Information about Europe Europa
Chapter

5 How the Environment Affects the Macroeconomy

Editor(s):
Ved Gandhi
Published Date:
June 1996
Share
  • ShareShare
Information about Europe Europa
Show Summary Details
Author(s)
David Pearce and Kirk Hamilton 

Standard texts in economic development have little or nothing to say about the natural or non-built environment. A well-known review of the state of development economics, for example, contains neither an essay on nor mention of the environment-development link (Ranis and Schultz, 1988). The same is true of more recent surveys (e.g., Balasubramanian and Lall, 1991). Those texts that have taken the environment on board provide a welcome relief to the general picture, but even here the analysis tends to be confined to descriptive issues (e.g., Hogendorn, 1992). In contrast, officially sponsored documentation on the environment-development interface, such as that from the World Bank, the United Nations, or Organization for Economic Cooperation and Development, has increased dramatically in recent years (Pearce and Warford, 1993; Eröcal, 1991; Bartelmus, 1986; Pearce, Whittington, and Georgiou, 1994), not to mention the ever-growing number of texts specifically on economics and the environment. All this suggests that, despite the efforts of those who have sought to uncover the environment-development-environment interactions, much of this work has still to filter down to those who effectively provide introductions to development planning, and, perhaps more seriously, to those who define the cutting edge of development economics.

This past neglect of the environment in development economics has a human cost. We argue below that this cost shows up as the deterioration of human capital through ill health and premature mortality arising from environmental risks, as forgone GNP because of the failure to recognize the high economic rate of return to many environmental investments, and as the erosion of the natural capital base on which the development of many economies depends.

Environmental Deterioration and Erosion of Human Capital

A growing body of epidemiological studies is identifying substantial economic costs from urban air pollution in the developing world. Table 1 assembles results from several city studies of health damage from air pollution. A UK national damage estimate is shown for reference. Valuing health symptoms and risks of mortality in economic terms is controversial and has additional complications in developing country contexts. Essentially, if life and health appear to have a low value generally, this will show up in the economic magnitudes and may give rise to the view that one developing country resident is worth less than one industrial country resident. But as proportions of available income, there is no reason to suppose that willingness to pay for avoiding ill health is less in a low-income country than in a high-income country. Moreover, what matters in most policy decisions is the set of priorities for action determined by the country in question. A willingness to pay offers one way of setting such priorities.

Table 1.Air Pollution Damage to Human Health in Cities
Mortality CostMorbidity CostTotal Health costCostCost per Capita
Coverage(In millions of U.S. dollars)(In percent of GDP)(In U.S. dollars)
Particulate matter
Costs of all exposure
UK 1993111,8009,40021,1502.7848
China 1990241,67019,30060,97011.152
Cairo 1990s3186–992157–472343–1,464n.a.38–161
Jakarta 1990411344157n.a.19
Benefit of percentage reduction
Mexico City 19905480358850n.a.50
58 percent reduction
Bangkok 19896138–1,315302–309440–1,624n.a.57–209
20 percent reduction
Santiago 1990s 786270n.a.15
15 percent reduction
Sulphur oxides
Bangkok 198900.20.2n.a.<1
6 percent reduction
Santiago 1990s700.10.1n.a.<1
8 percent reduction
Nitrogen oxides
Jakarta 19904011n.a.<1
Santiago 1990s701 or 17*1 or 18n.a.<1 or 4
49 percent reduction
Ozone
Cairo3011**11n.a.1
50 percent reduction
Mexico City 199050102102n.a.6
21 percent reduction
Bangkok 1990s609–369–36n.a.1–5
20 percent reduction
Santiago 1990s703333n.a.7
69 percent reduction in VOCs and 49 percent reduction in NOx
Lead
Jakarta 1990s4263662n.a.8
All exposure
Mexico 19895n.a.125–130125–130n.a.7–8
All exposure
Bangkok 1990s6291–1,4706–8297–1,478n.a.39–193
20 percent reduction
Total health damage costs of air pollution
Cairo3186–992157–472343–1,464n.a.38–161
(50 all exposure, PM only)
Jakarta413882220n.a.27
(All exposure, PM, lead, NOx)
Mexico54805901,070n.a.63
(All exposure, PM, lead, ozone)
Bangkok6429–2,785317–353746–3,138n.a.97–402
(Benefits of 20 percent reduction in PM, lead, SOx, and ozone)
Santiago7896 or 112104 or 120n.a.22–25
(Benefits of package of measures)

D.W. Pearce and T. Crowards, “Assessing the Health Costs of Particulate Air Pollution in the 1 United Kingdom," Centre for Social and Economic Research on the Global Environment, University College London and University of East Anglia, 1995, unpublished. Assumes a value of statistical life of $2.25m and population at risk of 25.2 million. Damage done by PM10 only.

