IPOl.l(Y RFSEARCH WORKING PAPER 1448 Environmental Regulation - E prottnhas :' a not been resricted to wealthy and Development nations. Starting at the iowest Ievel of deveopment' _ . . . ~~~~~~~~~~~~regulatioriinrawses steadigly-. A Cross-Country Empirical Analysis with. income. per capita. The. characteristic progression, is: Susinita Dasgupta firom natural resource Ashoka Mody protection, through Subhendu Roy regulation of vater pollution,. D)avid Wbeeler to air pollution control. I The World Bank Polic Research Departmecnt M i Environmeint, Infrastructure, and Agriculture Division April11995 [POL I(Y RESI ARCII WORKING PAPER 1448 Summary findings l)asgupta, Mody, Roy, and Wheeler develop natural resource protectin. With increased urhaniaiii'ue comparative indices of environmental policy and and industrialization, countrics move from initial performance for 31 countries using a quantified analysis regulation of water pollution to air pollution contrnl of reports prepared for the Ulnited Nations Conference The authors highligilt the importance of institutional on Environment anid Developmentn developmcnt. Environmental regulationi is moBre In cross-country regressions, they find a very strong, advanced in developing countries with relatively secuirc continuous association betwcen their indicators and property rights, effective legal and judicial systems, and national income per capita, particularly whcn adjusted efficient ptublic administration. for purchasing power parity. Their results suggest a charactcristic progression in development. Poor agrarian economies .ocus first on This paper - a product of the Environment, Infrastructure, and Agriculture Division, Policy Research Depat tment -- is part of a larger effort in the department to study the relationship between environmental regulation and economic development. Copies of the paper are available frec from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Elizabeth Schaper, room NIO-037. extension 33457 (27 pages). April 1995. The Policy Research Working Paper Series disseminates the findings of uwk in pogress to encourage the exchange of idLas abIout development issues. An objectiuv of the series is to get the findings out quickly. even if the presentations are less than fully polished. The papers carry the names of the authors and should be used and cited accordingly. The findings, interpretations. and conclusions are the authors own and should not be attributed to the World Bank. its F.recutive Board of Directors. or any of its mernber countries. Produced by the Policy Research Dissemination Center ENVIRONMEMTAL REGULATION AND DEVELOPMENT: A CROSS-COUNTRY EMPIRICAL ANALYSIS by Susmita Dasgupta* Ashoka Mody Subhendu Roy David Wheeler S. Dasgupta and S. Roy are Consultants and D. Wheeler is Principal Economist in the Environment, Infrastructure and Agriculture Division of the World Bank's Policy Research Department. A. Mody is Principal Economist in the Private Sector Development and Privatization Division of the World Bank's Cofinancing and Financial Advisory Services Department. EXECUTIVE SUMMARY Since the Stockholm Conference on Environment and Development in 1972, many countries have taken steps to mitigate environmental damage. More systematic comparative analysis of countries' environmental performance would undoubtedly help clarify the major policy issues and options. Unfortunately, comparable data on regulatory measures are available only for developed countries, and even these data are frequently scanty. In this paper, we undertake a comparative assessment using environmental reports presented to tlLe United Nations Conference on Environment and Development (UNCED, 1992) by 145 countries. From the information in these reports, we have developed a set of indicators which measure the status of environmental policy and performance. This paper describes our methodology, the indices, aiid some results from a statistical analysis of their relationship to other more conventional measures of socioeconomic development. The UNCED reports are similar in form as well as coverage, and permit cross-country comparisons. To an impressive degree, they seem to reflect real environmental conditions and issues. For this exercise, we have randomly selected 31 UNCED reports from the total of 145 (see Table 2A, p. 6). These 31 countries range from highly industrialized to extremely poor, they are drawn from every world region, and they range in size and diversity from China to Jamaica. Our analysis focuses on three dimensions of environmental policy and performance: Overall, "Green" sector, and "Brown" sector. We develop and test a set of hypotheses about regulatory development which can be summarized as follows: * Overall environmental performance should be positively correlated with: 1) Income per capita; 2) Degree of popular representation; 3) Freedom of information; 4) Security of property rights; 5) Development of the legal and regulatory system. Controlling for these variables, * "Green" sector indices