Morocco - Environment Review (Vol. 2 of 2) : The Working Papers

Report No. 16570-MOR CONFIDENTIAL 16570-MOR VOL. 2 Kingdom of Morocco Environment Review (In Two Volumes) Volume II Working Papers May 23, 1997 Natural Resources & Environment Division Maghreb and Iran Department Middle East and North Africa Region C ON A IV Document of the World Bank RL E M P This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Kingdom of Morocco: Morocco Environment Review Working Paper I COORDINATED TRADE AND ENVIRONMENT POLICIES FOR MOROCCO by David Roland-Hoist Mills College and CEPR in collaboration with Maurizio Bussolo University de los Andes and Fedesarrollo 20 April 1997 1511 Josephine Street Tel: (1) 510-845-4567 Berkeley, CA 94703 USA FAX: (1) 510-845-2155 Internet dwrh@mills edu The authors would like to thank Dominique van der Mensbrugghe, John Beghin, and Sdbastien Dessus for comments The opinions expressed here are those of the authors and should not be attributed to any of their affiliated institutions or the World Bank CONTENTS I EXECUTIVE SUMMARY . .. . .1 2 INTRODUCTION .. .... ..4 3 ECONOMIC ACTIVITY AND THE ENVIRONMENT IN MOROCCO . . . 6 4 THE MACROECONOMIC REFORM SCENARIO 9 41 GROWTH AND EMISSIONS 10 5 TRADE AND ENVIRONMENT POLICY SCENARIOS 13 51 LIBERALIZED TRADE WITH THE EUROPEAN COMMUNITY 13 5 2 ENVIRONMENTAL POLICIES . . ......... . 17 5 3 COMBINED ENVIRONMENTAL AND COMMERCIAL POLICIES 19 6. CONCLUSIONS .... .. . ... 23 7 R E F E R E N C E S . . . . 2 4 8 ANNEX - TECHNICAL SPECIFICATIONS .....26 8 1 THE MOROCCAN TRADE AND ENVIRONMENT MODEL. ... 26 8 1.1 Production . ...... . 26 8 1 2 Income Distribution and Absorption 29 8 1 3 International Trade... . ...... . ... . ...... .. 29 8 1 4 Model Closure and Dynamics.. .... 31 8 .1 5 E m issions . ..... ...... .. ... .. . . .......... .... . . ............ ... .3 1 8 1 6 Policy Instruments .. ..... . ......... .. . . 32 8 2 DECOMPOSITION OF POLLUTION EFFECTS . ....... 33 83 POLLUTION CATEGORIES .. . 34 1. EXECUTIVE SUMMARY An empirical simulation model was used to evaluate linkages between economic and environmental policies in Morocco In addition to assessing the status quo pollution intensity of economic activities, we examined the environmental implications of trade policies, the economic implications of environmental policies, and the prospects for coordinating the two to reconcile growth and abatement objectives Preliminary results indicate that the environmental effects of trade liberalization are potentially detrimental, but these can be significantly offset, without seriously compromising growth, by carefully targeted abatement measures An evaluation of status quo production and consumption patterns indicates that pollution intensities vary widely across the Moroccan economy, and some of these differences have implications for trade-environment linkages. Export agriculture is much more pollution- intensive than Food agriculture Another leading export sector, Mining, is among the most pollution-intensive in the economy Textile production, on the other hand, exhibits about average overall pollution intensity, although some of its chemical effluents have high toxicity The macroeconomic reform and stabilization program currently under way in Morocco can be expected to significantly increase the scale of domestic pollution, while keeping its composition and intensity (per unit real GDP) relatively constant. This is because of the gradual and relatively homogeneous nature of the macro-induced economic expansion While this pollution growth poses a new challenge for environmental management, its uniform and stable pattern provides an ideal context for design and implementation of more comprehensive policies toward monitoring and abatement By implementing the agreed terms of its FTA with the European Community, Morocco will remove significant and relatively distortionary import protection Doing so will provide considerable impetus to its economy but, in the absence of other policy measures, most of the growth will arise from intensification of the country's traditional comparative advantages, especially in primary exports. For this reason, the FTA could significantly increase pollution in Morocco The FTA will induce significant economic growth in Morocco, with attendant scale effects on domestic pollution The ultimate environmental effects are more detrimental than indicated by scale effects alone, however, because of shifts in the composition of domestic production and consumption toward more pollution- intensive products These results may be somewhat too pessimistic because they do not take account of two potentially important but uncertain factors Firstly, there is significant scope for pollution - I - reduction if Moroccan reforms stimulate foreign direct investment, significant technology transfer, and greater private sector efficiency Secondly, it is reasonable to expect Moroccan environmental standards and policies will eventually be harmonized with those of its largest trading partner. Both these trends could do much to moderate or even attenuate pollution growth in the country Although targeted environmental policies such as pollution taxes remain to be officially enunciated, they can lead to significant reductions in the level and intensity of pollution in Morocco Emission-specific pollution taxes or incentive-equivalent quota schemes, for example, can reduce the targeted emissions by up to 25 per cent at modest cost in terms of domestic economic growth. It is also apparent that significant linkages exist between effluents, and reductions can be obtained in a large family of pollutants from specific taxes on a small subset As a practical matter, this can lead to considerable savings in terms of administrative and monitoring capacity When trade liberalization is coordinated with economic instruments to reduce pollution, the growth potential of the former is largely fulfilled along with the beneficial effects of the latter It is apparent from these preliminary results that the economic opportunity the FTA represents can be realized, but its negative environmental consequences largely or completely offset, provided Morocco commits itself to developing the capacity for detailed pollution monitoring and management. Failure to do so will lead to unsustainable trajectories of pollution growth, and significantly higher economic and social costs further down the road By implementing the IPPS database in a policy simulation framework, the GE model used here represents a valuable diagnostic tool to support trade and environmental policy In the present context, it reveals the macroeconomic benefits of reform, while illustrating the importance of improved environmental regulation and resource management Further analysis within the GE framework can facilitate better integration of economic and environmental policies, minimizing growth-environment trade-offs -2- Table 1: Summary of Policy Recommendations Issue Analysis Recommendation Status Quo Pollution Pollution intensities vary Target pollution-intensive sectors for widely across domestic policies to achieve reduced emission loads production activities Much and promote innovation and efficiency higher in Export Agriculture Emphasis on a relatively small number of than in Food Ag Very high in activities/effluents can achieve significant Mining Moderate in Textiles reductions in aggregate pollution Macroeconomic Essential to sustainable The macro reforms provide a good Stabilization and growth, but likely to induce opportumty to develop and implement a Reform significant increases in comprehensive policy for pollution pollution levels Intensities by monitoring and reduction Gradual and activity and pollutant remain relatively homogenous growth under this about the same scenario provide a good reference case for stabilizing economywide pollution levels Free Trade Expanded access to external Import tariffs should be replaced by a Agreement with the markets is essential accelerate nondistortionary tax, uniform across all EC investment, employment, and domestic and imported commodities and income growth in Morocco raised only to the minimum level necessary to offset declining tanff revenue Non-tanff barriers should be abolished FTA and the A senous potential threat Pollution-intensive expanding sectors should Moroccan Comparative advantages in be targeted with mitigating environmental Environment some dirty activities are policies, including effluent taxes, investment intensified Pollution intensities and innovation incentives, and abatement rise significantly in some major incentives. Conformity with EC standards effluent groups. will help to achieve this Environmental When properly targeted, Implement emission taxes targeted very Policies economic instruments can narrowly to effluent types. Look for linkages achieve significant reduction of between different emission groups to emission loads minimize administration and monitonng costs Incentive-equivalent policies, such as tradable peruts, should also be considered Coordinated Trade Show considerable potential to While liberalization should be as complete and Environment fulfill the growth promise of as possible, targeting of economic Policies external opening, in tandem instruments to mitigate pollution should be with reductions in domestic more specific The right combination of pollution effluent taxes can yield most of the growth arising from trade liberalization, while reducing domestic pollution in several major -effluent categories simultaneously -3- 2. INTRODUCTION Trade and environment linkages are under increasing scrutiny, and a vast literature has emerged on the subject (see Dean (1992), Cropper and Oates (1992), Beghin, Roland- Hoist and van der Mensbrugghe (1994) for surveys). Pressures to use trade instruments for environmental ends have been debunked as a blunt and inefficient approach to environmental policy (Anderson and Blackhurst (1992)) Even in a second best world, the optimal policy to abate emissions would be a targeted uniform tax per unit of pollution as this would directly discourage the emissions of pollutants (Perroni and Wigle (1994), Braga (1992) Despite the insights of economic theory, however, many practical questions still remain While being aware of environmental problems and the need for coherence between trade and environment policies, most countries are now engaged in liberalizing their external trade Will the realization of comparative advantages bring with it a risk of specialization in dirty activities? Would implementation of domestic environmental taxes be detrimental to international competitiveness? Empirical research tends to confirm that developing economies specialize in dirty industries (Hettige, Lucas and Wheeler (1995), Low and Yeats (1992), Birdsall and Wheeler (1992)) However, studies do not find strong evidence that stricter environmental regulations of OECD countries have significantly influenced competitiveness (OECD (1993)) This might suggest that developing economies have a real comparative advantage in dirty productions, and hence a trade-off between trade liberalization and environmental preservation could occur. An increasing popular tool for assessing these questions is the general equilibrium policy simulation model (see e g. Lee and Roland-Hoist (1996), Beghin et al (1995)) Its main advantage lies it combining detailed and consistent real world databases with a sound framework of economic theory. Following the same line of investigation, this report offers a quantitative analysis of linkages between economic activity and the environment in Morocco, specifically evaluating the impact of independent and coordinated trade and environmental policies. The Moroccan trade and environment model developed for this analysis (see Annex I for more detailed information) has three features of particular relevance to the present analysis First, it embodies a high level of disaggregation for pollutants, products, sectors and types of households This model can be used, for example, to simulate policies targeted to reduce the loads of specific air emissions, measuring, at the same time, the effect on related water and soil pollutants. Trade policy reform, and its attendant structural adjustment resource reallocation, do not have uniform outcomes across sectors Likewise, expansion or contraction of individual activities does not have uniform environmental -4- consequences Because of its disaggregated economic structure and calibration to the IPPS database (Hettige, Lucas and Wheeler (1995)), this model yields detailed information on a broad spectrum of economy-environment linkages A number of aspects of the Moroccan economy make this type of analysis especially relevant Its relatively high degree of openness rate (about 70 per cent)', make this economy very sensitive to trade reforms Its exports are concentrated in the primary sectors and textiles, whose production requires extensive use of polluting chemical intermediates Policies of emission taxation might impose significant costs here and weaken Morocco's export competitiveness. For analogous reasons, both types of policies could be expected to have important consequences for income distribution between rural and urban households Pressures to implement stronger environmental policies can also arise from Morocco's main trading partners, namely the EC In fact, the new Euro-Mediterranean policy devised by the EC includes conservation policies among the critical issues for sustainable development of the region For example, the latest EU financial policy agreement (1992), although still emphasizing priorities in crop diversification and export promotion, explicitly adds the environment as a new concern. This makes training, technical assistance and investment related to environmental conservation eligible for EC regional assistance. The report is organized as follows. The next section surveys the status quo in terms of existing pollution intensity in