H.K. Florig, The Benefits of Air Pollution Reduction in China," Resources for the Future, Washington, D.C., 1993, unpublished. Adjustments made to the original estimates to give a value of statistical life of $45,547 based on a US/UK VOSL of $2.25m multiplied by the ratio of GNP per capita in China to GNP per capita in the United States. For a justification for using this ratio see A. Alberini and others, Valuing Health Effects of Air Pollution in Developing Countries: The Case of Taiwan, Resources for the Future, Washington, D.C., Discussion Paper 95–01, 1995. Morbidity effects are restricted activity days (RADs), which are valued at a daily GNP per capita of $1.29.

Estimate s of mortality and RADs taken from Chemonics International and Associates, Com- paring Environmental Health Risks in Cairo, Egypt, Vols. 1 and 2, Report to US AID, Egypt, September 1994. Value of statistical life taken to be $2.25m x GNP per capita Egypt/GNP per capita USA= $62,021. RADs valued at daily GNP per capita of $1.75 per day. Population taken to be 9.08 million. Estimates of hospital admissions from ozone polution valued at $260; minor restricted activity days and days or respiratory symptoms valued at $0.4, asthma attacks valued at $2.5. Unit values taken from (Sources in footnote 4) as GNP per capita in Egypt and Indonesia is very similar.

World Bank, “Indonesia Environment and Development: Challenges for the Future," Environment Unit, Country Department III, East Asia and Pacific Region, World Bank, Washington D.C., March, 1994. Value of statistical life of $75,000 and population at risk of 8.2 million. Morbidity effects include RADs, outpatient visits, hospital admissions, respiratory illness among children, asthma attacks, and respiratory symptoms. See also B. Ostro, “Estimating Health Effects of Air Pollution: A Methodology with an Application to Jakarta," PRDPE, World Bank, Washington, D.C., March 1994.

World Bank, “Thailand: Mitigating Pollution and Congestion Impacts in a High Growth Economy," Country Operations Division, Country Department 1, East Asia and Pacific Region, World Bank, Washington D.C., February 1994. Bangkok population of 7.67 million assumed. Value of statistical life of $336,000 based on compensating wage differentials in Bangkok for risky occupations.

S. Margulis, “Back of the Envelope Estimates of Environmental Damage Costs in Mexico," Working Paper WPS 824, Country Department 2, Latin America and the Caribbean Regional Office, World Bank, January 1992. Value of statistical life of $75,000 assumed based on human capital approach. Population of 17 million assumed.

World Bank, “Chile: Managing Environmental Problems—Economic Analysis of Selected Issues," Environment and Urban Development Division, Country Department 1, Latin America and the Caribbean Region, World Bank, Washington D.C., December 1994. Estimates are based on dose-response functions for mortality and morbidity converted to work days lost, each work day being valued at US$9.55. Population of Santiago taken to be 4.8 million. Control costs for this package of measures were estimated at $60m, so that, even without considering other pollutants, the benefits of reduced PM10 exceed the costs of control. Other benefits arise from the associated control of ozone, NOx, and SOx. Alternative estimate for NOx assumes NOx is credited with half the benefits of avoided ozone pollution damage.

D.W. Pearce and T. Crowards, “Assessing the Health Costs of Particulate Air Pollution in the 1 United Kingdom," Centre for Social and Economic Research on the Global Environment, University College London and University of East Anglia, 1995, unpublished. Assumes a value of statistical life of $2.25m and population at risk of 25.2 million. Damage done by PM10 only.