should be positively correlated with: 1) Rural population density; 2) Agricultural and forest production share of national output. * E"Brown" sectors indices should be positively correlated with: 1) Particular focus on public health, indexed by life expectancy; 2) Urban share of total population; 3) Urban populaFion density; 4) Manufacturing share of national output. Our analysis of overall regulatory performance reveals strong cross-country associations with income per capita, security of property rights, and general development of the legal and regulatory system. Surprisingly, however, we find only insignificant or perverse associations with degree of popular representation and freedom of information. For both the Green and Brown indices, performance is again strongly associated with income per capita, freedom of property and (in small samples) measures of regulatory efficiency. The two specifically rural-sector variables (population density; proportion of GDP in agriculture and forestry) are only weakly associated with the Green index. T'e fit is much better for the Brown index: degree of urbanization, population density and manufacturing share in GDP all have the expected signs and relatively high significance. Life expectancy as a proxy for public health priority has no independent effect. In summary, our findings suggest that a detailed, quantified analysis of the UNCED reports can yield comparable and plausible indices of environmental policy performance across countries. Cross-country variations in our environmental index are well- explained by variations in income per capita, degree of urbanization and industrialization, security of property rights, and general administrative efficiency. 1. Introduction Since the Stockholm Conference on Environment and Development in 1972, many countries have taken steps to mitigate environmental damage. General environmental legislation is already common, although detailed rules and regulations are still far from universal. In many developing countries, it is clear that enforcement of environmental laws has been hampered by inadequate staffing and funding. Anecdotes abound, but more systematic comparative analysis of countries' environmental performance would undoubtedly help clarify the major policy issues and options. Unfortunately, comparable data on regulatory measures are available only for developed countries, and even these data are frequently scanty. At present, therefore, comparat:.ve analysis must begin with basic data construction. One promising source is the set of environmental reports presented to the United Nations Conference on Environment and Development (UNCED, 1992) by 145 countries. The reports are reasonably comparable because the UN imposed a standard reporting format. Using a multidimensional survey of 31 national UNCED reports, we have developed a set of comparative indices for the status of environmental policy and performance. This paper describes our methodology, the indices, and some results from a statistical analysis of their relationship to other more 1 conventional measures of socioeconomic development. In the following section, we begin with a description of the UNCED reports. Section 3 explains our indexing method, while Section 4 sets out some preliminary hypotheses about the relationships linking environmental policy and performance to socioeconomic development. Section 5 reports and discusses some statistical tests of the hypotheses; and Section 6 concludes the paper. 2. The UNCED Reports As part of the preparations for the United Nations Conference on Environment and Development (UNCED - Rio de Janeiro, June 1992), all UN member governments were asked to prepare national environmental reports. Detailed preparation guidelines were laid down at the First Preparatory Committee meeting in Nairobi in August, 1990.' The UNCED secretariat suggested that the reports be prepared by working groups representing government, business and non-governmental organizations (NGO's). The guidelines recommended that the reports provide information on: (i) the drafting process; (ii) problem areas; (iii) past and present capacity building initiatives; (iv) recommendations and priorities for environment and development; {v) financial arrangements and funding requirements; (vi) environmentally sound technologies; ! United Nations General Assembly document A/CONF.151/PC/8 and A/CONF.lSl/PC/B/Add.1 2 (vii) international cooperation; and (viii) expectations about UNCED. The resulting reports are similar in form as well as coverage, and permit cross-country comparisons. Undoubtedly, the participation of NGO's has helped assure that the UNCED reports are not mere government handouts. To a striking degree, they seem to reflect real environmental conditions and issues. While we recognize that self-reporting always carries the risk of misrepresentation, we should also note that almost all currently available enivironmental information is self-reported by firms and governments. The UNCED reports differ principally in the absence 3f any formal sanction for misreporting. 