Moroccan production and consumption This is followed by discussion of a baseline growth scenario, where macroeconomic stabilization and reform is implemented and its environmental consequences are summarized Section 5. TRADE AND ENVIRONMENT POLICY SCENARIOScontains the main empirical results of the report, with three subsections devoted to different policy perspectives In Section 5.1 LIBERALIZED TRADE WITH THE EUROPEAN COMMUNITY Section 5.2 ENVIRONMENTAL POLICIES taxes in isolation, assessing their consequences in terms of environmental and economic growth objectives. Finally, Section 5.3 COMBINED ENVIRONMENTAL AND COMMERCIAL POLICIES combines the approaches of the other two to examine the prospects for explicit coordination of trade and environment policies Only a small subset of the many possible combined scenarios can be considered within the limits of this report, but the results for policy coordination are very promising. It is apparent even from this preliminary exercise that this kind of research capacity could do much to support more coherent approaches to reconciling growth and environmental objectives The report closes with concluding remarks (Section 6. CONCLUSION references (Section 7 REFERENCES), and a series of technical annexes Defined as the sum of exports plus imports on total income (1995) -5- 3. ECONOMIC ACTIVITY AND THE ENVIRONMENT IN MOROCCO Trade policy has been central to the debate on development strategies in Morocco for a long time, but the environmental dimension of this country's commercial relations has emerged more recently While Morocco has made considerable progress in negotiating improved access to export markets, progressive tightening of European Community and national rules as regards environmental issues, and the harmonisation of EC arrangements for industrial standards, all have the potential to become significant non-tariff barriers to trade Thus, a comprehensive approach to Moroccan-EC trade relations must now consider the environmental implications of economic policy and the economic implications of environment policy The Moroccan trade and environment GE model, which gives a detailed account of direct and indirect economy-environment linkages, is well suited to support this type of policy coordination Before using it to examine the dynamic evolution of trade and environment variables, however, this section assesses the baseline pollution intensities of Moroccan production and consumption. This provides a reference point for evaluating the growth of the economy and its attendant environmental consequences Looking at the economy from a static perspective, Table 2 presents estimates for sectoral emission intensities of production in 1995, i e. the volume of emissions per unit of real output The Moroccan economy has been disaggregated (for a summary presentation) into 9 sectors. food agriculture (FAg), export agriculture (ExpAg), mining and other primary products (Min), food products (FPr), Textiles (Texl), highly polluting manufacturing (PollM), other manufactured products (OthM), polluting services (PollS) and services with low rates of pollution (NPollS) 2 The last column displays economy-wide averages weighted by sectoral outputs, the last 3 rows show (respectively) per cent shares of sectoral production, export to output ratios and EC per cent shares of total exports From this summary table, it is possible to observe the distribution of emission intensities across sectors. This depends on the composition of sectoral intermediate demands (the term aC,, discussed in the Annex) and on the composition of sectoral outputs (the term fi, X""'). A sector would then have a higher pollution intensity (E/X,oU'' ), the more polluting intermediates it consumes and the higher the value of its own 8, coefficient By considering the relative weights shown in Table 2 (last three rows), it is also possible to identify the more polluting industries in volume terms and to infer the environmental consequences of increased market access of Moroccan products in the EC Consider pollution intensity by production sector The aggregate group of polluting manufactures (PollM) represents the worst case in terms of bio-accumulative substances, oxides and air polluting materials A tax proportional to emission intensity would therefore result in relatively higher production costs for this sector Conversely, a specifically 2 Food agnculture consists mainly of basic staple crops (cereals) and livestocks plus oilseeds. Export agnculture includes vegetables, fruits, fishing and other agricultural products Highly polluting manufacturing aggregates pulp and paper, chemicals, oil refinery, iron and steel and construction materials industnes Polluting services (PollS) are transports, electric power, and construction -6- targeted tax on toxic emissions and BOD substances would more significantly affect costs in polluting services (PollS) Export agriculture is notably more polluting than food agriculture (with the exception of bio accumulative substances) Textiles, which record the highest EC share of total exports enjoy relatively low emission intensities, whereas Mining shows the highest export to output ratio and fairly severe emission intensities. Table 2: Sectoral emission intensities for output - 1995 (kg per billion 1990 Dirham) FAR ExPAR Mm FPr Texl PollM OthM PollS NPollS Total TOXAIR' 1 0 1 3 3 8 27 140 197 30 195 2 5 8.3 TOXWAT 3 5 49 67 54 07 24 1 45 33 1 54 11.2 TOXSOL 24 3 1 147 67 05 992 100 1291 73 36.2 BIOAIR' 1 2 07 39 5 1 5 0 1 6084 347 1536 09 1298 BIOWAT 12 1 5 22 05 0 1 330 03 143 42 9.3 BIOSOL' 254 156 8242 294 06 127786 211 5 3161 3 165 27002 S02 52 7.2 110 88 03 34 8 2 1 93 48 108 N02 30 42 64 51 02 238 13 54 28 6.9 CO 05 07 42 09 00 537 02 131 05 116 VOC 13 19 35 38 04 82 33 24 24 33 PART 07 10 18 12 00 100 03 24 07 2.6 BOD 0 1 0 1 1 3 47 00 58 1 0 172 08 3.6 TSS 24 15 78 1 27 00 12196 00 2967 14 2564 Output % 11 4 2 8 9 18 4 11 34 100 X/Output 0 28 59 11 21 12 11 6 2 EEC X % 65 69 53 75 85 47 72 65 65 I Emission categones are defined in the Annex 2 Boo-accumulative pollution intensities are in grams per billion of Dirham In terms of volume, emissions from polluting manufacturing and services represent 82 per cent of the total toxic emissions, 99 per cent of total bio-accumulative pollutants, 80 per cent of total oxides, 65 per cent of other air pollutants and 99 per cent of other water pollutants Although production is the dominant environmental player in the economy, consumption of goods and services also causes significant pollution, particularly in some emission categories The analogous intensity results for consumption are given in Table 3 These estimated intensities cover only final goods and do not consider packaging or household garbage Except for water bio-accumulatives, consumption pollution is ascribed only to polluting manufactured products, mainly in association with consumption of refined fuels or chemicals Bio-accumulative metals and toxic waste released through consumption, similarly to production, are associated with soil degradation. -7- Table 3: Sectoral emission intensities of final consumption - 1995 (kg per billion 1990 Dirham) FAg ExpAg Mm FPr Texi PoIIM OthM PollS NPoIIS Total TOXAIR 00 00 00 00 00 188 00 00 00 19 TOXWAT 00 00 00 00 00 440 00 00 00 45 TOXSOL 00 00 00 00 00 387 00 00 00 39 BIOAIR' 00 00 00 00 00 45 00 00 00 05 BIOWAT 00 00 00 00 00 00 00 2753 00 393 BIOSOL' 00 00 00 00 00 377 00 00 0.0 38 S02 00 00 00 00 00 59 1 00 00 0.0 60 N02 00 00 00 00 00 344 00 00 0.0 35 CO 00 00 00 00 00 51 00 00 00 05 VOC 00 00 00 00 00 243 00 00 00 25 PART 00 00 00 00 00 82 00 00 00 08 BOD 00 00 00 00 00 26 00 00 00 03 TSS 00 00 00 00 00 00 00 00 00 00 Cons% 16 6 0 10 4 10 11 14 28 100 I Bio-accumulative pollution intensities are in grams per billion of Dirbam -8- 4. THE MACROECONOMIC REFORM SCENARIO Two major policy issues will influence Moroccan economic growth over the next decade macroeconomic stabilization and trade liberalization. Our evaluation of trade-environment linkages for this country will be done with respect to a baseline scenario of macroeconomic growth over the period 1995-2005 This section describes the baseline and the environmental significance of growth under a scenario of macroeconomic reform The sustainability of current fiscal trends in Morocco has been questioned by a variety of financial authorities A synopsis of macroeconomic reforms currently under discussion between the Bank and the government of Morocco is given in Roland-Hoist and Belghazi (1996), along with detailed numerical results are presented which contrast the effects of the proposed macro reforms with a do nothing approach.3 As the latter policy is generally agreed to be unsustainable, we take the macro reforms as the status quo trend over the next ten years The macroeconomic reforms have been extensively analyzed with the Bank's own RMSM-X simulation model. This yielded a series of annual projections for detailed national income accounts, trade, and monetary aggregates that reveal the macroeconomic implications of the Medium-term Economic and Financial Strategy While it is not as well-suited to macro forecasting, the GE model provides more detailed information on the real side of the economy, and the results in the next section are intended to elucidate the incidence of the proposed macro reforms To implement the GE model in the context of these reform scenarios in a consistent manner, it was calibrated to selected RMSM aggregates over a specific time interval These macro aggregates included real GDP, growth in government expenditure, productivity growth, foreign capital inflows, and world market price trends. In particular, RMSM assumptions were implemented about significant reductions in government expenditure on current and capital account, sharply rising foreign capital inflows, and GDP growth targets The GDP targets were achieved by endogenous computation of Moroccan capital productivity growth over the period 4 Although the two models can be calibrated closely, differences in their structure will inevitably lead to some divergence in detailed accounts 3 See also World Bank (1992, 1994abc, 1996), and IMF(1994) It is worth emphasising that, once calibrated, productivity growth is exogenous to the model. so it does not change under any of the policy scenarios Thus we do not capture so-called endogenous growth effects which might arise from policy reforms -9- The interval chosen was 1995-2005, the period of most immediate interest in policy discussions. Because of the fundamental nature of the proposed reforms, it is reasonable to expect that any trends established in this period will be extended beyond the next decade This is particularly true because the underlying growth strategy is based on increased investment and removal of domestic distortions, both of which will contribute to the sustainability of the reform measures For this reason, growth results reported for the simulations are likely to understate the ultimate benefits attributable to the complete reform package 4.1 GROWTH AND EMISSIONS Corresponding to the baseline economic scenario is a reference trend for emission. The joint evolution of economic activity and emission loads under the macro reforms can be seen Table 4, which presents long term pollution elasticities with respect to production and consumption for each effluent category These measure the per cent increase in a given type of emission load that results from a one per cent increase in either aggregate output or consumption. All are obtained from annual averages, under the Macro reform scenario, over the period 1995-2005. In particular, we assume no specific trade or emission tax policies are implemented Note that aggregate pollution grows more or less at the same rate as aggregate economic activity, i.e the elasticities are close to unity This result is not unexpected, since the medium term effect of macro reforms is to expand economic activities in a relatively uniform manner. Under such homogenous economic expansion, the composition of emissions would remain relatively constant, with loads increasing in direct proportion to aggregate output or consumption, respectively. Table 4: Emission elasticities - Macro reform scenario 1995-2005 Output Consumption TOXAIR 107 104 TOXWAT 101 101 TOXSOL 1.13 105 BIOAIR 1.15 1.36 BIOWAT 1.13 1.37 BIOSOL 1.15 1.36 S02 0.74 098 N02 078 098 CO 1 12 0.98 VOC 0.90 1 02 PART 0.92 098 BOD 1.11 1 12 TSS 1.15 0 The analysis of the decomposition of emission by origin, as shown by Grossman and Krueger (1992) can be instructive Three types of effects are distinguished within the variation of emission levels 1) the composition effect takes into account the modification of the proportion of polluting products in the aggregate output - 10- 2) the technological effect reflects changes in pollution due to alteration in the production technology 3) the scale effect describes the impact of increased volumes of output on the environment The estimated determinants total production and consumption emissions are presented displayed in Table 5 Observing the values in the scale effect column, one clearly sees the predominant role in environmental degradation (across all types of emission) attributable to an increase in the scale of economic activity. This is logical in the present scenario, as are the near-unit elasticities above, since the macro reforms tend to expand the economy in a relatively uniform manner The composition effect, the proportion of polluting goods and services, or the average pollution intensity, expands from 1995 to 2005 in both production and consumption This increases aggregate pollution volumes in all categories except sulfur and nitrogen oxide emissions. Finally, production technologies appear to be cleaner at the end of the period, specifically because of the improvements derived from the assumed gains in the energy efficiency factor (technology effect).5 Bio-accumulative emissions are exceptional