H.K. Florig, The Benefits of Air Pollution Reduction in China," Resources for the Future, Washington, D.C., 1993, unpublished. Adjustments made to the original estimates to give a value of statistical life of $45,547 based on a US/UK VOSL of $2.25m multiplied by the ratio of GNP per capita in China to GNP per capita in the United States. For a justification for using this ratio see A. Alberini and others, Valuing Health Effects of Air Pollution in Developing Countries: The Case of Taiwan, Resources for the Future, Washington, D.C., Discussion Paper 95–01, 1995. Morbidity effects are restricted activity days (RADs), which are valued at a daily GNP per capita of $1.29.

Estimate s of mortality and RADs taken from Chemonics International and Associates, Com- paring Environmental Health Risks in Cairo, Egypt, Vols. 1 and 2, Report to US AID, Egypt, September 1994. Value of statistical life taken to be $2.25m x GNP per capita Egypt/GNP per capita USA= $62,021. RADs valued at daily GNP per capita of $1.75 per day. Population taken to be 9.08 million. Estimates of hospital admissions from ozone polution valued at $260; minor restricted activity days and days or respiratory symptoms valued at $0.4, asthma attacks valued at $2.5. Unit values taken from (Sources in footnote 4) as GNP per capita in Egypt and Indonesia is very similar.

World Bank, “Indonesia Environment and Development: Challenges for the Future," Environment Unit, Country Department III, East Asia and Pacific Region, World Bank, Washington D.C., March, 1994. Value of statistical life of $75,000 and population at risk of 8.2 million. Morbidity effects include RADs, outpatient visits, hospital admissions, respiratory illness among children, asthma attacks, and respiratory symptoms. See also B. Ostro, “Estimating Health Effects of Air Pollution: A Methodology with an Application to Jakarta," PRDPE, World Bank, Washington, D.C., March 1994.

World Bank, “Thailand: Mitigating Pollution and Congestion Impacts in a High Growth Economy," Country Operations Division, Country Department 1, East Asia and Pacific Region, World Bank, Washington D.C., February 1994. Bangkok population of 7.67 million assumed. Value of statistical life of $336,000 based on compensating wage differentials in Bangkok for risky occupations.

S. Margulis, “Back of the Envelope Estimates of Environmental Damage Costs in Mexico," Working Paper WPS 824, Country Department 2, Latin America and the Caribbean Regional Office, World Bank, January 1992. Value of statistical life of $75,000 assumed based on human capital approach. Population of 17 million assumed.

World Bank, “Chile: Managing Environmental Problems—Economic Analysis of Selected Issues," Environment and Urban Development Division, Country Department 1, Latin America and the Caribbean Region, World Bank, Washington D.C., December 1994. Estimates are based on dose-response functions for mortality and morbidity converted to work days lost, each work day being valued at US$9.55. Population of Santiago taken to be 4.8 million. Control costs for this package of measures were estimated at $60m, so that, even without considering other pollutants, the benefits of reduced PM10 exceed the costs of control. Other benefits arise from the associated control of ozone, NOx, and SOx. Alternative estimate for NOx assumes NOx is credited with half the benefits of avoided ozone pollution damage.

Table 1 suggests several conclusions about priorities in air pollution control. First, two air pollutants, particulate matter and lead, are especially damaging. The exact manner in which particulate matter may cause health damage is not known with certainty, but statistical associations with respiratory illness and premature death are strong. Because the studies tend to use consensus dose-response functions, overall damage costs will be mainly determined by the size of the population at risk and the unit economic values used to value symptoms and statistical life. Costs per capita in developing countries appear fairly consistent in the range of $20–160, while the estimates for Bangkok suggest that overall damages could be very much higher. If the dose-response function is linear through the origin, as much of the epidemiology for PM10 suggests, per capita damage could be as high as $1,000 in Bangkok ($209 is for a 20 percent reduction). Lead damage is similarly high in the Bangkok study at $39–193 per capita. The Mexico and Jakarta studies show much lower per capita damages at around $8 per person. Health damage from lead shows up in the form of reduced IQ scores in children, hypertension in adults (mainly males), coronary heart disease, and mortality. Table 1 suggests that tropospheric ozone may rank as a significant health-damaging pollutant, with consistent per capita cost estimates of perhaps $5 per person. Ozone is a secondary pollutant where main precursors are NOx and volatile organic compounds (VOCs). In some studies, damage from VOCs is allocated to these two pollutants, although the allocation rule appears not to be generally agreed upon. Nitrogen oxides and sulphur oxides are generally revealed to be of limited relevance for health damage.