3. Quantifying Environmental Performance For this exercise, we have randomly selected 31 UNCED reports from the total of 145 (see Table 2A, p. 6). These 31 countries range from highly industrialized to extremely poor, they are drawn from every world region, and they range in size and diversity from China to Jamaica. Our survey considers the state of policy and performance in four environmental dimensions: Air, Water, Land and Living Resources. We analyze the apparent state of policy as it affects the interactions between these four environmental dimensions and five activity categories: Agriculture, Industry, Energy, Transport and the Urban Sector. Although many overlaps 3 undoubtedly exist, we attempt to draw a separate assessment for the interaction of each activity category with each environmental dimension. Our survey assessment uses twenty five questions to categorize the state of (i) environmental awareness; (ii) scope of policies adopted; (iii) scope of legislation enacted; (iv) control mechanisms in place; and (v) the degree of succeus in implementation.2 The status in each category is graded "High, Medium, Low," with assigned values of 2, 1 and 0 respectively. For each UNCED country report, all twenty-five questions are answered for each element of the matrix in Table 1. With 20 elements in the matrix, 500 assessment scores are developed for each country. We compute four composite indices by adding scores within each environmental dimension. We also calculate a total score to provide a composite index of the state of environmental policy and performance. Finally, we have used our scoring system to establish separate indices for three particularly interesting policy dimensions: the extent of environmental awareness; enactment of policies; and success in implementation. We use all three sets of indices for the cross-country analysis reported in Section 5. 2 The survey instrument is included in the Appendix. All country scores are available on request. 4 Table 1 Evaluation Format Sector/ Air Water Land Living Activity Resources Agriculture Industry Energy ____ ____ Transport _ Urban _ - _ Using the four dimensional indices and a composite index, we summarize our results as country rankings in Table 2A. Actual values are displayed in Table 2B. Table 2A also ranks countries on the basis of per capita C-NP (PCGNP) and per capita GDP estimates compiled by the UN International Comparisons Program (ICPGDP) . The ICPGDP computation explicitly adjusts the standard income data to take account of purchasing power parity. Where countries in our sample are not covered in the most recent International Comparisons Program Study (Phase V, 1985), we have adopted a World Bank estimate. The 1985 figures have been extrapolated to 1990 using World Bank estimates of real per capita GDP growth. Table 3 presents summary statistics for the four dimensional performance indices, whose possible maximum values are all 250. The results suggest fairly similar distributions with the exception of Air, which has a significantly lower mean and greater variance. Our statistical results suggest that air pollution gets relatively low priority in poor countries but 5 Table 2A Sample Counnry Runking.: Income and Envrcnnmental Perfonuance Indices Country PCGNP ICPGDP Air Water Land Living Resources Switzerland I I . 2 2 I 2 FInland 2 3 4 3 3 4 4 Germany 3 2 1 2 . Netherlands 4 4 3 4 4 3 3 Ireland 5 5 5 5 4 5 5 ICora 6 7 7 B 7 7 Trinidad 7 6 10 II 1I 12 11 Brazil 8 IU 12 16 16f 15 SAfrica 9 9 8 9 9 10 9 Bulgana 10 7 6 6 6 6 6 Janmica If 16 BI 8 7 X S Tunisia 12 13 9 10 10 11 10 Thailand 13 11 15 24 Is 23 19 Jordan 14 12 17 14 15 22 16 ParAguay 15 14 24 20 20 17 21 Papua NG 16 21 28 27 29 30 29 Philippines 17 17 is 24 20 18 20 Egypt lS 15 21 12 24 27 22 Zambia 19 26 '2 23 20 20 23 Ghana 20 20 18 19 Is 18 17 Pakistan 21 19 13 14 13 13 13 China 22 18 15 16 12 9 12 Kenya 23 24 23 16 16 16 18 India 24 23 13 13 14 i4 14 Nigena 25 22 26 21 25 24 24 Bangladesh 26 25 25 29 27 29 26 Malawi 27 27 Is 22 23 21 27 Bhuman 28 30 30 31 30 28 30 Ethiopia 29 31 31 30 31 31 31 Tanzania 30 29 29 28 28 26 28 Mozambique 31 28 27 26 26 25 25 6 Tibil. 211 Sample Coiuniy DR,Al lnmonne and Env innmctital performance Indices C'onltry I'tGNI' ICPODP Air Waler Land I.iving Env .3I9W01 i IS I 'rXM_ Resources Swucrlane3d J2.,hH0 21 .o90 231 240 J3 238 947 linland 2h,040 15,620 214 229 231 220 894 mnany 22.320 16,920 236 242 241 _ 232 951 Ncthcrlinds 17,320 14.600 219 220_j 229 226 90 lIcand 9.550 9.130 203 22 3 229 216 187 Korea 5.410) 7.190 -SO 170 189 177 686 Trinidad 3.610 8.510 I1l 149 159 13R 564 Brazil 2.680 4.780 113 127 130 123 15 S.Afnca 2.530 5.500 136 165 173 145 619 Bulgana 2.250 7,900 168 198 199 185 750 Jamaica 1.500 3.030 114 168 193 158 633 Tumnsia 1.440 3.979 128 158 161 142 589 Thailand 1.42D 4.610 98 113 129 109 449 Jordar 1.240 4,530 95 131 138 I10 474 Paraguay I.110 3.120 84 117 123 119 443 Papua NG 860 1.500 54 91 100 84 329 Philippines 730 2.320 93 