again, registering very low magnitudes. The actual mechanics of the technology effect in this model require some elaboration The specification of production technology is discussed in detail in the appendix (see especially Figure 1), but it is worth emphasizing its main features. In this economic model, the concept of technology refers not to an engineering specification of some (eg ) machine, but to a given mix of inputs and factors used to obtain output. There is some substitution between the two groups, but the intermediate bundle contains fixed proportions of inputs Factors are substitutable between three groups, labor, capital, and energy, where only the latter creates pollution in use The energy factor is further decomposable into petroleum and electrical substitutes, each of which has different pollution characteristics. Thus a firm can reduce pollution at any of these three levels, substituting intermediates and factors, substituting energy and non-energy factors, and substituting between energy sources All such substitution is dictated by the relative prices of each constituent (input or factor), and these are in turn determined by relative scarcity and taxes/subsidies For example, in the macro reform scenario, real exchange rate depreciation raises the cost of imported oil, and this induces a shift of energy demand to electricity, reducing the pollution intensity of domestic production (column 3 of Table 5). This effect is particularly evident in the hydrocarbon gas categories (S02, N02), emissions of which grow more slowly that aggregate output or consumption (Table 4). While such a specification allows for a relatively complex adjustment process, it should be emphasized that important features of the pollution/technology nexus are not explicitly captured here There are, for example, no specifications of technological innovation or transfer that explain how changing input/factor mixes are achieved. Secondly, we do not It is worth emphasising that, in its present form, the model captures three sources of technological gains 1) substitution of value-added (labor, capital, and energy) for other polluting inputs, 2) new capital acquisition 3) energy source substitution (electricity or petroleum-based fuel) -11- explicitly consider the possibilities of abatement, clean up, or other end-of-pipe technologies We simply assume that domestic firms respond to relative prices of inputs/factors by changing their mix with respect to some empirical substitution elasticities In summary, with no policy intervention, economic growth results in a significant increment of emissions, despite output and consumption shifts towards less polluting products and implementation of cleaner technology Table 5: Decomposition analysis of emission variations, 1995-2005 (Macro reform scenario) Production Consumption Composition Technology Scale Composition Technolopv Scale Variations in Volumes (1000 metric tons) TOXAIR 1 7 -03 280 0 1 00 26 TOXWAT 1 7 -14 363 00 00 60 TOXSOL 120 -07 125 1 02 00 54 BIOAIR' 500 -07 4549 02 00 07 BIOWAT' 3 1 00 322 162 00 61 9 BIOSOL' 10529 -14 1 94693 1 6 00 60 S02 -ll -50 306 -0 1 00 80 N02 -05 -29 200 -0 1 00 46 CO 40 -05 400 00 00 07 VOC 0 1 -08 10 1 01 00 3 3 PART 02 -07 80 -0.0 00 1 1 BOD 10 00 124 00 00 04 TSS 1005 -13 8994 00 00 00 Variations in percentages TOXAIR 6 -1 95 3 0 97 TOXWAT 5 -4 99 1 0 99 TOXSOL 9 -1 92 4 0 96 BIOAIR 10 0 90 20 0 80 BIOWAT 9 0 91 21 0 79 BIOSOL 10 0 90 20 0 80 S02 -4 -21 125 -1 0 101 N02 -3 -18 120 -1 0 101 CO 9 -1 92 -1 0 101 VOC 1 -8 108 2 0 98 PART 3 -9 106 -1 0 101 BOD 7 0 93 8 0 92 TSS 10 0 90 0 0 0 1) Bio-accumulative pollution intensities are in metnc tons - 12- 5. TRADE AND ENVIRONMENT POLICY SCENARIOS Given the baseline growth path for the Moroccan economy, we now examine interactions between trade and the environment, environmental policy and the economy, and finally the effects of jointly implementing both policies. In most cases, the economy-environment linkages are significant They are also complex enough that policy makers relying on intuition or rules of thumb would have difficulty anticipating them In particular, we find that trade liberalization of the kind being implemented in the Moroccan FTA with the EC will, in the absence of mitigating environmental policies, increase both the levels and intensities of important pollutants Secondly, targeted pollution abatement policies can achieve significant abatement, but with unintended and detrimental effects on economic growth. Third, policies which coordinate the two can achieve a measure of both objectives, significant economic growth with reduced emissions and concomitant environmental damage Since the FTA will ultimately entail greater harmonization with EC emission standards, it is reasonable to expect Morocco to implement some form of coordinated policy regime The first round of simulations consider a progressive reduction of each type of emission (i e it consists of 13 different experiments). A target in terms of emissions abatement is exogenously fixed as follows. Emissions levels are reduced with respect to the reference scenario by 2 per cent in 1992, 8 per cent in 1998, 17 per cent in 2001 and 25 per cent in the end of the period The instrument used to reach this target is a uniform tax per unit of emission paid by the agent causing that pollution and endogenously determined by the model 6 The second scenario simulates a policy of trade liberalization through a progressive reduction in import tariffs These international trade distortions are decreased ad valorem by 5 per cent with respect to the Macro reform case in 1992, by 27 5 per cent in 1998, by 58 5 per cent in 2001 and completely abolished at the end with a perfect convergence of domestic and international prices 5.1 LIBERALIZED TRADE WITH THE EUROPEAN COMMUNITY Moroccan import protection is relatively high by international standards and very unequal across economic activities While these distortions do not appear large on an import- weighted average basis, they confer significant price advantages upon selected domestic activities, particularly in agriculture This section examines effects of concluding the Free 6 See the technical specification of the model for more details - 13 - Trade Agreement between Morocco and the European Community Assuming the macro reforms and fiscal stabilization of the last section, we simulate abolition of Moroccan import restraints against goods from the EC This liberalization follows the letter of the agreement, phasing out tariffs in four years and NTBs in six. Some important differences characterize the simulated FTA, however First, we do not exempt agricultural products as is done in the current version of the agreement. Secondly, we assume Morocco implements the FTA with explicit provision for an alternative revenue source to offset declining trade taxes, i.e a revenue-neutral uniform (RNU) VAT. This is done to mitigate adverse fiscal effects and to better isolate the pure effects of the trade policy. To render more legible the results presented in the following tables, emissions are aggregated in five groups: toxic pollutants (TOXAIR, TOXWAT, TOXSOL), bio- accumulative metals (BIOAIR, BIOWAT, BIOSOL), oxide emissions (SO2, NO2, CO), other air pollutants (VOC, PART) and other water pollutants (BOD, SS). These aggregations are consistent in physical terms and their constituents tend to move in the same direction In fact, emissions show a high degree of correlation within each group. The five are then presented (column headings) twice, for production and consumption respectively, with respect to each policy (row) considered For example, the first row of Table 6 summarizes the emission elasticities for the baseline or macro reform scenario.9 Recall that numbers in this table refer to percentage changes in the relevant emission load (column), given a one-percent change in aggregate production or consumption, under each of fifteen hypothetical policy scenarios (rows) Thus the elasticities measure the estimated pollution intensity of aggregate production and consumption The second row of the table contains analogous results for the FTA, and the differences are arresting Evidently, trade liberalization with respect to the EC, in the absence of mitigating environmental measures, will increase the pollution intensity of Moroccan supply and demand in most categories of emissions. Indeed, in some cases, emissions intensities rise appreciably as a result of external opening The main reason for this is intensification of Morocco's traditional comparative advantage, particularly in activities related to export agriculture, mining, and polluting manufacturing. For comparison with another GE simulation of Morocco's FTA, the reader may consult Rutherford et al (1996) 8 In the following sub-sections the results from the above simulations are presented. For the sake of clanty only global or very aggregated results are shown, even though the model is run with about fifty sectors over five two-year time intervals These results are weighted averages of those in Table 4 above and are presented here for reference - 14- Table 6: Emission elasticities. Independent trade and environment policies Emission by Group With respect to production With respect to consumption Policv Tox Bio Nox Air Wat Tox Bio Nox Air Wat Baseline 109 1.15 091 091 1 15 1.03 1.37 098 101 1 12 FTA 1.26 1 63 1.11 1.05 164 1.15 143 100 1 11 1.17 Enussion Taxation TOXAIR 067 0 20 0.49 0.63 0 20 0 73 1 05 0 95 0.87 0.52 TOXWA T 059 043 0.16 036 043 0.52 0.80 0.73 080 020 TOXSOL 068 0.32 0.50 064 031 0.77 0.96 0.97 094 0.45 BIOAIR 092 062 070 0.77 0.62 096 1 29 0.98 097 1.07 BIOWA T 073 054 060 0.71 053 096 080 0.98 097 1.08 BIOSOL 0.92 064 071 077 063 0.96 1 30 098 097 1 07 S02 1 03 1 16 0.57 0.63 1 16 0.92 1 28 067 086 0.94 N02 100 1 14 055 0.62 1 14 0.90 123 067 086 085 CO 0.91 059 068 075 0.59 0.95 129 097 096 1.06 VOC 085 041 0 17 0.29 041 0.73 1 34 0.63 057 1.11 PART 081 0.49 0.33 048 049 084 1 19 073 083 0.88 BOD 0.77 097 0.70 078 096 083 075 1 00 102 0.13 TSS 093 0.65 0.71 078 0.64 096 1.30 098 097 1.07 Notes Tox toxic pollutants, Bio bio-accumulatives metals. Nox sulphur, nitrogen and carbon oxides, Air other air pollutants, Wat other water pollutants Table 7 reveals the composition of the FTA-induced pollution changes In particular, note that both composition and technology effects contribute to increased pollution elasticities, with the former being the larger component In other words, Moroccan specialization in polluting activities is mainly explained by a shift in the output composition towards more pollution-intensive products, rather than by the use of dirtier technologies across all industries This composition effect reflects the fuller exploitation of Moroccan comparative advantage in polluting sectors The technology effect is generally positive under the FTA for two reasons. Firstly, import liberalization lowers the relative prices of many polluting inputs, including chemicals. Secondly, the price petroleum products shipped from the EC falls relative to both petroleum from elsewhere and domestic electricity The first of these effects induces trade diversion, the second substitution in technology for a more pollution-intensive energy source The scale effect plays the most important role in absolute terms, significantly raising aggregate emissions levels Analogous reasoning applies to consumption, but the changes in emissions are smaller because the composition of domestic consumption is less responsive to the FTA. - 15 - Table 7: Decomposition of emission variations due to the FTA, 1995-2005 Production Consumption Composition Technolog Scale Composition Technology Scale Variations in Volumes (1000 metric tons) TOXAIR 69 0 1 363 04 00 3 1 TOXWAT 1 9 -06 42 1 0.7 0.0 71 TOXSOL 316 14 1596 07 0.0 63 BIOAIR 2506 43 6814 04 00 10 BIOWAT 145 04 46 1 19.3 00 71 2 BIOSOL 5413 5 899 14331 8 34 00 79 S02 -3 5 -39 33.9 0.0 00 8.9 N02 -22 -23 220 0.0 00 5.2 CO 214 00 597 00 00 0.8 VOC 05 -05 121 04 00 40 PART 14 -0.5 102 0.0 00 12 BOD 05 02 139 0.1 0.0 0.4 TSS 523 5 8.5 1368 3 0.0 00 00 Variations in percentages TOXAIR 16 0 84 12 0 88 TOXWAT 4 -1 97 8 0 92 TOXSOL 16 1 83 10 0 90 BIOAIR 27 0 73 30 0 70 BIOWAT 24 1 75 21 0 79 BIOSOL 27 0 72 30 0 70 S02 -13 -15 128 0 0 100 N02 -12 -13 125 0 0 100 CO 26 0 74 0 0 100 VOC 4 -5 100 9 0 91 PART 13 -5 92 0 0 100 BOD 4 1 95 11 0 89 TSS 28 0 72 0 0 0 1) Bio-accumulative pollution intensities are in metric tons It thus appears that trade liberalization, of the type contemplated by Morocco with respect to the EC, will move the country toward higher degrees of domestic pollution intensity. Not only will total emissions grow, but in the absence of abatement policies they will grow significantly faster than output and consumption This result is by no means inevitable for two reasons Firstly, in some countries, trade liberalization appears to shift economic activity toward less pollution intensity 10 Second, policy coordination with targeted economic instruments holds to promise to mitigate pollution without sacrificing growth objectives The next two sections will better elucidate the second option In closing this section, however, it must be noted that these results are probably too pessimistic, since relatively limited scope for technological improvements has been allowed To better elucidate the linkages between trade liberalization and emissions, Table 8 presents results on changes in output, consumption, and total emissions by aggregated 10 See e g Beghin, Roland-Holst, and van der Mensbrugghe (1994 and 1997) - 16- sector group." Even at this relatively aggregate level, the sources of pollution arising from trade liberalization are apparent. On the production side these include export agriculture, mining, textiles, and polluting manufactures These generally represent an intensification of Morocco's traditional comparative advantages, such as tree crops (and upstream agrochemicals) and mining