These observations bear on policy. They suggest that priority actions in countries where Western ambient standards are not met should focus on particulate matter and lead. The former can be addressed by energy conservation and vehicle-traffic control, especially the latter, since inhalable particulate matter tends to be associated with vehicle emissions rather than stationary sources, such as power stations. The practice of requiring new power stations in Eastern Europe and developing countries to be fitted with flue gas desulphurization equipment (FGD) becomes questionable since SOx is not seen to be associated with significant health damage, although very high concentrations in black spots probably do have health consequences. Lead emissions can be addressed primarily through reductions in the lead content of gasoline. Overall, then, the air pollution studies suggest a focus on transport rather than on traditional power stations, although health damage from the latter can be significant.

The epidemiology of water pollution is well understood, but estimating dose-response functions for waterborne pollution and human health remains very difficult. Contamination of water supplies has known health effects, but these effects can vary substantially according to personal hygiene behavior and the amount of water available. It has been estimated that in 1979 some 360–400 billion working days were lost in Africa, South America, and Asia because of water-related diseases. At even a nominal 50 cents per day, this suggests that these continents lost $180–200 billion in forgone GNP each year. The combined GNP of the three continents in 1979 was about $370 billion, so that GNP was a staggering 35 percent below its potential value because of waterborne diseases alone. This figure would be offset only partially by the costs of treating water supplies and avoiding contamination.

A review of over 80 studies of water quality and quantity reveals that improved water and sanitation can be expected to reduce diarrhoeal mortality by 55–60 percent and morbidity by 25 percent (Esrey, 1990). Table 2 assembles some estimates of water pollution control benefits for two cities and one country. The Mexico study suggests that the benefits of control could be substantial and on a par with those from the control of particulate matter air pollution and lead exposure. A review of past policies in Chile on typhoid suggests that the modest actions taken in 1984 were justified on benefit-cost grounds, while more drastic action might well have had costs in excess of benefits. In 1991 a cholera outbreak prompted quick and effective action on educating the public concerning the consumption of unwashed vegetables and the banning of sewage irrigation for growing vegetables. Benefit-cost analysis suggests a benefit-cost ratio of 5:1 in favor of the emergency actions, with most of the benefits coming from the avoided costs of restrictions on food exports.

Table 2.Water Pollution Damage to Health, Selected Cities
Mortality CostMorbidity CostTotal Health costCostCost per Capita
Coverage(In millions of U.S. dollars)(In percent of GDP)(In U.S. dollars)
Jakarta13003303n.a.37
Fecal contamination
Mexico23,600small3,600n.a.212
Intestinal disease
Santiago3
Typhoid
1985–900.50.91.4n.a.<1
19910.50.70.9n.a.<1

World Bank, “Indonesia Environment and Development: Challenges for the Future," Environment Unit, Country Department III, East Asia and Pacific Region, World Bank, Washington D.C., March, 1994. Value of statistical life of $75,000 and population at risk of 8.2 million. Assumes 7,000 diarrhea related deaths per year. Improved water quality and sanitation can reduce such deaths by 55–60 percent a year, so that 3,800–4,200 deaths could be avoided. Some 360,000 fewer diarrhoeal episodes per year are estimated to be saved by improved water quality.

S. Margulis, “Back of the Envelope Estimates of Environmental Damage Costs in Mexico,” Working Paper WPS 824, Country Department 2, Latin America and the Caribbean Regional Office, World Bank, January 1992. Value of statistical life of $75,000 assumed based on human capital approach and applied to the whole of Mexico.

World Bank, “Chile: Managing Environmental Problems—Economic Analysis of Selected Issues,” Environment and Urban Development Division, Country Department 1, Latin America and the Caribbean Region, World Bank, Washington D.C., December 1994. Direct costs of typhoid only and based on forgone production, that is, the human capital approach. Indirect effects excluded from health effects include effects on farm profitability from loss of exports of farm produce.

World Bank, “Indonesia Environment and Development: Challenges for the Future," Environment Unit, Country Department III, East Asia and Pacific Region, World Bank, Washington D.C., March, 1994. Value of statistical life of $75,000 and population at risk of 8.2 million. Assumes 7,000 diarrhea related deaths per year. Improved water quality and sanitation can reduce such deaths by 55–60 percent a year, so that 3,800–4,200 deaths could be avoided. Some 360,000 fewer diarrhoeal episodes per year are estimated to be saved by improved water quality.