113 123 118 447 Egypt 600 3.100 92 134 118 97 441 Zambia 420 810 87 115 123 114 439 Ghana 390 1.720 93 124 129 118 464 Pakistan 380 1.770 [Os 131 144 128 SOB China 370 1.950 98 127 151 153 529 Kenya 370 1,120 85 127 130J 121 463 India 350 1.150 105 132 143 127 507 Nigeria 290 1.420 75 106 114 105 400 Bangladesh 210 1,050 77 89 109 91 366 Malawi 20W 670 93 116 122 III 352 Bhutan 190 510 39 54 70 93 256 Ethopia 120 310 20 56 67 75 218 Tanzania 110 540 50 90 103 98 341 Mozambique 80 620 56 98 112 102 37_ 7 increases more rapidly in importance with income. By contrast, low income countries such as Tanzania, Mozambique, Bhutan and Bangladesh seem to focus first on the natural resources which are critical to their livelihood -- soils, forests and water. Table 3 Indices of Environmental Policy-Summary Measures for 31 Countries Resource Mean s.d. Maximum Minimum Air 113.84 56.61 236.0 20.0 Water 140.61 50.91 242.0 54.0 Land 149.03 48.26 241.0 67.0 Living 137.B4 46.70 238.0 75.0 4. The Political Economy of Exvironmental Management: Some Preliminary Hypotheses Environmental degradation affects national welfare by damaging human health, economic activities and ecosystems. Because environmental problems represent a classic externality, some government regulation is generally warranted. From an economist's perspective, desirable regulation should weigh two factors: the benefits associated with reduced environmental damage and the opportunity cost of mitigation. In reality, the extent and focus of government intervention will also reflect national political and institutional considerations. 8 4.1 Benefits The demand for environmental quality should increase with income per capita, and we would expect this to be strongly reflected in the country scores. In addition, demographic and sectoral differences may play an important role. For example, economies with high rural population densities and heavy dependence on agriculture and forest extraction should be particularly concerned with agricultural water supply, soil erosion, and deforestation. In our Evaluation Format (Table 1), the relevant scoring cells are located at the intersection of Agriculture with Water, Land and Living Resources.3 If environmental policy reflects basic economic considerations in resource-dependent economies, we would expect country scores in these dimensions to be positively correlated (ceteris paribus) with rural population density and the share of agricultural and forest production in national output. By contrast, urbanized and industrialized economies should exhibit more concern with the potential health impacts of air and water pollution on densely populated areas. The relevant cells in this context are located at the intersections of the Air and Water columns with Industry, Energy, Transport and Urban. We would expect country scores in these dimensions to be correlated with the urban share of national population, urban population density, and the share of manufacturing in national output. 3 Agriculture includes wood production from plantations and primary forests. 9 4.2 Opportunity Costs Governments must make resource allocation decisions with constrained budgets, so we would expect the benefits of environmental improvement to be weighed against opportunity costs. In particular, environmental management lias to share a limited social welfare budget with public health, education and other needs. Therefore the poorer the country, the more limited environmental management resources are likely to be. This should be another source of positive correlation between income per capita and country scores. 4.3 Political Economy Political and institutional factors may also contribute significantly to cross-country variation in environmental policy and performance. Attention to environmental problems should reflect the political power of affected interest groups, the quality of their information about environmental damage, and the effectiveness of legal and regulatory institutions. Many environmental problems pit broad public interests against the profitable pursuit of manufacturing and extraction. Thus, we might expect our environmental performance indices to be correlated with measures of degree of popular representation, freedom of information and education. Performance should also be superior where legal and regulatory systems are relatively efficient. Finally, environmental objectives may be promoted 10 more strongly in economies where secure property rights lead to longer planning horizons. 