As the share estimates in the last three columns indicate, the pollution intensity of mining is well above average Export agriculture's direct pollution does not appear to be above average, but its upstream linkage to agrochemical production induces significant industrial pollution. Such activities would be attractive candidates for more targeted pollution taxes. Consumption growth induces added emissions from polluting manufactures and polluting services (especially transportation and electricity) These are more typical emergent pollution groups in a growing economy, and would appear to merit increased regulatory vigilance Table 8 Sectoral Adjustments Resulting from the FTA (percentages) Real Real Total Output Emission Share Output Consumption Emissions Share Share Ratio FoodAgri 032 1 30 075 993 0 13 0.01 ExpAgri 7 98 1 90 999 407 005 001 Mning 12 51 0.00 15.00 262 0 75 0.29 FoodPr -2.69 4 19 -004 722 0 14 002 Textile 14 15 674 15 84 8.17 004 000 PollMan 757 1067 1049 1836 8476 462 OthMan -476 1965 -434 3 81 030 008 PollServ 699 623 924 11.48 13 35 1 16 NPollServ 5 72 2 27 7 96 3435 048 001 Total 1 000 10000 The first three columns represent differences with respect to the baseline scenano in 2005 Shares in the last three columns are for the 1995 base The last column contains ratios, not percentages 5.2 ENVIRONMENTAL POLICIES Despite the apparently negative environmental consequences of FTA-type trade liberalization for Morocco, it is now well established that import protection is an inefficient way to promote environmental ends 12 Morocco is already committed to developing economic instruments for pollution abatement, so it is reasonable to ask how these can help fulfill the economic promise of trade liberalization while offsetting its negative environmental side-effects To better understand the efficacy of such economic instruments, we use the model to evaluate the economic and environmental effects of targeted pollution taxes. The results indicate that in Morocco there is significant scope for pollution abatement with these policy tools I Table 12 in the Appendix details the same structural adjustments at the full sectoral disaggregation 12 See Beghin, Roland-Hoist, and van der Mensbrugghe (1996) for a recent survey of this issue - 17- Referring back to Table 6, the last thirteen rows correspond to hypothetical pollution taxes In particular, for each of the corresponding thirteen effluent types, a uniform (across producers and consumers) tax was levied on emissions of that effluent with the objective of achieving 25 per cent aggregate abatement against baseline emissions 1 For example, a uniform tax is levied on water pollutants (TOXWAT) so that 25 per cent abatement with respect to the Macro reform is the target for the year 2005. In this case, average yearly growth rates for bio-accumulative emissions in the period 1995-2005 would be equal to 0 43 times the average growth rate in production or 0 80 times the corresponding growth rate for pollution originating in consumption Clearly, such targeted abatement taxes have significant potential to mitigate pollution. It is also noteworthy that they exhibit strong positive interactions, i.e when only one emission is taxed, other emissions may drop significantly. In many cases, abatement in secondary pollution groups exceeds that of the target group This indicates that benefits of individual instruments may far exceed their direct pollution objectives and, just as importantly in a country like Morocco, such linkage can be exploited to reduce the number of instruments and their associated administrative costs. Aggregate reduction in the emission loads results primarily from decreases in production generated emissions This is in turn due to a shift of production towards less polluting activities and a shift within activities toward cleaner mixes of inputs. A more detailed analysis effects would show significantly lower output for those sectors producing highly polluting goods (e.g. PolIM), up to 20 per cent with respect to the reference scenario in the year 2005 (composition effect). It would also indicate that emission abatement in the other industries is obtained through diminished individual pollution intensity (technology effect), as the result of substituting toxic intermediates for more labor, capital, and cleaner energy. 14 Table 9 Real GDP, export and import growth rates (per cent average yearly rates 1995-2005) Baseline Tox Bio other ave FTA Air Wat Sol Air Wat Sol S02 N02 CO VOC PART BOD TSS RGDP 4.0 39 38 38 40 38 40 41 40 40 39 39 38 40 3.9 4.9 Export 4.7 4 1 44 43 44 43 4.4 4 4 4 4 4 44 45 44 44 11.1 Import 3.5 30 33 32 32 32 3.2 3 5 35 32 33 33 34 32 33 28.5 13 That is, the tax rate is computed endogenously for an exogenous abatement target This acts just like a tradeable pollution rights mechanism, where the equilibnum pnce of the pollution permits is the implied tax. 14 In these scenarios, domestic savings are higher when abatement policies are implemented, even with no anticipation of future emission taxes by the agents. Revenues from these taxes are assumed to be redistributed to households in proportion to their baseline income taxes This revenue redistribution scheme, combined with the correspondence of high income tax rates and high savings, results in faster capital accumulation and growth for the economy - 18- Given the efficacy of these instruments in terms of reducing pollution intensity, what then are their growth costs? Quite low, judging from the results of Table 9 Most of the abatement policies reduce the growth due to macro reform by less than one percentage point, and in many cases the redistributive effects of the effluent taxation appear to completely offset losses due to the tax distortion " The average yearly GDP growth rate in the simulations is found in the range of 3.8 and 4.1 per cent, very close to the Macro reform rate of 4 0 per cent It should also be noticed that these policies do not significantly affect Moroccan external competitiveness, for its aggregate export (in real term) growth rate decreases at most (in the case of an air toxic emission abatement policy) by 0 3 percentage points with respect to the Macro reform Under the FTA, by contrast, real exchange rate depreciation makes exports more competitive (rising 11. 1 percent annually), while abolition of barriers to EC goods boosts imports by over 28 percent. These policies are much less effective in reducing emission loads generated through final consumption This can be explained by two factors Firstly, in the model households reactions to emissions taxes are sluggish, in the sense that there is no 'technological" improvement in their consumption pattern. This slow reaction is also confirmed by empirical data observed for OECD countries, and is primarily due to the households' low replacement rate of durable equipment (housing, vehicles), the main source of pollutants (fuels, chemicals) Secondly, most of the tax burden (as a per cent of taxpayer income) is on producers, these being the main polluters through their production activities 5.3 COMBINED ENVIRONMENTAL AND COMMERCIAL POLICIES The growth benefits of trade liberalization and the environmental benefits of targeted economic instruments invite the prospect of coordinating these policies. This would be true in any economy seeking optimally to pursue both objectives, but it should be particularly attractive to Morocco because it has already committed itself to an external policy (the FTA) that will apparently have negative environmental consequences. What then are the possibilities for using trade policy and pollution taxes in concert? This section presents a partial menu of hypothetical cases that combine FTA liberalization with targeted emission taxation. Even these limited results indicate that significant scope exists for realizing the growth potential of expanded external markets while simultaneously reducing aggregate emission loads. Apart from the intuitive appeal of such policy coordination, Moroccan trade liberalization with no associated measures of pollution control seems unsustainable for three additional reasons On one hand, given the current patterns of Moroccan comparative advantage, tariff elimination will favour pollution-intensive sectors. This tendency could be exacerbated if Moroccan trade partners, among whom the EC plays a major role, tighten 15 Note that these abatement taxes are not second. or probably even third, best Here they are implemented in the context of many other distortionary instruments, including import taxes and existing Moroccan producer taxes. - 19- their own environmental policies In these countries production costs of locally polluting goods can be expected to rise, further promoting cheaper imports from Morocco 16 Secondly, anti-pollution pressures might originate from neighbouring countries, inducing the Moroccan government to adopt higher environmental quality standards Over time, such demands could also intensify locally, as pollution levels and domestic incomes rise in tandem and social tolerance of the former diminishes. All these factors contribute to make coordination of trade liberalization and abatement policy a more realistic long-term scenario for Morocco Third, there is rising pressure to harmonize environmental standards under the FTA, and Morocco would be well advised to anticipate this by identifying activities for voluntary compliance Table 10 presents the pollution elasticities for the baseline scenario and thirteen experiments with combined trade and emission-specific pollution taxes For the latter group, we simply consider the FTA in concert with each of the abatement policies simulated in the previous section Strictly speaking, these policies are not coordinated, i.e they represent a menu of examples rather than a specifically designed policy mix. It is unlikely that any coordinated policy would be confined to one emission alone, but we make no attempt in the present analysis to devise or propose an optimal mix of trade and emission tax reforms Looking at the interactions of trade reform and emission taxation one at a time, the incremental contribution of each type of tax, as well as its indirect effect on other pollution, are made more transparent The results of a combined approach appear promising, indicating significant reductions in emissions could be attained together with most of the trade-induced gains in real income and consumption Invariably in these experiments, individual pollution taxes mitigate the negative environmental effects of trade liberalization, i e. pollution elasticities are below the values registered when the FTA is implemented without emission taxes. In some cases, combined policies even achieve elasticities below those of a pure abatement policy (compare for instance row "TOXAIR"in Table 6 with the corresponding row in Table 10), a win-win combination In other cases, abatement is more effective without trade liberalization (oxides) 16 A country exporting pollution-intensive goods is simultaneously exporting a pollution service, retaining the effluent content of production at home and spanng the importing country this environmental cost There is a significant component of this environmental asymmetry in North-South trade, quite independently of the pollution haven phenomenon, which deals mainly with capital flows (see Low: 1992). For more on the this notion, see Lee and Roland-Hoist (1996) Note also that this effect is distinct from the pollution haven effect, which is a capital market response to differential environmental standards - 20 - Table 10: Emission elasticities - Coordinated trade and environmental policies Emission by Group With respect to production With respect to consumption Poliev Tox Bio Nox Air Wat Tox Bio Nox Air Wat Baseline 1.09 1 15 091 091 1 15 103 1 37 098 1.01 1.12 FTAonly 126 163 111 105 164 115 143 100 111 117 TOXAIR+ 063 008 043 059 008 083 1 11 0.97 096 062 TOXWAT+ 058 040 0 12 032 040 0.62 087 078 090 030 TOXSOL+ 067 028 044 058 028 090 103 099 104 060 BIOAIR+ 089 061 065 072 061 1.05 1 30 1.00 106 1 12 BIOWAT+ 069 045 052 065 045 1 05 083 100 106 1 13 BIOSOL+ 090 061 0.65 073 062 105 131 1.00 106 1 13 S02+ 1.18 1 59 079 078 1 59 105 134 069 096 1.02 N02+ 1.16 1.58 078 078 159 1.04 1 31 071 097 0.95 CO+ 088 057 062 070 057 105 130 099 105 112 VOC+ 078 026 006 021 026 082 1 34 068 066 1 17 PART+ 079 046 031 046 047 097 122 080 095 100 BOD+ 091 137 085 087 137 095 084 102 112 020 TSS+ 090 062 066 0.73 062 105 131 100 106 113 Notes Tox toxic pollutants, BRio bio-accumulatives metals, Nox sulfur, nitrogen and carbon oxides, Air other air pollutants, War other water pollutants The + notation means the pollution tax is implemented in concert with the FTA Emission elasticities with respect to consumption are almost always slightly higher when abatement is undertaken in concert with trade liberalization. In other words, the share of polluting products demanded under the FTA are highest, under pure emission taxation lowest, and somewhere in between when the policies are combined. This happens because the FTA induces both household income growth and a shift of consumption towards more 17 pollution-intensive products Output of polluting products (PollM) increases in response to intensified Moroccan comparative advantage. In the case of combined policies, this composition effect can be less important, being partially offset by factor substitution that reduces emission loads per unit of output In fact, this technology effect is reflected in the substitution of toxic intermediates for more labor and capital value added. Larger savings, generated from increased household incomes, reduce capital rental rates and allow enterprises to make the new investments necessary to reduce emissions without excessively raising output costs It is worth emphasizing that coordinated policies of the type discussed above will inevitably emerge during the course of FTA implementation This is because a