S. Margulis, “Back of the Envelope Estimates of Environmental Damage Costs in Mexico,” Working Paper WPS 824, Country Department 2, Latin America and the Caribbean Regional Office, World Bank, January 1992. Value of statistical life of $75,000 assumed based on human capital approach and applied to the whole of Mexico.

World Bank, “Chile: Managing Environmental Problems—Economic Analysis of Selected Issues,” Environment and Urban Development Division, Country Department 1, Latin America and the Caribbean Region, World Bank, Washington D.C., December 1994. Direct costs of typhoid only and based on forgone production, that is, the human capital approach. Indirect effects excluded from health effects include effects on farm profitability from loss of exports of farm produce.

The limited analyses available on the control of water pollution in cities suggest that economic damages to human health can be very large and benefit-cost ratios typically favor intervention in water quality treatment. In terms of the human capital base, the air and water pollution studies suggest that both kinds of pollution are taking a heavy toll on human life and well-being. If so, pollution control is not a “luxury good” to be afforded after the development process has taken off, but a prior requirement for sustainable development. This suggests a wholly different picture to that implied by the neglect of environment in development economics generally.

Rate of Return on Environmental Investments

The social costs of health damage from pollution already hint at the high rate of return on environmental expenditures. These returns can, by and large, be expected to show up in conventional market terms as gains in GNP as traditionally measured. Other environmental investments, such as afforestation, may also generate high rates of return in conventional terms. The challenge for the analyst is to cast the net wide enough to ensure that all benefits and costs are measured. Anderson’s study of afforestation in Northern Nigeria is a good example (Anderson, 1987, 1989). The benefits of afforestation included:

  • halting the erosion of fertile soil (since trees typically reduce erosion)
  • raising current levels of soil fertility
  • producing tree products—fuelwood, poles, fruits
  • producing fodder both from increased productivity of soils and from the use of forest fodder.

The economic rates of return (in percent) that resulted for shelterbelts (planting trees mainly for wind protection) and farm forestry (intermixing trees and crops) are indicated in the following table:

ShelterbeltFarm Forestry
Base case14.919.1
Low yield, high cost13.114.5
High yield16.2
No erosion13.516.6
More rapid erosion13.615.5
Soil restored + yield jump16.921.8
Wood and fruit benefits4.77.4

Calculation of timber costs and benefits alone in the Kano area has tended to show rates of return of around 5 percent, which has to be compared with the cut-off rate, usually double this rate. In other words, afforestation for timber production does not pay. But once the other benefits are included, dramatic increases in rates of return can be secured. The analysis shows that counting wood benefits only produces negative net present value and correspondingly low economic rates of return. But if allowance is made for the effects of trees on crop yields and for expected rates of soil erosion in the absence of afforestation, the picture is transformed for both farm forestry and shelterbelts. Other examples of high return environmental investments are given in Pearce, 1993.

Capturing Nonmarket Returns

Many environmental investments will yield mixes of market and nonmarket returns, where the distinction between market and nonmarket reflects the essentially arbitrary dividing line set by the way conventional accounting systems operate. While economists are right to insist that nonmarket gains and losses are intrinsically no different from market gains and losses (both translate into changes in the state of human well-being), nonmarket effects are by definition not traded and hence do not translate into empowerment over resources. Put simply, they do not by definition have associated cash (or in-kind) flows. This is why demonstrating the monetary value of environmental assets is not enough if change in the development planning ethos is to take place. It is as important to show how such nonmarket values can be captured by the developing countries. The issue of capture is now the subject of an emerging literature that aims to find ways in which developing countries can exploit their comparative advantage in environmental assets, from biological diversity to carbon storage in tropical forests. Table 3 shows estimates of the economic values residing in tropical forests. What stands out is the high value of tropical forests as stores of carbon, contingent on global warming being a proven phenomenon. Since the main proximate cause of deforestation is the burning of forests for conversion to agricultural or pasture land, one of the benefits of finding alternative land uses is the avoided global warming damage from released carbon dioxide through burning. These carbon values clearly exceed, sometimes by many multiples, the conversion value of the land. This suggests trades whereby the developed world seeks the attenuation of land-use options in tropical forests in return for encashable transfers. This is the essence of joint implementation, tradable development rights and franchise agreements (Panayotou, 1994). Early forms of such trades already exist in the form of joint implementation whereby carbon emissions in a developed economy are offset by carbon sequestration or carbon emission reduction in developing countries.