4.4 Predicted Relationships Within this simple framework, we can make some predictions about the probable strength and direction of empirical relationships across our sample countries. We consider cross- country variations in three sets of indices: (1) Overall policy and performance, along with separate scores for Air, Water, Land and Living Resources; (2) a "Green" index (interaction of Agriculture with Water, Land and Living Resources) and (3) a "Brown" index (interaction of Industry, Energy, Transport and Urban with Air and Water). We have also decompnosed the Green and Brown indices into three subindices: Awareness of environmental problems; enactment of regulations; and success in implementation. However, as Table 4 indicates, the subindices are so highly correlated with the composite indices that more detailed analysis seems unnecessary. Table 4 Correlation Matrix: Component Scores Green Subindices _ Composite Awareness Enactment iSuccess Composite 1 l Awareness .906 1 l Enactment .982 .858 1 1Success .968 .866 .910 1 11 Brown Subindices IComposite Awareness Enactment Suce7ess Composite 1 _ _ Awareness .953 1 Enactment .989 .926 1 Success .984 .934 .951 1 To summarize briefly, the following predictions are consistent with our hypotheses: * Overall environmental performance should be positively correlated with: 1) Income per capita; 2) Degree of popular representation; 3) Freedom of information; 4) Security of property rights; 5) Development of the legal and regulatory system. Controlling for these variables, * G.reen indices should be positively correlated with: 1) Rural population density; 2) Agricultural and forest production share of national output. * Brown indices should be positively correlated with: 1) Particular focus on public health, indexed by life expectancy4; 2) Urban share of total population; 3) Urban population density; 4) Manufacturing share of national output. 4 We recognize some risk of endogeneity, but we regard it as minimal in this case. Life expectancy is influenced by many policy and other variables which are not directly related to environmental concerns. 12 5. Results 5.1 Income and Environmental Performance The correlation between income and composite environmental rankings is clear in Table 2A. Comparisons of bivariate regressions on the two income measures, recorded in Tables 5A and 5B, reveal significantly tighter fits for ICPGDP. The income elasticity of environmental policy performance is positive and highly significant in all environmental dimensions. Air seems to have a much higher income elasticity than the others. The scatter of the composite environmental index (Env) against ICPGDP (Figure 1) indicates that the relationship is continuous over the entire range of incomes. 5.2 Political Economy and Institutional Variables For the reasons previously noted, effective environmental management may be seriously handicapped by lack of political, civil, and economic liberty; lack of an independent judicial system; and an inefficient or corrupt bureaucracy. To test these ideas, we have fitted regressions with several sets of institutional indicators previously used in the literature. In each case, limited availability of the indicators has forced us to run regressions on subsamples of countries. Our first test employs a widely-used set of political, civil and economic liberty indicators developed by Gastil.5 These 5 See Scully (1992) for details. 13 Table 5A Impact of PCGNP on Environmental Indicators Dependent Intercept In PCGNP Adjusted R2 Variable . ln Air 2.70 0.27 0.71 (11.93) (8.70) ln Water 3.55 0.19 0.72 (22.84) (8.80) ln Land 3.79 0.17 0.72 (27.70) (8.75) _ ln Living 3.73 0.16 0.74 (29.60) (9.26) ln Env 4.89 0.19 0.76 (34.80) (9.78) * t-statistics in parentheses. Table 5B Impact of ICPGDP on Environmental Indicators Dependent Intercept ln ICPGDP Adjusted R2 Variable ln Air 1.29 0.42 0.79 (4.06) (10.59) . ln Water 2.59 0.30 0.78 (11.53) (10.30) ln Land 2.97 0.25 0.76 (14.52) (9.82) __ ln Living 3.03 0.23 0.71 l ______________ (13.88) (8.53) ln Env 3.97 0.29 0.79 (18.72) (10.79) 14 Figure 1 Overall Environmental Performance vs. ICP Income Per Capita 7.00 *. I . *~~~~~~~~~~~~~. 9 6.50! * 9 X 6.00 9 _- I 9 9 5.50 5.00 6.00 7.00 8.00 9.00 10.00 11.00 ln ICPGDP 15 indicators are available for 29 of our selected 31 countries. Among the aspects that appear most relevant for our study are: freedom of property (FOP), freedom of information (FOI), freedom of print media (FPM), freedom of broadcast media (FBM), freedom of peaceful assembly (FPA) and the Gastil-Wright classification of types of economic system (TES) by degree of commercial freedom. In our regressions, only FOP and FOI are statistically significant (Table 6). Each of these indicators is coded 1 to 5, with higher scores for lower liberty, so the expected sign of the coefficients is negative for both indicators. Freedom of property has the expected sign, but the other result is quite surprising: Controlling for income and property rights, greater freedom of information is associated with lower environmental index values. We have no explanation for this anomaly, and we have dropped FOI from our final regressions (Table 9). Table 6 Impact of Liberty Indexes on Environmental Indicators Dependent Intercept ln ICPGDP ln FOP ln FOI Adjusted Variable R2 ln Air 1.42 0.41 -0.36 0.27 0.80 (2.97) (8.17) (-2.39) (2.24) ln Water 2.86 0.27 -0.26 0.18 0.82 (9.54) (8.44) (-2.80) (2.38) ln Land 3.17 0.23 -0.18 0.12 0.77 (10.28) (7.16) (-1.90) (1.57) ln Living 3.22 0.22 -0.27 0.16 0.74 (9.57) (6.27) (-2.57) (1.90) ln Env 4.18 0.27 -0.26 0.18 0.82 ._ _ (13.43) (8.25) (-2.72) (2.25) 16 As a second test, we have employed measures of bureaucratic delay and contract enforceability (or relative degree to which contractual agreements are honored) from Business Environmental Risk Intelligence, Inc. (BERI) ,h Scores for the BERI indicators are available for only fourteen of our thirty-one countries and are set so thlat positive relationships with environmental Table 7 Impact of BERI Indexes on Environmental Indicators Dependent Intercept ln ln Delay| in Adjusted Variable ICPGDP Contract R2 ln Air 1.99 0.32 0.19 0.81 (3.48) (3.23) (0.56) ln Water 3.21 0.18 0.31 0.72 (6.19) (2.04) (1.00) ln Land 3.25 0.20 0.18 0.68 l__________ (6.18) (2.19) (0.57) ln Living 2.99 0.21 0.24 0.66 ____ __ (4.87) (1.99) (0.64) . ln Env 4.29 0.22 0.23 0.74 l__________ (7.96) (2.40) (0.72) ln Air 2.05 0.32 0.16 0.81 (2.24) (2.10) (0.34) ln Water 3.45 0.15 0.35 0.72 (4.15) (1.11) (0.82) ln Land 3.43 0.18 0.22 0.68 (4.12) (1.26) (0.52) ln Living 3.01 0.22 0.17 0.65 l__________ (3.06) (1.34) (0.33) ln Env 4.42 0.21 0.23 0.73 (5.13) (1.47) (0.52) _ 6 For a discussion of these indicators, see Keefer and Knack (1993). 17 management would be consistent with our prior hypotheses about the effect of judicial and administrative efficiency. The regression coefficients are positive, as expected, but none are statistically significant (Table 7). Finally, we have tested a set of indicators which directly reflect the efficiency of the legal and judicial system (LJS) and the level of red tape in the bureaucracy (RTB). These were developed by the Country Assessment Service of Business International, Inc.7 Unfortunately, the measures are available for only twelve of the thirty-one countries in our sample. In separate regressions for this subset of countries, both LJS and RTB emerge as significant explanatory variables. Since they are collinear, we have computed their first principal component (PC1) and used it as a composite regressor. When it is included with ICPGDP (Table 8) the results show substantial improvement in the explanatory power of the regressions: The adjusted R2 increases between 9% and 24%. The change in outliers indicates that the improvement is especially striking for Ireland, India and Thailand. 5.3 Green and Brown Indices For both Green and Brown indices, the regressions reported in Table 9 suggest that performance is again strongly associated See Wheeler and Mody (1992) for details. 18 with income per capita, freedom of property and (in small samples) measures of regulatory efficiency. The two rural-sector variables (population density; proportion of GDP in agriculture and forestry) are only weakly associated with the Green index (Table 9a). The fit is much better for the Brown index: degree of urbanization, population density and manufacturing share in GDP all have the expected signs and relatively high significance (Table 9b). Life expectancy as a proxy for public health priority has no independent effect. 6. Summary Using a multidimensional survey analysis of the UNCED reports, we have developed a set of comparative indices of environmental policy and performance in thirty-one countries. We find a strong positive correlation between our environmental indicators and the level of economic development. The fit is substantially better when national incomes are adjusted for purchasing power parity. The income elasticity of the indices is positive and highly significant in all environmental dimensions. The pattern of elasticities suggests that protection measures for land and living resources precede those for water; action for reducing air pollution comes later. Some impact for institutional development is also suggested by our results, although the information base is quite limited. 19 Table 8 Impact of ICPGDP, LJS and RTB oi Environmental Indicators Dependent Intercept ln ICPGDP PCi Adjusted R2 Variable _ ln Air 1.60 0.38 0.76 ____________ (2.91) (6.02) In Air 3.35 0.18 0.26 0.95 l___________ (8.81) (4.07) (6.18) In Water 2.59 0.29 0.72 l ___________ (5.57) (5.35) ln Water 4.13 0.11 0.23 0.96 (16.68) (3.73) (8.37) ln Land 2.79 0.27 0.70 ____________ (6.19) (5.16) ln Land 4.20 0.10 0.21 0.93 ____________ (13.15) (2.78) (5.96) in Living 2.79 0.27 0.70 ('.19) (5.16) ln Living 4.05 0.11 0.24 0.90 l___________ (9.12) (2.15) (4.91) _ ln Env 3.77 0.31 0.73 (7.79) (5.48) ln Env 5.35 0.12 0.23 0.95 _L. _ ____(18.08) (3.58) (7.15) The level of explanation in all regressions improves significantly with the addition of the Business International effectiveness indices for legal/judicial and administrative systems and the Gastil measure of property rights protection. Similar BERI measures are not significant, however. We also obtain insignificant or perverse results for all Gastil measures of degree of popular representation and freedom of information. 20 Table 9a Regression Results for ln(Green) Intercept InPCGNP ln ICPGDP In POP ln(Shoro of In (Pop. Adjusted agriculture density) RI __________ in GDP) _ -_______ _______nG P 3.31 0.16 0.71 (25.55) (8.66) _.... 