combination of domestic and external forces will impel Morocco toward harmonization of its environmental standards and policies with those of its main trading partner For this 17 The income elasticities of consumption for services used in the model are greater than unity This reinforces consumption shifts towards polluting services when income increases as a result of trade liberahLation An analogous argument holds for pollution intensive manufactures These consumption trends are typical of semi-industrialized countries with nsing incomes -21 - reason, the basic FTA results of Section 5 1 above are too pessimistic. The existing agreement is not specific enough in this regard, however, to model harmonization more explicitly For this reason, the model has simply been used to survey a menu of alternatives for policy coordination Ultimately, this research capacity could be used to support design and implementation of policies when they become more well defined - 22 - 6. CONCLUSIONS From a practical standpoint, there are two general lessons from this research. Firstly, it is apparent that there are significant linkages between trade and the environment in Morocco, and a coordinated approach to these two types of policy is essential to avert undesirable secondary effects Secondly, these linkages are quite complex, and it is unlikely that policy makers relying on intuition or rules of thumb alone will devise or implement anything approaching optimality More specifically, the results obtained above indicate that a free trade agreement of the kind Morocco negotiated with the EC will, in the absence of offsetting environmental policies, shift the economy toward more pollution-intensive production and consumption patterns In some major effluent categories, pollution can be expected to grow up to 65 percent more rapidly than GDP Thus the significant economic gains derived from trade liberalization can lead to environmental degradation In our analysis of independent abatement policies, the results identify three main advantages of specific taxes targeted to emission abatement First, sectoral output and emission loads no longer grow at same rates Secondly, the economic (growth) costs of these policies appear to be quite low. Third, even when only a single effluent is taxed, little substitution towards other pollutants is induced This is a direct consequence of linking emissions to input use rather than simply to output. A final set of simulation results illustrated the advantages of coordination between environmental and commercial policies When implemented in concert with the FTA pollution taxes can achieve significant abatement without offsetting most of the growth benefits of trade liberalization. In particular, the two policies can encourage more pollution efficiency in traditional sectors and alter patterns of specialization in favor of new, cleaner activities. In this sense, there is reason for optimism that careful policy coordination can indeed reconcile growth and environmental objectives The current model specification probably understates the potential gains of well-conceived abatement policies, given that it does not take into account the relationship between factor productivity and environment degradation, the introduction of new technologies, and household preferences towards the environment Even with these limitations, however, the present findings strongly support the implementation of economic instruments in an outward-oriented economy - 23 - 7. REFERENCES Anderson, Kym and Richard Blackhurst, eds (1992), The Greening of World Trade Issues, Ann Arbor University of Michigan Press Beghin, John, S6bastien Dessus, David Roland-Hoist, and Dominique van der Mensbrugghe (1996), "General Equilibrium Modeling of Trade and the Environment," Technical Paper No 116, OECD Development Centre, Paris, September Beghin, John, David Roland-Hoist, and Dominique van der Mensbrugghe (1994), "A Survey of the Trade and Environment Nexus Global Dimensions," OECD Economic Studies, 23, 167-192, Winter, 1994. Beghin, John, David Roland-Hoist, and Dominique van der Mensbrugghe (1995), "Trade Liberalization and the Environment in the Pacific Basin Coordinated Approaches to Mexican Trade and Environment Policy," American Journal ofAgricullural Economics Beghin, John, David Roland-Holst, and Dominique van der Mensbrugghe (1997), "Trade and Pollution Linkages Piecemeal Reform and Optimal Intervention," Canadian Journal of Economics. Birdsall, Nancy, and David Wheeler (1992), "Trade Policy and Industrial Pollution in Latin America. Where are the Pollution Havens?" in Low, P. (ed.), 159-168 Braga, C.A.P (1992), "Tropical Forests and Trade Policy The Case of Indonesia and Brazil," in Low, P (ed.), pp 173-195 Burniaux, J-M, Nicoletti, and J Oliviera-Martins (1992), "GREEN A Global Model for Quantifying the Costs of Policies to Curb CO2 Emissions," OECD Economic Studies, 19, 49-92 Bussolo, M., and D.W Roland-Holst (1994), 'A Social Accounting Matrix for Morocco," Unpublished manuscript, OECD Development Centre, Paris Bussolo, Maurizio, Roland-Hoist, David, and Dominique van der Mensbrugghe, 1995, 'A Dynamic General Equilibirum Model of Morocco," Unpublished manuscript Cropper, Maureen L , and Wallace E Oates (1992), "Environmental Economics. A Survey," Journal of Economic Literature, XXX, 675-740, June Grossman, G M, and A B Krueger (1992), "Environmental Aspects of a North American Free Trade Agreement," Working Paper No 644, CEPR, April Dessus, S6bastien, David Roland-Holst, and Dominique van der Mensbrugghe (1994), "Input- based Pollution Estimates for Environmental Assessment in Developing Countries," Technical Paper No 101, OECD Development Centre, Paris - 24 - Hettige, Mala, R E B Lucas, and David Wheeler (1995), "The Industrial Pollution Projection System," Policy Research Worlang Paper 1431, The World Bank, Washington, D C International Monetary Fund, Morocco - A Review of Adjustment Experience, IMF Mimeo, June 1994 Lee, Hiro, and David Roland-Holst (1996), "The Environment and Welfare Implications of Trade and Tax Policy," with Hiro Lee, Journal ofDevelopment Economics Low, P (1992), International Trade and the Environment, Discussion Paper No 159, The World Bank, Washington, DC Low, P, and A Yeats (1992), "Do Dirty Industries Migrate?" in Low, P (ed). OECD (1993), Environmental Policies and Industrial Competitiveness, OECD Documents, Paris Perroni, C , and R M Wigle (1994), 'International Trade and Environmental Quality How Important are the Linkages?" Canadian Journal of Economics. Radetzki, M (1992), "Economic Growth and the Environment," in P. Low (ed). Roland-Hoist, D., and S Belghazi (1996), "Morocco An Economywide Analysis of Selected Policy Reforms," report prepared for the World Bank, December. Rutherford, T F., E E. Rutstrom, and D Tarr (1996), "Morocco's Free Trade Agreement with the European Community: A Auantitatice Assessment," Economic Modelling, Forthcoming World Bank (1992), Kingdom ofMorocco: Issues and Prospects in the Public Sector, Gray Cover Report No. 101 57-MOR, June. World Bank (1 994a), Kingdom ofMorocco: Poverty, Adjustment, and Growth, Gray Cover Report No 11918-MOR, January World Bank (1 994b), Kingdom of Morocco: Preparing for the 21st Century - Strengthening the Private Sector in Morocco, Gray Cover Report No. 11894- MOR, July World Bank (1994c), Kingdom of Morocco- Public Expenditure - Issues and Outlook, Green Cover Report No. 13413-MOR, August. World Bank (1996), Country Economic Memorandum, Report No 14155-MOR - 25 - 8. ANNEX - TECHNICAL SPECIFICATIONS 8.1 THE MOROCCAN TRADE AND ENVIRONMENT MODEL The general equilibrium simulation model used in this paper originates directly from a prototype model18 built for the OECD Development Centre's research programme on sustainable development, environment, resource use, trade and technology and another model build for the Moroccan government's Ministry of trade '9 It is calibrated on the data contained in a Social Accounting Matrix estimated for the year 1995 and described in detail elsewhere 20 The version of the SAM used here includes 10 household categories (5 urban and 5 rural), 48 sectors, 3 labour types, 3 separated trading partners (EC, Non-EC Europe, and ROW) and 13 different polluting emissions The model is dynamic and solved recursively for the years 1995, 1997, 1999, 2001, 2003, and 2005. It includes approximately 100 generic equations, describing agent behaviour, market clearing and other accounting relationships. The following sub-sections briefly illustrate the main characteristics of the model. The detailed model algebraic structure is presented in Beghin, Dessus, Roland-Hoist, and van der Mensbrugghe (1996) 8.1.1 Production The Constant Elasticity of Substitution (CES) production function is a nested structure, taking into account the optimising behaviour in the choice of production factors Assuming constant returns to scale, output results from two composite goods non-energy intermediates and energy plus value added The intermediate aggregate is obtained combining in fixed proportions (Leontief structure) all products The value added and energy components are decomposed into two parts, aggregate labour and capital (including energy). Labour is a composite of 3 occupational groups. The capital-energy bundle is further disaggregated into its basic components By distinguishing between 'hew" and 'bid" vintages, the capital existing at the beginning of each period, or already installed, can be separated from that resulting from contemporary investment (putty/semi- putty production function) Finally, the energy aggregate includes two energy substitutes fuel (mainly oil) and electricity Figure 1 depicts the aggregation of production factors. 18Se Beghin, Dessus, Roland-Host and van der Mensbrugghe (1996) See Bussolo, Roland-Hoist and van der Mensbrugghe (1995) 20 See Bussolo and Roland-Hoist (1994) - 26 - Substitution elasticities reflect adjustment possibilities in factor demand originating from variations in their relative prices Consider particular values21 0 00 between intermediates and value added with old capital plus energy, 0 50 between intermediates and value added aggregate incorporating new capital plus energy; 0 12 between aggregate labour and old capital-energy bundle, 1 00 between aggregate labour and new capital-energy bundle, 0 40 among different types of labour; 0 00 between old capital and energy; 0 80 between new capital and energy, 0.25 among different sources of energy associated with old capital, 2 00 among those associated with new capital. 21 These elasticities are derived from the most recent relevant literature. In fact, they are mostly derived from background studies done for the construction of the OECD GREEN model. See for instance Burmaux, Nicoletti and Oliveira-Martins (1992) - 27 - Figure 1: Nested Production Function Production =(00,05) Intermediate consumption Value added and (excluding energy) Energy consumption = (1 5-3 0) (0 12,1 0) Domestic Imported Labour Capital-Energy Intermediates Intermediates aggregate aggregate a = (0 4) =(00,08) Skilled Unskilled Employer Energy Capital a = (0 25, 2 0) Petroleum Electncity Notes 1 CES substitution elasticities are differentiated by capital vintage, new capital (higher) elasticities are shown to the right of the comma 2 No substitution is possible among intermediates. Domestic products can be substituted with the corresponding foreign ones - 28 - 8.1.2 Income Distribution and Absorption Labour income is allocated to households according to a fixed coefficient distribution matrix derived from the original Moroccan SAM Likewise, capital revenues are distributed among households, corporations and rest of the world Corporations save the after-tax residual of that revenue Private consumption demand is obtained through maximisation of household specific utility function following the Extended Linear Expenditure System (ELES). Household utility is a function of consumption of different goods and saving. Income elasticities are different for each household and product and vary in the range 0 20, for basic products consumed by the household with highest income, to 1.30 for services. Government and investment demands are disaggregated in sectoral demands once their total value is determined according to fixed coefficient functions. 