock production are projected for China, of the order of 55-60%. Substantial expansion is also projected in Southeast Asia, of between 29-38%. The smallest projected expansion of livestock production is for Japan (8-9%) and in the EU where beef and non-ruminant output expanded by around 10% whereas milk output was constrained by the current quota arrangements. For New Zealand, livestock output expansion is projected to increase by 21% and 37% for beef cattle and milk, and around 15% for nonruminant livestock.
To translate these changes in livestock and milk sector outputs to estimates of changes in N-waste production, we formed estimates of projected animal numbers. Changes in the latter will differ from sectoral output changes due to our assumed productivity growth and any substitution that occurs between livestock and the other primary factors of production (10). The problem is that livestock is but one component of the factor `capital' in livestock production, and the database used here does not allow further disaggregation of this factor. Due to relative factor price changes, substitution between capital and the other factors in livestock production will occur. One possibility therefore is to use the changes in the quantity of capital demanded by firms in the livestock sectors as proxies for changes in livestock numbers. However, there are reasons to suggest this is unrealistic. First livestock capital may comprise only a small share of total capital used in livestock production. Second, livestock capital may be less sensitive to changes in relative factor prices than other items of capital: for example, a rise in wages relative to capital prices may encourage the adoption of labour-saving capital inputs rather than an increase in livestock numbers. Our approach was to use the percentage changes in the primary factor composite (11) in livestock production as the proxy for the change in livestock numbers, the assumption being that livestock numbers change proportionately with this value-added composite. Results are given in the first three data columns of Table 1. The largest increases in livestock numbers are projected in China. But to put these increases for China in some context, increases over the 1985-95 period were 36% (pigs), 122% (chickens), 58% (beef cattle) and 135% (dairy cattle). Note that for the EU, milk output was held constant in the projections due to the milk quota but dairy cow numbers decrease by 13% -- this is the effect of the increased productivity in this sector.
3.3. The Uruguay Round Scenario
The Uruguay Round (UR) trade liberalisation is simulated from the updated 2005 database. That is, the 1995-2005 projection was performed first, and the UR scenario simulated from the updated 2005 benchmark dataset. Modelling the full UR implementation is not possible using version 4 of the GTAP database since some liberalisation is already evident in the base 1995 dataset. However we can simulate projections and tariff reductions that have the effect of moving the 1995 database to a post-Uruguay Round scenario. (12) Ideally, this simulation would include the appropriate cuts to import tariffs, export subsidy quantities and expenditures, quantitative market access targets, and elimination of the Multifibre Arrangement (MFA) quotas.
We use estimates of non-agricultural tariff reductions from work by Francois and Strutt (1999). This work draws on version 3 of GTAP, which has a carefully developed set of post-Uruguay Round tariff rates, and also the GATT/WTO integrated database. These reductions are applied to the tariff equivalents in the "other natural resources", "other processed foods" and "manufacturing and services" sectors. MFA quota removal is also modelled.
Reforms stipulated in the UR Agricultural Agreement can be grouped under domestic support, export subsidies and market access, and were phased in over a six-year period (10 years for developing countries) from 1995. Considerable problems exist in modelling this Agreement. It would not be appropriate to reduce the output subsidies in the GTAP database by the agreed 20%, since many domestic subsidies were excluded from the UR reduction commitment, and the reduction applied to the total agricultural subsidy outlay rather than those at the individual commodity level. For tariffs, it was agreed in the UR that they be reduced by an average of 36%. Thus governments had considerable leeway in reducing tariffs, and some reduced tariffs on less-sensitive commodities by more than 36% so allowing those on politically-sensitive commodities to be reduced by less. Further, the agreed rates refer to bound tariffs whereas the tariffs in the GTAP database are applied rates. Because tariffs were often bound at levels above the applied rates, the agreed tariff reductions may have little impact on the rates actually applied. The agreed reductions in export subsidies (a 36% reduction in subsidy expenditures and a 21% reduction in subsidised export volumes) have been considered to be the most effective part of the Agricultural Agreement in terms of its trade liberalisation impacts. But some countries have also taken action in an effort to weaken these commitments, and these have led to formal disputes within the WTO. As a compromise and as an approximation to the degree of reform that might be achieved, our UR scenario in this study incorporates, perhaps optimistically, 20% reductions in all tariffs and export subsidies relating to the agricultural commodities, but no reductions in output subsidies (domestic support).