2.60 0.23 0.71 (12.29) (8.65) 2.75 0.20 -0.11 0.06 0.09 0 64 (4.69) (3.85) (-1.31) (0.93) (1.32) 3.27 0.17 -0.16 0.09 0.73 (11.11) . . | (5.38) (2.19) _ _ (1.34) Table 9b Regression Results for ln(Brown) Intercp lnPCGNP lnICPGDP ln FOP ln(Urban ln(Popu- in ln(Life Adj Rz /total lation (Manuf. expect- popula- density) share ancy) ________ ____ ____ _______ tion) of GDP) _ _ _ _ 3.81 0.21 0.76 (24.25) (9.7S) 2.73 0.32 0.82 (12.40) (11.75) 3.91 0.20 -0.19 0.14 0.06 0.16 -0.34 0.82 (2.63) (2.27) (1.98) (1.46) (2.30) (2.04) (-0.67) 2.94 0.16 -0.20 0.14 0.06 0.15 0.83 (8.02) ._ !2.65) (2.20) (1.46) (2.25) (1.95) . Table 9c Green/Brown Impacts of ICPGDP, FOP and Regulatory Efficiency v Yariable |ntercept ln ICPGDP ln FOP tln RTB ln LSJ Adj R2 ln(Green) 3.84 0.03 -0.17 0.39 0.93 ____________ 1(9.37) (0.52) (1.93) j(3.37) ln(Brown) 3.95 0.09 -0.07 0.36 0.14 0.98 (9.44) 1 (2.69) - (1.09) (4.20) (1.07) _ J 21 Decomposition of overall environmental performance into Brown and Green sectors yields some additional insight into the impact of demographics and economic structure on regulation. Controlling for income, comparative analysis of the Brown sector indices suggests a very significant country response to environmental pressures from industrialization and urbanization. However, our results do not reveal an equivalent response on the Green side beyond the effect of variations in income per capita. In summary, our findings suggest that a detailed, quantified analysis of the UNCED reports can yield comparable and plausible indices of environmental policy performance across countries. Cross-country variations in our environmental index are well- explained by variations in income per capita, de:-ree of urbanization and industrialization, security of property rights, and general administrative efficiency. 22 7. References Keefer, P. and Knack, S., 1993, "Why Don't Poor Countries Catch Up? A Cross Country National Test of an Institutional Explanation", Center for Institutional Reform and the Informal Sector, University of Maryland at College Park, Working Paper No.60 Scully, G.W., 1992, Constitutional Environments and Economic Growth, (Princeton: Princeton University Press) UNCED, 1990, Draft Format for National Reoorts (Geneva: United Nations Conference on Environment and Development) UNEP, 1992, Savinq Our Planet: Challenges and Hopes (Nairobi: United Nations Environment Programme) Wheeler, D. and A. Mody, 1992, "International Investment Location Decisions", Journal of International Economics, 33, pp. 57-76 World Bank, 1992, World Development Report 1992: Development and the Environment, (World Bank, Washington D.C. and Oxford University Press, New York) World Resources Institute, 1992, World Resources 1992-93, (World Resource Institute, New York and Oxford University Press, New York) 23 APPENDIX Questionnaire for Evaluating Environmental Policy Performance 1. AWARENESS A. When did environmental awareness gain prominence? 2 Pre 1972 1 1972-89 O 1990+ B. How widespread is this awareness at present? 2 Mass awareness countrywide 1 Restricted to limited pockets of elite groups 0 Very little awareness C. The extent of awareness regarding global dimensions 2 Excellent 1 Reasonable 0 Very little 2. POLICY A. For how long has significant environmental policy existed? 2 Dates back to 1970s 1 Introduced in the last ten years 0 Very little so far B. How did the policy evolve? 2 As a felt need 1 Of late as a result of diffusion of knowledge 0 Yet to evolve significantly C. What is the coverage of the policy? 2 Comprehensive with clearly laid down targets 1 Some policy and some targets 0 Very little policy 24 3. LEGISLATION A. When did significant environmental legislation begin to be enacted? 2 Dates back to 1970s 1 Introduced in the last ten years 0 Very little so far B. How extensive is the legislation so far? 2 Comprehensive and supported by detailed rules and regulations 1 Sketchy; some rules and regulations 0 Only a few or none at all C. What is the extent of machinery for enforcement of laws? 2 Agency clearly entrusted with specified guidelines 1 Agency set up but yet to develop effectively o No agency or very little effort so far 4. CONTROL MECHANISM A. What is the nature of regulatory instruments? 2 Both command and control as well as economic 1 Only command and control 0 Hardly any mechanism B. What is the extent of power vested in the environmental protection agency? 2 Both formulation of policy as well as its enforcement 1 Only limited to policy 0 No agency or very little power C. What is the degree of decentralization of such an agency? 2 Extensive 1 Somewhat o Very little 25 D. What is the extent of allocation of funds to the agency? 2 Reasonably good for carrying out allotted tasks 1 Some but not enough for effective functioning o None or very little E. What is the extent of self regulation by polluters? 2 Extensive 1 Somewhat o Very little F. How widespread is the involvement of NGOs in regulation? 