8.1.3 International Trade The model assumes imperfect substitution among goods originating in different geographical areas. Imports demand results from a CES aggregation function of domestic and imported goods. Export supply is symmetrically modelled as a Constant Elasticity of Transformation (CET) function Producers allocate their output to domestic or foreign markets responding to relative prices. The model implements a two-stage procedure for determining both import demand and export supply For imports, consider Figure 2. At the first stage, aggregate demand is decomposed into a domestic component and an aggregate import component. At the second stage, aggregate import demand is allocated across the various trading partners. Export supply is treated in a symmetric fashion (Figure 3) Producers allocate production between domestic sales and aggregate export sales At the second stage, aggregate exports are sold to the various trading partners based on the relative price the exporter can receive in each market.22 22 Elasticities between domestic and foreign products are of comparable magnitude for imports demand and exports supply. Their values are 3.00 for agncultural goods, 2 00 for manufactured goods and I 50 for services Similar values are used for the second nesting - 29 - Figure 2: Import Demand Aggregation Aminglon Demand Domestic Demand Imported Demand Import Import Import Demand Demand Demand from EU from ROE from ROW Figure 3: Export Supply Aggregation Domestic Production Domestic Sales Aggregate Exports Exports Exports Exports to the EU to the ROE to the ROW The small country assumption is applied, Morocco being unable to change world prices Thus, its import and export prices are exogenous. The balance of payments equilibrium is determined by the equality of foreign savings (which are exogenous) to the value for the current account With fixed world prices and capital inflows, all adjustments are mediated by the real exchange rate- increased import demand, due to trade liberalization must be financed by increased exports, and these can expand owing to the improved resource allocation Price decreases among importables drive resources towards export sectors and contribute to falling domestic resource costs (or real exchange rate depreciation) - 30 - 8.1.4 Model Closure and Dynamics The equilibrium condition on the balance of payments is combined with other closure conditions so that the model can be solved for each period Firstly consider government budget Its balance is exogenous and the household income tax schedule shifts in order to achieve the predetermined net government position Secondly, investment must equal savings, which originate from households, corporations, government and rest of the world. The dynamic structure of the model results from that last condition, equilibrium between savings and investment. A change in the savings volume influences capital accumulation in the following period Exogenously determined growth rates (in this case from the macro reform baseline) are assumed for various other factors that affect the growth path of the economy, such as population and labour supply growth rates, labour and capital productivity growth rates and energy efficiency factor growth rate Agents are assumed to be myopic and to base their decisions on static expectations about prices and quantities. The model dynamics are therefore recursive, generating a sequence of static equilibria 8.1.5 Emissions Emissions are determined by either intermediate or final23 consumption of polluting products. In addition, certain industries display an autonomous emission component linked directly to their output levels. This is introduced in order to include some polluting production processes which would not be accounted for by only considering the vectors of their intermediates consumption It is assumed that labour and capital do not pollute. Emissions coefficients associated with each type of consumption and production are derived from a previous study on the determinants of polluting intensity for the US and here adapted to the Moroccan case (Dessus, Roland-Hoist, van der Mensbrugghe 1994).24 A change in sectoral output, or in consumption vectors, both in levels or composition, affects therefore emission loads. Formally, the total value for a given polluting emission takes the form. E = a,C, + 1,8, Xo"'P' + Z a JX ,Aington a 3I 3 23 Final consumption. in this context, is restricted to households, government and investment demand Exports are not considered since the analysis is limited to local emission 24 Instead of focusing on pollution output at individual industrial sources, the authors advocate moving back up the production process Factories producing pollution can be numerous and very dispersed geographically The evidence reported in their study indicates that only a few commodities are responsible for determining pollution levels when they are consumed as intermediates. The authors' econometric estimates indicate that over 90 per cent of the variation in emission of most toxic pollution can be explained by consumption of less than a dozen intermediate commodities. Their calculations are based on a 345 sector US input-output table (see Reinert and Roland-Hoist (1992)) and on the 1987 IPPS (Industrial Pollution Projection System) database developed at the World Bank for the US (Hettige, Martin, Singh, Wheeler (1995)) -31 - where i is the sector index, j the consumed product index, C intermediate consumption, A'""' output, A""ng"o final consumption (at the Armington composite goods level), a the emission load associated to one unit consumption of product j and 83, the emission load associated to one unit production of sector i Thus, the first two elements of the right hand side expression represent production-generated emissions, the third one consumption- generated emissions. There are 13 types of polluting substances Their volume is independently determined and measured in metric tons Toxic emissions in air (TOXAIR), water (TOXWAT) and soil (TOXSOL) depend primarily on consumption of chemicals (especially fertilizers for water pollution), oil derived products and mineral products Bio-accumulative emissions differ from the previous ones for their long term effects on bio organisms, due to their high lead (or other heavy metal) concentration Again, these are distinguished according to the medium where they are released into the air (BIOAIR), water (BIOWAT) and soil (BIOSOL) These emissions are a result of the use of mineral and metal products, generally found in construction-related sectors There are 5 types of toxic substances released in the air sulphur dioxide (S02), nitrogen dioxide (N02) and carbon monoxide (CO), volatile organic compounds (VOC) and suspended particulates (PART) Their levels depend primarily on fuels consumption- oil and coal derived products Finally, two additional categories of water polluting substances are considered: suspended solids (SS) and those measured by their biochemical oxygen demand (BOD) These emissions are related to the consumption of mineral products The household utility functions do not include among their arguments any term directly related to environmental qualities In other words, pollution levels do not explicitly affect household utility Despite the theoretical validity of such relationship, empirical applications would require estimates for utility values that household assign to environmental qualities Unfortunately statistical information on which these estimates can be based is still too limited (Perroni and Wigle 1994) Likewise, environmental degradation does not affect production factors productivity Productivity gains resulting from a greener are not measured in this model Thus, the potential gains from environmental protection policies are almost certainly going to be under-estimated 8.1.6 Policy Instruments The model includes a variety of important instruments of economic policy: direct and indirect taxes on production, consumption and revenues, tariffs and other taxes and subsidies on international transactions Each of these taxes/subsidies is differentiated by sector, product, household, production factor, consumption type or income source. A uniform tax on each unit of polluting emission (for type of toxic substance) is also introduced and paid by the polluter agents. This tax can be endogenously determined if specified levels of emission (abatement) are to be targeted, otherwise it can be exogenously fixed. In this latter case, emissions levels become endogenous - 32 - 8.2 DECOMPOSITION OF POLLUTION EFFECTS Consider the following identity, which simply states that total emission (for each type of pollutant) is equal to the sum of sectoral emissions 11= OE, Xoulp ~xupJ A uiu t ) The total variation in emission levels can then be measured as the sum of the just mentioned three effects by differentiating the shown identity* SE = 1 'O'+6.Xo'"'+pXuut ),~1~ E, t0p' I ut E, [ X o"'" EL X 0u"" X,'upu X,out, ) ro'u XU t"' ' Tot~' x,xfu~ where e is the differential operator, E total emission load, XT1'1 total output (in real terms), E, the sectoral emission loads and X,ou'' the sectoral outputs A similar formula is used in the case of emissions originating from final consumption. In this case, technological effect is null, given that each component of final consumption is associated to an emission coefficient invariant with time The emission loads variation due to a modified consumption vector takes the form: e E [O(X ~~~~I Amngtm E + Xrigo) E Xrmingon + X Annington X X angton where X"'n"'" is total final consumption (of the Armington composite good) in real terms and XA,"l"o final consumption in real terms of product i. - 33 - 8.3 POLLUTION CATEGORIES Table 11: Definitions of Major Pollution Groups TOXAIR' Toxic Air Pollutants TOXWAT Toxic Water Pollutants TOXSOL Toxic Soil Pollutants BIOAIR Bio-accumulative Air Pollutants BIOWAT Bio-accumulative Water Pollutants BIOSOL Bio-accumulative Soil Pollutants S02 Sulphur Dioxide N02 Nitrogen Dioxide CO Carbon Monoxide VOC Volatile Organic Compounds PART Fine Airborne Particulates BOD Biochemical Oxygen Demand TSS Total Suspended Solids 1) See Hettig et al (1995) for more details - 34 - Table 12 Sectoral Adjustments Resulting from the FTA Real Real Total Output Emission Share Output Consumption Emissions Share Share Ratio HARDWHEAT 177 157 265 109 002 002 SOFTWHEAT 118 161 204 116 0.03 002 BARLEY 171 158 259 081 002 002 MAIZE 138 163 228 021 001 002 RICE -11 08 327 -1050 000 0.00 000 LEGUMES 280 170 404 022 000 002 SUGAR -387 236 -248 149 005 003 OILSEEDS 008 170 154 019 000 001 RAWFIBRE 1 29 000 3 16 003 000 001 VEGETABLES 327 1 80 449 1 14 002 001 CITRUS 1790 311 2038 051 001 001 OLIVES 1 66 000 345 030 000 001 GRAPES 109 141 280 0 11 000 001 DATES 157 136 330 020 0.00 001 ALMONDS 1 83 1 42 322 038 000 001 OTHAGRI 3762 252 4097 022 000 001 LIVESTOCK 082 055 351 466 001 000 FORESTRY 048 159 407 028 000 001 FISHING 917 187 1443 100 001 001 PHOSPHMIN 1251 000 1493 262 075 029 PETROLEUM -100 507 -100 3.21 021 007 ELECT 209 108 7 15 128 0.01 001 AGROFOOD 166 357 309 622 0.13 002 ANIMALFEED 2 15 325 307 023 000 002 BEVERAGE -3916 1345 -3911 077 001 001 TEXTILE 1133 697 1340 240 001 000 CARPETS 11095 1830 11806 0 17 000 000 APPAREL 2667 880 2689 2.94 003 001 LEATHER -287 -0 17 -235 1.47 000 0.00 SHOES -362 -030 -307 1 20 0.00 000 WOODPROD 458 503 -411 121 003 003 PAPERPUB -698 732 -643 151 008 005 STONECERAM -602 770 -532 294 11.08 377 METALS -2948 2309 -2906 1 10 0.41 037 METALPROD -1566 1425 -1507 208 062 030 NONELECMCH -11 76 1660 -1134 078 012 0 15 VEHMCLES -666 1759 -620 148 003 002 ELECMACH 058 2432 1 35 1.55 0.15 010 CHEMFERT 8625 1517 8737 3 30 48.55 1469 OTHERCHEM -1352 1356 -1280 273 2373 869 OTHMFG -25 16 1848 -2489 027 005 019 CONSTRUCT 737 733 922 688 1323 192 COMMERCE 701 000 1089 799 0.01 000 TRANSPORT 810 189 1167 331 011 003 COM1fUN1C 607 1 19 1067 086 000 000 BANKINSRE 467 278 960 320 001 000 OTHPRIVSRV 821 5 15 1092 1466 0.26 002 PUBLICADM -003 -003 368 763 019 003 Total 10000 10000 The first three columns represent differences with respect to the baseline scenario m 2005 Shares i the last three columns are for the 1995 base The last column contains ratios, not percentages -35 - Kingdom of Morocco: Morocco Environment Review Working Paper II AIR POLLUTION, HEALTH AND CLEANER FUELS: A Cost-Benefit Analysis for Casablanca/Mohammedia Prepared by Bjorn Larsen EMTEN World Bank' May 1997 I/This paper is based on a longer background paper by B Larsen prepared for the Morocco Environment Review Task manager is Yoko Eguch (MN INE). The analysis and conclusions in this paper do not necessarily represent the views of the World Bank. I SUMMARY A substantial share of particulates (PMIO), sulphur dioxide and lead pollution in Casablanca/ Mohanmedia is from the combustion of fuel oil, diesel, and leaded gasoline. The sulphur content in diesel and fuel oil is about I and 3 percent respectively, and the lead content in gasoline is about 0 5 g/1tr, all high by international comparison The high sulphur content is also contributing significantly to higher PMIO emissions. An important policy issue is therefore if the sulphur and lead content in major petroleum products should be reduced The annual social cost of health impacts associated with PMIO, sulphur dioxide and lead pollution in Casablanca/Mohammedia is conservatively estimated at US $185-190 million. Adjusting for the fact that some of the health impacts are likely to be a result of long-term exposure, the social cost per ton of PMIO and sulphur dioxide is estimated at US $16,000 and US $175 respectively. Based on the quality of fuels used in Morocco, the estimated social cost is equivalent to US $65 per ton of fuel oil and US $105 per ton of diesel. The health benefit of reducing the sulphur content in fuel oil from 3 to 1 percent is estimated at US $40 per ton of fuel oil used in industry without emission control technology. The corresponding estimated benefit is US $20 for a sulphur reduction from I to 0.05 percent in diesel used by road vehicles. These benefits are predominantly because of lower PMIO emissions associated with lower sulphur content in the fuels The health benefits of eliminating lead in gasoline is estimated at US $0 035-0 05 per liter (US $45-65 per ton). The lowest cost option to reduce sulphur in fuel oil and diesel, and provide sufficient supply of unleaded gasoline, is found to be investments in SAMIR refinery. About 75 percent of the investments would improve the economic viability of the refinery as well as fuel quality, while 25 percent would be primarily for improved fuel quality without direct economic benefits. The costs of investments in SAMIR compared favorably to import of low sulphur crude oil or import of low sulphur fuel oil and diesel and unleaded gasoline The analysis in this paper indicates that the health benefits of reducing the sulphur content to I percent in fuel oil and to 0.05 percent in diesel, and supplying unleaded gasoline, outweigh the incremental cost of improving the fuels by a factor of 2.5-3 The SAMIR refinery analysis also indicates that these fuel quality improvements are technically feasible. In order to refine the estimates, systematic air quality monitoring data and a more detailed assessment of the investment costs at SAMIR are required. A share of diesel, fuel oil and gasoline is used outside of urban areas where the health impacts of pollution can be expected to be minimal. An available option is therefore to first concentrate the use of cleaner fuels to the urban areas. 