Simulation results from the UR scenario described above suggest that beef cattle sector output will fall in Japan, Korea, and Southeast Asia by up to 4%, with a larger 10% reduction in the EU. Beef cattle output expands 1%-5% in all other regions. Production of non-ruminant livestock declines in Japan, Southeast Asia and Australasia, but in the latter region milk output expands substantially. Milk production declines in Northeast and Southeast Asia and the EU. There is also a substantial shift out of cereals production in Japan, Korea and Southeast Asia, and by less in the EU. The UR reforms encourage expansion of the manufacturing and services sector in the EU, Japan, Korea and China. The final three columns of Table 1 are our estimates of the changes in livestock numbers in 2005 resulting from implementation of the Uruguay Round.
4. Environmental Impacts of Economic Growth and Trade Liberalisation
By applying the N-waste coefficients ([n.sub.ir]) to the simulated changes in livestock numbers, we obtain an estimate of the separate impacts of economic growth and the UR reforms on the level of N-waste outputs from livestock production
4.1. Projections of livestock N-waste
Table 2 reports the total changes in N-waste output by region. The largest increase in N-waste output is estimated for China -- not surprising given the size of that country, its projected rapid growth rate and its likely comparative advantage in at least some forms of livestock production. But as shown in data columns 3 and 4 of Table 2, nearly all of the increase, and the subsequent environmental problems, can be traced to the effects of economic development. Very little additional livestock-induced environmental problems result from the UR trade liberalisation. Livestock environmental problems from increased N-waste production were also indicated in both North and South America. In the former region, like China, the majority of the emerging problem is due to livestock expansion as part of economic growth, rather than trade liberalisation per se. This is not the case in South America, however, where the UR trade reforms encourage expansion of the beef and dairy sectors. In New Zealand, livestock N-waste output might increase by around 15% over the 1995-2005 period, with half this increase resulting from livestock sector expansion encouraged by the UR liberalisations. Livestock environmental problems may also become more severe in Southeast Asia, but we estimate this to be entirely due to the effects of economic development: the UR reforms actually reduce this problem somewhat since livestock production in this region could contract.
We project a more than 10% reduction in livestock N-waste outputs in the EU over the 1995-2005 period. This reduction is driven almost equally by the downsizing of the livestock sector that results from future economic development and changing comparative advantage within the region, and the additional downsizing enforced by the reforms of the Uruguay Round. Similar remarks can be made in the case of Japan.
4.2. Decomposition of Environmental Effects
From these changes in N-waste outputs, three sources of environmental effects of policy changes can be identified: the change in the level of aggregate economic activity, the change in the contribution of each sector to output, and the change in production technology. This decomposition is useful for disentangling the causes of changes in environmental damage (see for example Fredriksson 1999, Strutt and Anderson 2000). Increased aggregate economic activity leads to increased demands for all goods and services and therefore increased emissions. The change in output due to the aggregate activity effect is the proportional change in aggregate real output in the economy multiplied by the initial output in each sector. This gives the change in the scale of output in each sector with all sectors growing at the aggregate growth rate of the economy. The second effect is the intersectoral composition effect. Because some sectors are more polluting than others, changes in the composition of output will change pollution, even if aggregate output were to remain constant. The intersectoral effect is measured by allowing the composition of output to change while maintaining aggregate output at its initial level. Thirdly, there is the technology effect. We can capture part of the technology effect by identifying the impact on environmental damage of assuming an improvement in livestock productivity. Improved productivity implies that the same marketable outputs can be produced from a smaller number of livestock. Since we assume N-waste output is determined by livestock numbers, the consequent technology effect will lessen environmental damage. (13) Changes in technology will therefore change the amount of degradation caused by each unit of output in each sector. (14) The summation of the scale, composition and technology effects will equal the total change in N-waste output that we project.