2 Extensive 1 Somewhat o Very little G. What is the progress of preparation of a national environmental action plan (NEAP)? 2 NEAP with detailed plans for identifiable regions have been prepared 1 Only a sketchy NEAP or plans for some regions O No action so far 5. MEASURE OF SUCCESS A. What is the trend in environmental indicators? 2 Improving 1 Not much headway but steady 0 Deteriorating B. Roughly what percentage of GDP is being devoted for environmental control measures? 2 More than 1% 1 Some but less than 1% 0 Almost none C. What is the market share of pollution control industries in total industrial production? 2 Above the global average 1 Around average 0 Below average 26 D. What is the prevalence of environmental incidents/accidents? 2 Almost none 1 A few O Considerable E. How good is the availability of environmental data? 2 Extensively compiled 1 Sporadically available O None or very little F. What is the extent of interest in environmental studies and R & D? 2 Widespread 1 Somewhat O None or very little G. How widespread is the involvement of NGOs in the environmental movement? 2 Considerable 1 Somewhat 0 None or very little H. What is the prevalence of environmental litigation? 2 Considerable 1 Somewhat O None or very little I. What is the level of media interest in environmental issues? 2 Very high 1 Somewhat O None or very little 27 Policy Research Working Paper Series Contact Title Author Date for paper WPS1425 On the Intersectoral Migration ol Donald Larson February 1995 J. Jacobson Agricultural Labor Yair Mundlak 33710 WPS1426 Russian Unemployment: its Simon Commander February 1995 V. Reid Magnitude, Characteristics, and Ruslan Yemtsov 35195 Regional Dimensions WPS1427 Corporate Govemance and Equity Stijn Claessens February 1995 F. Hatab Prices: Evidence from the Czech 35835 and Slovak Republics WPS1428 Short-Term Supply Response to a Bruno Boccara February 1995 M. Pfeiffenberger Devaluation: A Model's Implications Fabien Nsengiyumva 34963 for Primary Commodity-Exporting Developing Countries WPS1429 The World Trade Organization's Bemard M. Hoekman March 1995 F. Hatab Agreement on Government Petros C. Mavroidis 38535 Procurement: Expanding Disciplines, Declining Membership? WPS1 430 Intergovemmental Fiscal Relations Richard M. Bird March 1995 G. Coward and Poverty Alleviation in Viet Nam Jennie I. Lilvack 80494 M. Govinda Rao WPS1431 The Industrial Pollution Projection Hemamala Hettige March 1995 A. Williams System Paul Martin 37176 Manjula Singh David Wheeler WPS1432 Using Financial Futures in Trading Ignacio Mas March 1995 K. Binkley and Risk Management Jesus Saa-Requejo 81143 WPS1433 Enterprise Restructuring in Eastern Bernard Hoekman March 1995 F. Hatab Europe: How Much? How Fast? Gerhard Pohl 35835 Where? Preliminary Evidence from Trade Data WPS1434 Govemment Credit Policy and Charles W. Calomiris March 1995 P. Infante Industrial Performance (Japanese Charles P. Himmelberg 37642 Machine Tool Producers, 1963-91) WPS1435 The Political Economy of Formal Arup Banerii March 1995 D. Evans Sector Pay and Employment in J. Edgardo Campos 37496 Developing Countres Richard H. Sabot WPS1436 An Empirical Model of Sunk Costs Mark J. Roberts March 1995 J. Ngaine and the Decision to Export James R. Tybout 37959 Policy Research Working Paper Series (.:;;ntact ritle Author Date fur paper WPS143/ Travel Mode Substitution in Sao Joft.e Swait March 1995 C loiies Paulo. Estimates and Implications Gunnar S. Eskeland 37699 for Air Pollution Control WPS14S8 Trade Relorm, Efficiency, and Growth Ejaz Ghani March 1995 A Nokhostin Carl Jayarajah 34150 WPS1439 Nontarif Barriers Africa Faces: What Azita Amiadi March 1995 S Lipscomb Did the Uruguay Round Accomplish, Alexander Yeats 33 A18 and What Remains to Be Done? WPS1440 Poverty and Social Transfers in Christiaan Grootaert March 1995 N. Sachdeva Poland 82717 WPS1441 The Significance of Credits and Douglas Galbi March 1995 N. Castillo Subsidies in Russian Agricultural 33490 Reform WPS1 442 Energy Price Increases in Einar Hope March1995 C. Jones Developing Countries: Case Studies Balbir Singh 37699 of Co!ombia. Ghana. Indonesia. Malaysia. Turkey. and Zimbabwe WPS 1443 Policy-Based Finance, Financial Dimitri Vittas April 1995 P. Infante Regulation, and Financial Sector Akihiko Kawaura 37642 Development in Japan WPS1444 Roads, Lands, Markets, and Kenneth M. Chomitz April 1995 E. Schaper Deforestation: A Spatial Model David A. Gray 33457 ol Land Use in Belize WPS1445 Human Capital and Industry Wage Chris N. Sakellariou April 1995 I. Conachy Structure in Guatemala 33669 WPS1446 Review of Integrated Approaches Donna J. Lee April 1995 C. Spooner to River Basin Planning, Ariel Dinar 32116 Development, and Management WPS1447 Environmental Inspections and Benoit Laplante April 1995 E. SchaDer Emissions of the Pulp and Paper Paul Rilstone 33457 Industry: The Case of Quebec WPS1448 Environmental Regulation and Susmita Dasgupta April 1995 E Schaper Development: A Cross-Country Ashoka Mody 33457 Empirical Analysis Subhendu Roy David Wheeler