2 AIR POLLUTION, HEAL TH AND CLEANER FUELS A. The Approach There is substantial evidence that certain air pollutants significantly impact human health Numerous studies have been conducted in the United States, Europe, and in a few but increasing number of developing countries. They all conclude that air pollution causes increased mortality, respiratory disorders, and in some instances increased cancer and neurophysiological impairment. Major air pollutants of health concern are PM10 (particulates with diameter less than 10 micrometer), sulphur dioxide, and lead. Casablanca/Mohammedia is likely to be the urban area in Morocco with the most significant environmental health concern. The area has a population of 3 million, is the location of about 50 percent of industrial activity including two power pants and the largest refinery, and is the home of more than 400,000 vehicles A substantial share of PM10, sulphur dioxide and lead pollution in Casablanca/ Mohammedia are from the combustion of fuel oil, diesel, coal and leaded gasoline. The sulphur content in diesel and fuel oil is about 1 and 3 percent respectively, and the lead content in gasoline is about 0 5 g/ltr, all high by international comparison The high sulphur content is also contributing significantly to higher PM10 emissions. An important policy issue is therefore if, and to what extent, the sulphur and lead content should be reduced2 This issue is analyzed in this paper by estimating the health benefits and incremental fuel costs of supplying cleaner fuels in the Casablanca/Mohammedia urban area. B. Health Impacts of Air Polution Bl. Dose Response Functions In the absence of controlled studies linking health to air pollution in Morocco, an approach is to apply dose response functions estimated in the United States and Europe. Ostro (1994) presents a summary of these estimates for PMIO, sulphur dioxide, and lead, and Dockery et al (1993) and Pope et al (1995) provide estimates of long-term impacts of particulates on mortality. Mean estimates are presented in Table 1. The health impacts of lead related to hypertension, heart attacks and premature mortality are presented as a range because evidence is relatively weak. 2/The use of coal in Casablanca/Mohanumedia is mainly restncted to the Mohammedia power plant, which is equipped with particulate control technology The net benefits of improving the efficiency of the control technology, or installing a new high efficiency electrostatic precipitator, are higher than the use of cleaner coal. A cost-benefit analysis of cleaner coal is therefore not presented in this paper 3 Table 1: Annual Health Impacts per lug/m3 Change in Air Concentrations Dose Response Coefficients (representing central estimates) Particulates (PM 10) Premature mortality per 1 million people 16.6 RHA per 100,000 people 1.2 ERV per 100,000 people 23.5 RAD per adult 0 0575 LRI per child 000169 Asthma attacks per asthmatic 0 0326 RS per adult 0 183 Chronic bronchitis per 100,000 6 12 Sulphur dioxide (SO2) Premature mortality per 1 million people 2 4 RS per 1000 children 0.018 Chest discomfort per adult 0 01 Lead (Pb) Reduction in IQ points per child 0 975 Cases of hypertension per 1 million males 0-72,600 Cases of non-fatal heart attacks per 1 0-340 milhon males (40+ yrs of age) Premature deaths per I million males 0-350 (40+ years of age) Source: The coefficient for premature mortality related to PMIO is the average of estimates presented in Dockery et al (1993) and Pope et al (1995), adjusted for age distribution in Morocco. All other coefficients are from Ostro (1994). RHA is respiratory hospital visits; ERV is emergency room visits; RAD is restricted activity days; LRI is lower respiratory illnesses, and RS is respiratory symptoms Dose response coefficient for asthma attacks has been adjusted for differences in the prevalence of asthma between the United States and Morocco. Lead pollution causes impairment of neurophysiological development in children Impact on IQ points is often used as an indicator of such impairment. B2. Estimated Health Impacts and Social Cost Lack of systematic air quality monitoring data makes it difficult to estimate the health impacts of air pollution in Casablanca/Mohammedia. Limited monitoring data, however, suggests that air concentration levels of total particulates, S02, and lead could exceed 150 ug/m3, 100 ug/m3, and 1 0 ug/m3, respectively. A level of total particulates of 150 ug/m3 could indicate a PMIO level of 70-80 ug/m3 For comparison, monitoring data from about 30 cities in India shows an average concentration level of PMIO of about 100 ug/m3. 4 To be relatively conservative, health impacts of 30 ug/m' of PM10, 50 ug/m' of SO2, and 0.5 ug/m3 of lead have been estimated for Casablanca/lohammedia based on the coefficients in Table I and an exposed population of 3 million. Estimated total annual social cost of these impacts are presented in Table 2 The most significant estimated health impacts are premature mortality from respiratory disorders and an increase in the number of individuals with bronchitis associated with PM10 Table 2: Estimated Annual Social Cost of Air Pollution in Casablanca/Mohammedia Number of Persons or Social cost Total social cost Cases per person or case (US $ million) (US $) Particulates (PM1O) Per 30 ug/m3 Premature mortalhty* 1,500 43,000 64.5 RHA 1,100 1,000 1 1 ERV 21,000 25 05 RAD in adults 3,800,000 6 22.8 LRI in children 40,500 100 4.0 Asthma attacks 350,000 8 2 8 RS in adults 12,000,000 1 12.1 Chronic bronchitis 5,500 10,000 55.1 Subtotal 162 9 Sulphur dioxide (SO2) Per 50 u/m' Premature mortality 360 43,000 15.5 RS in children 700 1 - Chest discomfort in adults 1,100,000 1 1 1 Subtotal 16.6 Lead Per 0 5 u/m IQ points lost 26,000 225 5 9 Hypertension 0-40,000 20 0-0.8 Non-fatal heart attacks 0-94 600 - Premature mortality 0-96 43,000 0-4 1 Subtotal 5 9-10.8 TOTAL 185.3-190.3 * PMIO related mortality is based on extrapolation of the average of estimates in Dockery et al (1993) and Pope et at (1995). All other estimated health impacts are based on extrapolation from Ostro (1994). 5 The monetized social cost of premature mortality is based on willingness to pay (WTP) for a reduction in the probability of early death. The mean statistical value of life based on WTP studies is on the order of US $3 million in the United States. Alternative valuation techniques, such as observations of wage premiums in higher risk occupations, tend to confirm a mean statistical value of life of US $3 million WTP reflects to a large extent income and wage levels. Thus the figure has been adjusted by the income differential between Morocco and the United States. An important issue is that most WTP studies have been conducted among individuals in their prime age with an average remaining life of 35 years. It can be expected that those who die prematurely from air pollution are individuals with a poorer than average health status, and therefore lower than average remaining life expectancy. The number of years lost due to air pollution is not captured in available studies. Some of the premature mortalities may only be avoided for a few days in the absence of air pollution, while the estimated increase in cardio-pulmonary mortality associated with PM1O in Pope et al (1995) and Dockery et al (1993) may reduce the life time by 20 years or more. An average loss of 10 years has been applied here. Thus the statistical value of life based on WTP has therefore been adjusted downwards by a factor of 3 5. The social cost of morbidity is also based on estimates in the United States. For respiratory hospital admission and emergency room visits, the estimates are based on treatment cost and lost earnings. The cost of the other morbidity impacts is based on WTP to avoid such impacts, and are most often higher than the cost of medical treatment and lost earnings, reflecting the value individuals place on avoiding pain and suffering. The cost of IQ losses from lead pollution is based on studies of the correlation between IQ and life time earnings, controlling for other factors explaining income level All cost figures are adjusted by the income differential between Morocco and the United States B3. Issues For Consideration Several important issues need to be discussed regarding the estimates in Table 2: First, if actual concentration levels of PMIO, sulphur dioxide, and lead in Casablanca/Mohammedia are "only" 30 ug/m', 50 ug/m3, and 0 5 ug/m3 respectively, a question arises as to what extent health impacts occur at these relatively low levels of pollution. The two recent studies by Pope et. al (1995) and Dockery et. al (1993) do not detect a lower threshold below which pollution has insignificant health impacts, and other studies have concluded the same. Similarly, studies of lead pollution indicate health impacts at very low levels. Second the estimated health impacts of PM1O and sulphur dioxide are likely to represent some, and perhaps significant double counting. This is because PMIO and sulphur dioxide are often highly correlated, which makes it difficult to separate the impacts A correction for double counting would only change the total health impacts by a maximum of 15 percent 6 Third, an important issue is to what extent extrapolation of estimated dose response coefficients from the US or Europe to Casablanca/ Mohammedia is realistic. The Pope et al (1995) and Dockery et al (1993) studies seem to indicate that the increased mortality associated with PM 10 is primarily because of an observed increase in cardio-pulmonary mortality and to a lesser extent an increase in lung cancer If the share of these causes of mortality is lower in Casablanca/Mohammedia - than in the United States (for the age groups above 35 years of age), it may be that extrapolation would tend to overstate the actual mortality impact of PM10. This is difficult to assess and can only be confirmed if studies are conducted in Casablanca/Mohammedia B4. Annualizing Estimated Health Impacts In order to undertake a benefit-cost analysis of options to reduce PM10, sulphur dioxide, and lead pollution, it would be necessary to estimate the health impacts of one year of a certain amount of pollution and compare the monetary value of those impacts to the cost each year of reducing that amount of pollution Although the estimated health impacts presented in Table 2 are annual impacts, some the health impacts are most likely a result of long-term exposure to air pollution This is particularly the case for premature mortality and chronic bronchitis If the increased mortality and chronic bronchitis associated with PMIO is a result of 20 years of exposure to PM10, then the social cost per year of pollution would average about US $100 million. This estimate is used as the basis for comparing benefits and costs of reducing pollution through cleaner fuel oil, diesel and gasoline C. POLLUTION DAMAGE COST OF DIRTY FUELS In order to translate estimated social cost of health impacts of air pollution into damage cost per ton of pollutant and damage cost per ton of fuel oil, diesel and leaded gasoline, it is necessary to have an estimate of total tons of each pollutant emitted annually Annual emission estimates for Casablanca/Mohammedia are presented in Table 3 The emission estimates are based on estimated fuel consumption in each sector in Casablanca/Mohammedia, and a set of emission coefficients for each sector and fuel The emission coefficients for coal used in the Mohammedia power plant are based on coal with 10 percent ash and 1 25 percent sulphur content, and 95 percent abatement efficiency of the particulate control technology. 7 Table 3: Estimated Total Emissions from Fossil Fuels in Casablanca/Mohammedia (000 tons in 1994) Year 1994 SO2 Lead Particulates (PM10) Road transport Gasoline vehicles 0.2 0.08 Diesel vehicles 4 8 1.8 (1.6) Subtotal 5 0.08 1 8 (1.6) Industry: Coal* n n Fuel oil** 29 2.2 (1.72) Diesel 2 0.2 (0.18) Subtotal 31 2.4 (1.9) Power plants. Coal 15 1.5 (1.1) Fuel oil 44 0.9 (0.7) Subtotal 59 24 (1.8) TOTAL 95 0.08 6.6 (5.3) *Coal is mainly used by the cement plants which are not located in Casablanca/Mohammedia. **Includes refinery fuel n indicates negligible. C1. Damage Cost Per Ton of Pollutant In order to estimate the damage cost of PMIO and sulphur dioxide, a relationship between air concentration levels and annual emissions must be established This is a highly complex relationship and depends on such factors as wind, precipitation, temperature, and size of particulates To arrive at an average, a linear relationship has been applied, which gives an estimated damage cost of about US $16,000 per ton of PM10, and US $175 for a ton of sulphur dioxide C2. Diesel Fuel The health damage cost of road transport diesel, the sector in which most diesel is used in Morocco, is presented in Table 4 on the basis of the damage cost per ton of pollutant and average emission coefficients per ton of diesel Cost estimates are presented for three qualities of diesel, i.e., for diesel containing 1 percent, 0 75 percent and 0.05 percent sulphur The emission coefficient for sulphur dioxide is straight forward as it is proportional to the sulphur content The difference in 8 particulate emissions by diesel quality is more complex, and is here based on estimates that about 2 percent of sulphur forms into sulphate and combines with water and hydrocarbons, resulting in about 0 1 kg of particulates per ton of diesel for each 0.1 percentage point sulphur content. This estimate is likely to be conservative because sulphur dioxide reacts in the atmosphere to form ammonium sulphate or sulfuric acid which contribute to particulate concentrations in the air The diesel used in Morocco has a sulphur content of about 1 percent, and its damage cost is estimated at about US $105 per ton, or US c 9.1 per liter. The incremental benefit of changing from a diesel with 1 percent sulphur content to 0.75 percent sulphur is estimated at US $5, or about US c 0.4 per liter, while the incremental benefit