When we decompose these environmental effects into the scale, technology and composition components, further insights are gained. As expected, the increased level of projected output from 1995-2005 drives relatively large scale effects (data column 1 of Table 3). The technology effect dampens these scale effects significantly, given our assumption of increased livestock productivity and therefore lower environmental damage per unit of output in the livestock sectors. The composition effect further dampens the scale effect in all cases. This positive impact of the composition effect on the environment is interesting as it indicates that the quantities of livestock (which cause the environmental damage in our model) are growing less rapidly than is the average rate of output growth across all sectors. For Japan, Australia and Europe the combination of the positive technology and composition effects are sufficient to overturn the negative scale effects on the environment (see data columns two and three of Table 3).
When the environmental results of the Uruguay Round are decomposed, we find that they are primarily driven by changes in the composition of output (15) The scale of output changes is relatively small in this comparative static model, but there are significant changes in the composition of output. In particular, Japan, Southeast Asia and Europe are projected to move resources out of their livestock sectors and thus reduce aggregate N-waste output levels.
4.3. Environmental damage relative to agricultural land and population density
The social cost of environmental damage from increased N-waste outputs is likely to be more closely related to changes in effluent outputs per hectare of agricultural land (N[L.sub.r]) and population density, rather than to the raw total N-waste output levels. Figure 1 shows the change in N-waste output per hectare, decomposed into the effects due to economic growth and those due to the UR trade reforms. The largest increase in N-waste output per hectare due to economic growth occurs in China, and the largest such reduction in Japan followed by the EU. But as far as the impacts of trade liberalisation are concerned, by far the largest impact on increased N-waste output per hectare occurs in New Zealand. By comparison, this increase is much smaller in North and South America and Australia.
[FIGURE 1 OMITTED]
High levels of N-waste output per unit of agricultural land tend to pose the greatest environmental problems to humans in regions with high human population densities (such as in the Netherlands). Figure 2 indicates a relationship between the estimated changes in N-waste outputs per hectare and human population density in each region. This figure focuses on the changes in N-waste output that may result from the UR trade liberalisation alone. As can be seen, trade liberalisation appears likely to be accompanied by increased environmental pressure from livestock in countries with relatively low population densities, such as New Zealand and North and South America. In these regions the human consequences of such damage may be relatively low. On the other hand, our results suggest trade liberalisation may lead to reduced potential livestock environmental problems in the densely populated countries of the EU and Northeast and Southeast Asia.
[FIGURE 2 OMITTED]
Irrespective of the extent of the increased environmental damage from livestock production relative to human population densities, we would emphasise that the first-best policy response is to address the environmental problem with carefully targeted environmental policies, rather than by using trade policy. Model results from this comparative static simulation project a welfare gain to New Zealand due to the UR trade reforms of US$480 million (1995 prices). While this gain does not recognise any loss of welfare due to increased environmental degradation, scope clearly exists for the net gains from trade reform to remain significant and positive, even after taking account of any environmental degradation.
Growth in incomes, population and urbanisation are fuelling a rapid increase in demand for animal products, particularly in developing countries. This is encouraging the expansion of livestock numbers and the intensification of livestock production. Our study suggests that continuation of these forces will, in the absence of appropriate intervention, further exacerbate the negative environmental impacts of livestock production in many regions of the world. However in the highly industrialised economies of Japan and the EU, future economic development appears likely to feature a relative decline in the size of the livestock sector.