Table 3.Local and Global Conservation Values in Tropical Forests(In U.S. dollars per hectare, present values at 8 percent)
MexicoCosta Rica (Carbon values adjusted)Indonesia (Carbon values adjusted)MalaysiaPeninsular
Timber1,2401,000–2,0004,0751,024
Non-timber products77538–125325–1,23896–487
Carbon storage650–3,4003,0461,827–3,6541,015–2,7092,449
Pharmaceutical1–9021–103
Ecotourism/Recreation820913–35
Watershed protection<1
Option value80
Non-use value15
Note: Adapted from Kumari, 1994, but with additional material and some changed conversions. All values are present values at 8 percent discount rate, but carbon values are at 3 percent discount rate. Uniform damage estimates of $20.3 tC have been used (Fankhauser and Pearce, 1994), so that original carbon damage estimates in the World Bank studies have been reestimated.
Note: Adapted from Kumari, 1994, but with additional material and some changed conversions. All values are present values at 8 percent discount rate, but carbon values are at 3 percent discount rate. Uniform damage estimates of $20.3 tC have been used (Fankhauser and Pearce, 1994), so that original carbon damage estimates in the World Bank studies have been reestimated.

Asset Depletion and the Savings Rule

The final issue we address is the role that environmental assets and natural capital should play in development analysis. Given the centrality of savings and investment in economic theory, it is perhaps surprising that the effects of depleting natural resources and degrading the environment have not, until recently, been considered in the measurement of national savings. This omission may be explained both by the models economists use and the fact that the UN System of National Accounts (SNA) ignores depletion and degradation of the natural environment. This is not intended to be excessively critical of the SNA, which measures market activity very well, as is its intent. It is nonetheless true that the glasses we look through as economists tend to color or restrict our view of the problems we face.

Valuing depletion and degradation within a national accounting framework is an increasingly viable proposition, as a result both of the progress in the techniques of valuation of environmental resources (Freeman, 1994) and of the expanding foundation that theoretical developments are placing under the methods of green national accounting (Mäler, 1991; Hamilton, 1994a). The first application of these accounting methods to the measurement of net savings appeared in Pearce and Atkinson, 1993. This study combined published estimates of depletion and degradation for 20 countries with standard national accounting data to examine the true savings behavior (what Hamilton, 1994a, calls “genuine saving”). By this measure many countries appear to be unsustainable.

Enlarging the concept of net saving to include the depletion of natural resources is in many ways the most natural alteration of traditional savings concepts, because the depletion of a natural resource is, in effect, the liquidation of an asset and so should not appear in any measure of net national product or, by extension, net savings. While minor technical issues remain, the methods of valuing the depletion, discovery and growth of commercial natural resources in the context of the SNA are by now well developed (Hamilton, 1994a; Hill and Harrison, 1994).

More problematic is the valuation of environmental degradation. While UN guidelines for environmental accounting (United Nations, 1993) favor valuing this degradation in terms of maintenance costs (the cost of restoring the environment to its state at the beginning of the accounting period), the latest theoretical approaches (Hamilton and Atkinson, 1995) suggest that the marginal social costs of pollution are the correct basis for valuing waste emissions to the environment. The model supporting this conclusion is included as an annex to this paper.

To give the flavor of what results from the formal approach to green national accounting, the following expression adapts the expression for economic welfare from Hamilton, 1994a:

Here C is consumption, I investment, n the unit resource rental rate net of emission taxes on production, R resource extraction, g resource growth, σ the marginal social costs of pollution emissions e, the natural dissipation of pollution d, and PB the willingness of consumers to pay for environmental services B. For nonliving natural resources the term in g is zero, while d is zero for pollutants with cumulative effects. External trade is ignored in this expression for reasons of simplicity.