of changing from 0 75 percent sulphur diesel to 0 05 percent is estimated at about US $15, or about US c 13 per liter It should be noted, however, that the damage cost of transport diesel (and incremental benefits of lower sulphur diesel) could be significantly higher because of the immediate human exposure to vehicle emissions in urban areas. Table 4: Estimated Polution Damage Cost of Road Transport Diesel in Casablanca/Mohammedia Damage Cost Damage Cost (US $/ton of fuel) (US c/ltr of fuel) 1% Sulphur diesel 105 9.1 0 75% Sulphur diesel 100 8.7 0 05% Sulphur dresel 85 74 * More than 90 percent of particulates from transport diesel are reported to be PMIO C3. Fuel Oil The damage cost of fuel oil is presented in Table 5 on the basis of the damage cost per ton of pollutant and average emission coefficients per ton of fuel oil Cost estimates are presented for two qualities of fuel oil, i.e., for fuel oil containing 3 percent and 1 percent sulphur. As for diesel, the emission coefficient for sulphur dioxide is straight forward. Particulate emissions by fuel oil quality are more complex, and is here based on the relationship between particulate emissions and sulphur content reported by US EPA for heavy fuel oil (kg per 1000 liters, i.e., 1.25 S + 0.38 where S is the sulphur content. The fuel oil used in Morocco has a sulphur content of about 3 percent, and its damage cost is estimated at about US $65 per ton. The incremental benefit of changing to a fuel oil with 1 percent sulphur is estimated at US $40 per ton. 9 Table 5: Estimated Polution Damage Cost of Fuel Oil in Casablanca/Mohammedia Damage Cost (US $/ton of fuel) 3% Sulphur fuel oil 65 1% Sulphur fuel oil 25 * PM10 from combustion of heavy fuel oil is reported to average about 80 percent of particulates. C4. Leaded Gasoline Leaded gasoline is the predominant source of lead pollution in Casablanca/Mohammedia in the absence of point sources such as large lead smelters. If the lead content of gasoline averages 0.5 g/ltr, then the social cost of lead in gasoline is estimated at US $0 035-0.05 per liter of gasoline The incremental cost of providing unleaded gasoline is very unlikely to exceed US $0.015 per liter The cost of lowering the lead content is likely to be even less, which consequently would be justified at even lower levels of lead concentrations. D. COST-BENEFIT ANALYSIS OF CLEANER FOSSIL FUELS There are basically three options available to reduce the sulphur content in the diesel and fuel oil consumed in Morocco: (a) import of low sulphur crude, (b) import of low sulphur diesel and fuel oil, and export domestically produced products; and (c) investments in the refineries (particularly SAMIR) The incremental cost per unit of fuel is estimated for each of the three options Dl. Incremental Cost ofLower Sulphur Crude Oil Crude oil prices, at any point in time, depend to a large extent on supply and demand conditions for various characteristics of crudes. Two approaches have been used here to assess the incremental cost of lower sulphur crude oil. The first approach involves calculating the price differentials of crudes with different sulphur content, but with crude fractions that satisfy the product mix currently demanded in Morocco, i.e., about 30 percent diesel and 40 percent fuel oil The second approach involves estimating a regression equation for a much larger number of crudes to estimate the incremental price of lower sulphur crudes in general, not taking into account the crude fractions. The first approach, based on eight crude oils with 30 percent diesel and 40 percent fuel oil fractions, indicates that a crude with one percentage point lower sulphur costs on average about US $1.1-1 3 more per barrel during the last ten years. Based on the second approach, the following regression equation was estimated: 10 P = a + b API + cS where P is crude oil price per barrel, API is the gravity, and S is sulphur content in the crude The equation was estimated for data on 40 internationally traded crudes Estimates were performed separately for price data from year-end 1991, 1992, 1993, 1994 and 1995 On average, crudes with one percentage point lower sulphur content (but similar API grade) cost about US $1 1 more per barrel. Thus, the two approaches give relatively similar results. To illustrate the magnitude of incremental cost per ton of diesel and fuel oil, a comparison is made between currently imported crudes and low sulphur crudes The total incremental cost of importing about 6.2 million tons of low sulphur crudes would be about US $72 million per year If this incremental cost is allocated to diesel and fuel oil, the incremental cost per ton would be about US $15.2 per ton D2. Incremental Cost of Impordng Cleaner Petroleum Products Several qualities of heavy fuel oil and diesel in terms of sulphur content are traded internationally and would be available to Morocco. Spot price differential or incremental costs of heavy fuel oils in three markets, and diesel in two markets were analyzed. The incremental cost of a fuel oil with I percent sulphur versus a fuel oil with 3 percent sulphur in the US and Europe averaged about US $14 per ton during 1987-96, while the average price differential during 1996 was on the order of US $10-12 per ton. The price differential between 0 3 percent and 3.5 percent sulphur content fuel oil in the Singapore market was about US $13 per ton during 1987-96, but only US $6 during 1996. The incremental cost of a diesel with 0.05 percent sulphur versus a diesel with I percent sulphur averaged about US $8-9 per ton in 1996 Based on the incremental cost per ton of fuel, the total incremental cost of meeting domestic demand in Morocco through imports of diesel with 0.05 percent sulphur and fuel oil with 1 percent sulphur would be on the order of US $50 million per year, or an average of about US $10 per ton of diesel and fuel oil. These figures do not include incremental transport costs. D3. Incremental Cost of Refinery Upgrading An alternative option to import of low sulphur crude or low sulphur diesel and fuel oil would be upgrading investments in SAMIR refinery4 SAMIR petroleum product output is primarily for the 3/From Saudi Arabia, UAE, Qatar, Iran, Oman, Dubai, Egypt, Tunisia, Algeria, Libya, Nigeria, United States, Norway, and the UK 4./This section is based on an analysis of cost and technological options prepared by T S Nayar in the Industry and Energy Department, Oil and Gas Division, at the World Bank. 11 domestic market. The refinery's configuration yields a large share of low value heavy fuel oil, which limits its economic viability To provide a range of upgrading options, an initial evaluation was undertaken of five alternatives to improve the economic viability of SAMIR: investments in (1) hydrocracker; (2) fluid catalytic cracker (FCC); (3) mild hydrocracking (MHC), (4) mild hydrocracking with 50 percent high sulphur crude and 50 percent low sulphur crude; and (5) vacuum gas oil (VGO) desulphurization The initial evaluation indicates that alternative (3) with investments 'in MHC would best satisfy a combination of economic return on investment, anticipated future domestic product demand, and modesty in terms of total investment cost. The analysis of alternative (3) was conducted in two stages: First, investments were evaluated to improve the economic viability of the refinery, without including investments for which the primary objective is to provide cleaner fuels Second, investments were evaluated for which motivation is predominantly environmental improvements, i.e., to provide cleaner fuels Economic Viability Investments The investments under alternative (3) to improve the economic viability of SAMIR include a mild hydrocracker, increased vacuum distillation capacity, visbreaker, hydrogen plant, sulphur plant, and off-site facilities Total investment cost is roughly estimated at US $285 million ' These investments would increase the supply of higher value diesel (by about 0 7 million tons) for which there is a domestic deficit and reduce the supply of low value fuel oil for which there is a domestic surplus. The value added of the incremental diesel output would provide a rate of return on investment estimated at more than 15 percent. The economic viability investments would also contribute to improved fuel quality. With the crude oils currently used, the sulphur content in diesel would on average be reduced from about I percent to about 0.75 percent In fuel oil, the sulphur content would be reduced from about 3 percent to about 1 percent if the refinery uses the heavy bottom residue oil in its own furnaces. Environmental Investments. Investments in SAMIR to improve fuel quality predominantly for environmental improvements could include a diesel hydrodesulphurization (IDS) unit, an isomerization unit, and any off-site facilities. Estimated cost would be roughly US $90 million. The diesel HDS would further reduce the sulphur content in diesel from about 0.75 to no more than 0.05 percent, at a cost of about US $5 per ton of diesel. The isomerization unit, necessary for only a portion of the gasoline output, would allow the refinery to supply high octane unleaded gasoline, meeting all domestic gasoline demand at an incremental cost of less than US $0.015 per liter. The incremental cost of the three alternatives are presented in Table 6, indicating that the SAMIR refinery investments are likely to be the least cost option No cost is presented for the investments that would improve the economic viability of SAMIR refinery since the return on 5/Investment costs include 25 percent off-site facilities and 20 percent pnce contingencies 6/Investment costs include 25 percent off-site facilities and 20 percent price contingencies. 12 investment is positive However, the use of the cleaner fuels resulting from those investments has an opportunity cost reflected by international spot price differentials. This opportunity cost is applied in the comparison of health benefits and incremental fuel costs of cleaner fuels With the plans to privatize the refinery, the most practical solution would be to specify fuel quality standards and then allow the refinery and the petroleum sector to choose the least-cost solution, based on other potential refinery modifications that private owners may deem desirable. Table 6: Incremental Cost of Options to Reduce Sulphur in Diesel and Fuel Oil Options: Incremental cost (US $/ton) SAMIR refinery investments: Economic viability Diesel (0.75% S) and fuel oil (1.0% S) Environmental Diesel (0.05% S) $5/ton Low sulphur crude import Diesel and fuel oil $15/ton Low sulphur diesel/fuel oil Diesel and fuel oil (weighted average) $1 0/ton* import *not including incremental transport cost to Morocco 4. Costs and Benefits of Lower Sulphur Fuels Compared The preceding analysis indicates that the option of SAMIR refinery investments to reduce the sulphur content in diesel and fuel oil represents the lowest cost of the alternatives considered. The investments would bring the sulphur content in diesel to less than 0.05 percent and sulphur in fuel oil to about 1 0 percent. The benefits in terms of health improvements of reducing sulphur in these fuels (and thus also reduce particulate emissions) to those levels were estimated at about US $40 per ton of fuel oil used in industry and power plants without emission control technology and about US $20 per ton of diesel used in road transport These benefits are more than 3 times higher than the international spot price differentials of 3 percent and 1 percent sulphur fuel oil, and more than 2.5 times higher than the differentials of 1 percent and 0.05 percent sulphur diesel Some caution, however, must be taken in interpreting the estimated benefits and costs First, systematic monitoring data are needed before a more accurate estimate can be provided. Second, the benefit and cost analysis of cleaner fuel oil, diesel and lead in gasoline is for Casablanca/Mohammedia 13 only The health impacts of dirty fuels used outside of urban areas can be expected to be minimal An available option is therefore to concentrate the use of cleaner fuels to major urban areas. E. Conclusions The analysis in this paper indicates that the health benefits of reducing the sulphur content in fuel oil from 3 percent to 1.0 percent, and in diesel from 1 percent to 0 05 percent substantially exceed the cost of sulphur reductions for fuel oil used in industry and power plants without emission control technology and for diesel used in road transport. Similarly, the health benefits of eliminating lead in gasoline are estimated at 2.5-3.5 times the incremental cost of unleaded gasoline In the short-term, however, a practical strategy could be to reduce the lead level in gasoline from currently 0 5 g/1tr to no more than 0. 15 g/ltr, and reductions to levels below 0 15 g/ltr should be considered The analysis was undertaken for Casablanca/Mohammedia, stipulating air concentrations of PMIO at 30 ug/m, of sulphur dioxide at 50 ug/m3, and of lead at 0.5 ug/m3. If, however, air concentrations in reality are higher, the estimated benefits of pollution reduction would be higher as well. 14 References: Direction de la Statistique (1995) "Annuaire Statistique du Maroc " Morocco Dockery et al (1993) "An Association Between Air Pollution and Mortality in Six U.S Cities " The New England Journal of Medicine, vol 329 (24). 1754-59 European Commission (1991) "Forecast of Emissions from Road Traffic in the European Communities " Report EUR 13854 EN. Ostro (1994) "Estimating the Health Effects of Air Pollutants " Policy Research Working Paper No. 1301. World Bank. Pope et al (1995) "Particulate Air Pollution as a Predictor of Mortality in a Prospective Study of U.S. Adults." American Journal of Respiratory and Critical Care Medicine, vol 151(3). 669-74 U S. EPA (1985) " Compilation of Air Pollutant Emission Factors." Volume 1. United States 15 CATALOGUERS/ILE CONFIDENTLAL Report No.: 16570 MOR Type: SR