Whether reforms to trade policies will enhance or degrade the natural environment is an empirical matter, and will depend partly on how the altered economic incentives affect outputs of pollution-intensive relative to pollution-extensive industries and sectors. Dairy production is one of the world's most highly protected agricultural activities, through high tariffs and (especially in the EU) substantial export subsidy payments. Consequently, our simulation of the Uruguay Round Agricultural Agreement suggests a contraction of the dairy sectors for the EU and parts of East Asia, but expansion elsewhere. The beef sector also contracts in the above regions, as does non-ruminant livestock production in East Asia. To the extent that farm protection is highest in the high-income, densely populated countries of Northeast Asia and Western Europe, lowered farm protection could see less effluent from livestock and less fertiliser used in cropping, with relatively high gains to society due to high population densities in these regions. Furthermore, some of the farm production is likely to shift to other regions of the world, where human population densities are much lower and farm production systems are more extensive. Thus the additional environmental damage in the latter countries could be much less than the reduction in environmental damage in the densely populated regions (Anderson and Strutt 1996). Extensive livestock production systems also tend to utilise less grain-feeding than intensive systems, with increased reliance on nitrogen-fixing pasture plants, both suggestive of net environmental gains from the relocation to extensive systems. Our quantitative analysis confirms these effects.
Even in the absence of specific environment-enhancing policies and activities, we suggest the UR trade liberalisation would reduce the level of N-waste output from livestock farming in parts of Europe and Asia. Trade liberalisation may increase livestock environmental problems in countries such as New Zealand and North and South America but due to their relatively low population densities, the human consequences of such damage may be relatively low. On the other hand, trade liberalisation leads to reduced livestock production in the densely populated countries of the EU and Northeast and Southeast Asia, and therefore offers the potential of overall gains in environmental quality. Decomposition of these environmental impacts reveals that they are primarily driven by changes in the composition of output. In particular, Japan, Southeast Asia and Europe are projected to move resources out of their livestock sectors and thus reduce aggregate N-waste output levels.
While we did not model changes in environmental policy, (16) improved policy ought to be considered if the projected environmental damage due to trade policy reforms is to be reduced or avoided. For example, with rapid growth in the decade to 2005, China is likely to experience a large increase in livestock pollution in the absence of appropriate environmental policies to dampen the impact. New Zealand may also need to further consider appropriate environmental policies to limit the impact of livestock pollution due to growth and trade reform. While the absolute increase in livestock production is not as large for New Zealand as for many other countries, our results indicate that per farmed hectare, it might experience a greater increase in livestock pollution with Uruguay Round liberalisation than any other region we model. However this comment should be tempered with mention of the low population density in New Zealand, which may limit the damage to human health.
There are of course a number of important tradeoffs and limitations with this type of work. In particular, with our focus on global trade reforms, we had to work at an aggregate level of analysis that required us to treat livestock pollution as a `national' problem. In reality, there often exist `hot spots' of pollution, for example in intensive pig production regions, the environmental impacts may be many times more severe than is indicated by national indicators. In the case of dairy farming in New Zealand, livestock pollution in Canterbury and the Waikato are likely to be of particular concern. Local level studies will therefore complement (and be complemented by) this work. In addition, we only consider environmental damage from one sector. (17) Changes in other sectors will also impact on the net national and international level of environmental damage. However, given the model and data we use, our analysis suggests that for world livestock production the aggregate environmental implications of trade policy reform appear to be small or even positive. On the other hand, in the absence of improved environmental policies, the aggregate environmental impact of structural change to 2005 is likely to be of much greater consequence to those concerned about environmental damage.
Table 1 Estimated changes in livestock numbers (%) Projection 1995-2005: no policy reform Beef Other Dairy cattle livestock cattle Australia -0.7 -0.8 -1.2 New Zealand 5.7 -0.4 20.1 Japan -5.4 -5.1 -4.3 Korea 0.7 1.0 1.4 China 24.1 25.0 29.0 SE Asia 7.3 7.5 14.8 Nth America 5.7 8.9 7.4 Sth America 0.9 1.6 1.4 EU -3.1 -4.2 -12.6 Due to the Uruguay Round: 2005 Beef Other Dairy cattle livestock cattle Australia 1.4 -0.4 2.2 New Zealand 4.7 -34.6 20.7 Japan -3.6 -1.4 -2.3 Korea -3.9 1.0 -0.1 China 0.7 0.5 0.4 SE Asia -3.1 -1.5 -6.7 Nth America 1.3 0.6 0.5 Sth America 2.8 0.3 1.4 EU -10.4 0.0 -2.8 Source: Authors' calculations (GTAP variable qva). Table 2 Change in N-waste outputs ('000MT) Total N-waste Projected 1995-2005 Total output change under: projected change 1995 2005 No policy Uruguay reform Round Australia 2069 2081 -16 28 12 New Zealand 736 852 59 57 116 Japan 608 564 -31 -13 -43 Korea 293 236 2 -1 2 China 11631 14588 2874 83 2957 SE Asia 1571 1647 118 -42 76 Nth America 9094 9786 588 103 691 Sth America 14408 14902 142 352 494 EU 8229 7302 -476 -450 -926 Source: Authors' calculations. Table 3 Decomposition of the change in N-waste outputs ('000MT) Projection: 1995-2005 Scale Tech- Composition Total Effect nology Effect Effect Effect Australia 612 -298 -330 -16 New 228 -106 -63 59 Zealand Japan 81 -87 -25 -31 Korea 90 -59 -28 2 China 7362 -3036 -1453 2874 SE Asia 638 -316 -204 118 Nth 3063 -1310 -1165 588 America Sth 4700 -2896 -1662 142 America EU 1469 -1185 -761 -476 Total 18243 -9293 -5691 3260 Uruguay Round Scale Composition Total Effect Effect Effect Australia 4 24 28 New 1 56 57 Zealand Japan 0.5 -13 -13 Korea 0.5 -5 -4 China 37 47 83 SE Asia 7 -49 -42 Nth 2 101 103 America Sth 23 329 352 America EU 10 -460 -450 Total 85 30 114 Source: Authors' calculations.
(*) Acknowledgements: This study was funded through the Foundation for Research, Science and Technology contract number IER501, from which support is gratefully acknowledged. Thanks are also due to the three referees who offered many useful comments and suggestions.
(1) This is particularly the case for dairy farming in Canterbury and the Waikato (see, for example, Neeley 2001 and Cassells and Meister 2001)
(3) Ideally, data would have been assembled on nitrogen balances, but these data are not generally available.
(4) Additional data was obtained through personal communication with B.A.Babcock, from unpublished manuscript "Manure production from hogs, cattle, chickens and turkeys".
(5) Much of this research refers to the Netherlands, where the environmental problems are among the most severe.
(6) Personal communication, Dr S-K Choi, Korean Rural Economics Institute.
(7) However, it must be recognised that average N-waste output at the national level may hide localised high-pollution pockets, such as in parts of the USA, France and the UK.
(9) Throughout this paper when we refer to productivity growth we are referring to the productivity of value-added.
(10) We assumed a zero substitution between intermediate production inputs and the individual primary factors. The version of GTAP used by us also assumes all pairwise elasticities of substitution among the primary production factors are equal (and non-zero).
(11) Including land, skilled labour, unskilled labour and capital.
(12) We do not fully include China in the Uruguay Round implementation modelled. The main difference would be whether China is excluded or included in getting expanded access to US and EU textile and clothing markets (see Anderson et al. 1997, and Strutt and Anderson 2000).
(13) For a discussion of other possible components of the technology effect, see Fredriksson (1999).
(14) For policy changes such as trade liberalization where we start from the appropriate updated database, we assume that the new technology is in place and that the trade reform itself does not change the environmental damage coefficients.
(15) Recall our earlier assumption that livestock productivity changes over the period 1995 2005, but does not change further with the Uruguay Round implementation. The technology effect is therefore zero for our Uruguay Round simulation. However, this may be regarded as pessimistic in that increased trade openness and income growth may be expected to further enhance livestock productivity.
(16) For some recent work on interventions to reduce livestock pollution, see Cassells and Meister (2001), Komen and Peerlings (1998), Reinhard et al. (1998) and Brouwer et al. (1999).
(17) And we focused on just one environmental indicator, when livestock pollution is multidimensional. When other indicators are available for a number of livestock types and countries, this shortcoming can of course be rectified.
Ahuja, V., Filmer, D., 1995. Educational Attainment in Developing Countries: New Estimates and Projections Disaggregated by Gender. World Bank Policy Research Working Paper 1489, Washington D.C., July.
Anderson, K., Dimaranan, B., Hertel, T., Martin W., 1997. Economic Growth and Policy Reform in the Asia-Pacific: Trade and Welfare Implications by 2005. Asia-Pacific Economic Review 3(1), 1-18.
Anderson, K., Hoekman, B., Strutt, A., 2001. Agriculture and the WTO: Next Steps. Review of International Economics 9(2), 192-214.
Anderson, K., Strutt, A., 1996. On Measuring the Environmental Impact of Agricultural Trade Liberalisation. In: Bredahl, M.E., Ballenger, N., Dunmore, J.C., Roe, T.L. (Eds.), Agriculture, Trade and the Environment: Discovering and Measuring the Critical Linkages. Westview Press, Boulder, Colorado, 151-72.
Anderson, K., Strutt, A., 1999. Impact of East Asia's Growth Interruption and Policy Responses: The Case of Indonesia. Asian Economic Journal 13(2), 205-218.
Brouwer, F., Hellegers, P., Hoogeveen, M., Luesink, H., 1999. Managing Nitrogen Pollution from Intensive Livestock Production in the EU. Report 2.99.03, Agricultural Economics Research Institute (LEI), The Hague, February.
Brouwer, F.M., Godeschalk, F.E., Hellegers, P.J.G.J., Kelholt, H.J., 1994. Mineral Balances at Farm Level in the European Union. Agricultural Economics Research Institute (LEI), The Hague.
Card, 1997. Agricultural and Environmental Outlook. Resource and Environmental Policy Division, Centre for Agricultural and Rural Development, Iowa State University, Ames.
Cassells, S. and Meister, A.D. 2001. Cost and Trade Impacts of Environmental Regulations: Effluent Control and the New Zealand Dairy Sector. Australian Journal of Agricultural and Resource Economics 45(2), 257-274.
de Haan, C., Steinfeld, H., Blackburn, H., 1997. Livestock and the Environment: Finding a Balance. Report of a study coordinated by the Food and Agricultural Organisation, the United States Agency for International Development, and the World Bank, European Commission Directorate-General for Development, Brussels.
Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S., Courbois, C., 1999. Livestock to 2020: The Next Food Revolution. Food, Agricultural and the Environment Discussion Paper 28, International Food Policy Research Institute, Washington,D.C.
FAOSTAT, FAO. Statistical Databases, http://apps.fao.org.
Francois, J., Strutt, A., 1999. Post Uruguay Round Tariff Vectors for GTAP Version 4. Mimeo, Global Trade Analysis Project, Purdue University, West Lafayette.
Fredriksson, P., 1999. Trade, Global Policy, and the Environment: New Evidence and Issues. In: Fredriksson, P. (Ed.), Trade, Global Policy, and the Environment. The World Bank, Washington, DC.
Gehlhar, M., 1997. Historical Analysis of Growth and Trade Patterns in the Pacific Rim: An Evaluation of the GTAP Model. In: Hertel, T.W.(Ed.), Global Trade Analysis: Modeling and Applications. Cambridge University Press, Cambridge and New York.
Goh, K.M., Williams, P.H., 1999. Comparative Nutrient Budgets of Temporate Grazed Pastoral Systems. In: Smaling, E.M.A., Oenema, O., Fesco, L.O. (Eds.), Nutrient Disequilibria in Agroecosystems. CAB International, Wallingford, U.K., 265-294.
Harrison, W.J., Pearson, K.R., 1996. Computing Solutions for Large General Equilibrium Models using GEMPACK. Computational Economics 9, 83-127.
Hertel, T.W.(Ed.), 1997. Global Trade Analysis: Modeling and Applications. Cambridge University Press, Cambridge and New York.
Hertel, T.W., Martin, W., Yanagishima, K., Dimaranan B., 1996. Liberalising Manufactures Trade in a Changing World Economy. In: Martin, W., Winters, L.A. (Eds.), The Uruguay Round and the Developing Countries. Cambridge University Press, Cambridge and New York, Chapter 7.
Huang, J., Bouis, H., 1996. Structural Changes in the Demand for Food in Asia. Food, Agriculture and the Environment Discussion Paper No. 11, International Food Policy Research Institute, Washington D.C., March.
Komen, M.H.C., Peerlings, J.H.M., 1998. Restricting Intensive Livestock Production: Economic Effects of Mineral Policy in the Netherlands. European Review of Agricultural Economics 25, 110-28.
Lambert, M.G., Renton, S.W., Grant, D.A., 1982. Nitrogen Balance Studies in some North Island Hill Pastures. In: Gandar, P.W. (Ed.), Nitrogen Balances in New Zealand Ecosystems, DSIR, Palmerston North, New Zealand, 35-39.
Ledgard, S.F., Penno, J.W., Sprosen, M.S., 1997. Nitrogen Balances and Losses on Intensive Dairy Farms. Proceedings of the New Zealand Grassland Association. 59, 49-53.
Leuck, D., Haley, S., Liapis, P., McDonald, B., 1995. The EU Nitrate Directive and CAP Reform: Effects on Agricultural Production, Trade, and Residual Soil Nitrogen, Foreign Agriculture Economic Report No. 255. Economic Research Service, U.S. Department of Agriculture, Washington, DC.
Martin, W., Mitra, D., 1996. Productivity Growth in Agriculture and Manufacturing. Mimeo, International Trade Division, The World Bank, Washington, DC.
Neeley, A., 2001. Water Protection a Juggling Act for Councils. New Zealand Herald, 16 April 2001. www.nzherald.co.nz.
OECD, 2000. Assessing the Environmental Effects of Trade Liberalisation Agreements: Methodologies. OECD, Paris.
Quin, B.F., 1982. The Influence of Grazing Animals on Nitrogen Balances. In: Gandar, P.W. (Ed.), Nitrogen Balances in New Zealand Ecosystems. Department of Scientific and Industrial Research, Palmerston North, New Zealand, 95-102.
Rae, A.N., 1998. The Effects of Expenditure Growth and Urbanisation on Food Consumption in East Asia: A Note on Animal Products. Agricultural Economics 18, 291-299.
Rae, A.N., Hertel, T.W., 2000. Future Developments in Global Livestock and Grains Markets: The Impacts of Livestock Productivity Convergence in Asia:Pacific. Australian Journal of Agricultural and Resource Economics 44(3), 373-423.
Reinhard, S., Lovell, C.A.K., Thijssen, G., 1998. Econometric Estimation of Technical and Environmental Efficiency: An Application to Dutch Dairy Farms. American Journal of Agricultural Economics 81, 44-60.
Steinfeld, H., de Haan, C., Blackburn, H., 1997. Livestock-Environment Interactions: Issues and Options. Report of a study coordinated by the Food and Agricultural Organisation, the United States Agency for International Development, and the World Bank. European Commission Directorate-General for Development, Brussels.
Strutt, A., Anderson, K. 2000. Will Trade Liberalisation Harm the Environment? The Case of Indonesia to 2020. Environmental and Resource Economics 17, 203-232.
Taha, F.A., 1992. Environmental Constraints in East and Southeast Asia: Potential for Changing Patterns of Agricultural Production and Trade. In: Pacific Rim Agriculture and Trade Report: Situation and Outlook Series. Economic Research Service, USDA, Washington, DC. 47-52.
World Bank, 1998. Global Economic Prospects and the Developing Countries. The World Bank, Washington, DC.
Allan N. Roe and Anna Strutt (*)
(*) Allan N Rae, Director, Centre for Applied Economics and Policy Studies, Massey University and Anna Strutt, Department of Economics, Waikato Management School, University of Waikato
|Printer friendly Cite/link Email Feedback|
|Author:||Rae, Allan N.; Strutt, Anna|
|Publication:||New Zealand Economic Papers|
|Date:||Dec 1, 2001|
|Previous Article:||Work experience and the gender earnings gap.|
|Next Article:||A cohort analysis of household income, consumption and saving.|