The measure of sustainable national income simply drops the last welfare term from this expression. The intuition behind this is clear: I − n (R − g) − σ(e − d) is the value of net investment when changes in natural resource stocks and stocks of pollutants, appropriately shadow priced, are included in addition to increments to the stock of produced assets.

While it is important to know what constitutes the sustainable level of national income, this measure is not particularly relevant for policy purposes. A shift in the level of national income does not carry a policy signal with regard to sustainable development, while the relative growth rates of sustainable income and GNP, for instance, are liable to give equivocal signals. Given that concerns about sustainable development are fundamentally concerns about the future, this suggests that adjusted measures of savings and wealth are more fertile territory for policy purposes.

The expression for genuine saving follows directly from the preceding:

Here GNP − C is gross saving as traditionally defined, with C being the sum of public and private consumption; gross saving includes the level of foreign saving as well. D is the value of depreciation of produced assets, while the last two terms represent the value of net depletion of natural resources and net accumulation of pollutants.

The importance of this measure of genuine saving is that it is a one-sided indicator of sustainability. While Asheim, 1994, and Pezzey, 1994, have shown that measuring positive genuine savings at a given point in time does not permit one to conclude that the economy is necessarily on a sustainable path, Hamilton and Atkinson, 1995, demonstrate that persistent negative genuine saving implies that the economy is not sustainable and that welfare will eventually decline. The foregoing references to “points in time” and “persistent negative savings” are important: sustainability is fundamentally a property of the path that the economy is on, rather than its instantaneous state.

It is possible to have apparently positive gross saving and negative genuine saving. So while it is easy to calculate gross savings rates from published national accounts data, this may give little indication of whether the economy is or is not on a sustainable path. This reinforces the point made earlier that economists may be biased in their conclusions with regard to economic performance.

Several policy issues are raised by negative genuine saving. All the policies concerning generating and mobilizing savings are relevant, including fiscal and monetary policy, levels of government current consumption, government investment policies (particularly with regard to the investment of resource rents), and the size and viability of the financial sector. Beyond these, there are questions of the performance of the resource sector, including tenurial arrangements and royalty regimes. Finally, there are questions regarding the policies for environmental protection and the optimality of pollution control levels.

Savings are not the only means to achieve economic sustainability. More efficient use of existing assets can lead to growth in many cases, although there is a limit, and the quality of investments arising from savings is also clearly important. Investment in primary education will give higher returns than many of the development projects seen historically in developing nations.

Evidence on Genuine Saving

Notions of genuine saving would have little impact if all countries were prudent managers of their portfolio of economic assets. There is abundant evidence, however, that many countries are not on a sustainable path. Figure 1 plots genuine savings on a regional basis for a range of developing countries.

Figure 1.Genuine Savings Rate by Region

Source: World Bank (1995).

The saving rates in Figure 1 account for the depletion of oil, major minerals, and net deforestation. The calculations of resource rents are crude, assumed to be equal to 50 percent of the market value of resource extraction and harvest, but sufficiently accurate (see Hamilton, 1994b) to serve as a useful indicator. Carbon emissions are the only pollutants considered in these calculations, with the global social costs of each ton of carbon valued at $20 in 1990 (Fankhauser, 1994). Depletion of fish stocks and degradation of soils, with the latter being particularly important in many developing countries, are not included in these figures.

The interregional comparison of savings behavior over this 30-year period turns up some interesting trends, although a number of caveats are required. Both South Asia and Latin America and the Caribbean had genuine saving rates that were moderately positive on average over this period, with savings briefly going negative at the time of the oil crisis in South Asia and at the time of the debt crisis (roughly speaking) in Latin America and the Caribbean. The short time series for the Middle East and North Africa shows a substantial decline in genuine saving, but the data coverage for these figures should encourage caution in their interpretation. In addition, the net price method of calculating rents may not be particularly appropriate for countries having many decades of reserves of oil (i.e., a discount factor should be factored into the depletion calculation, as in El Serafy, 1989).

The two trends that stand out in Figure 1 are for East Asia, where the rise of the “tigers” was associated with very strong saving performance and where primary resource activities are a declining proportion of GNP, and sub-Saharan Africa,1 which started dissaving roughly at the time of the oil crisis and has continued on this unsustainable path into the 1990s.

Figure 2 shows the results of a more refined calculation of genuine savings (based on Hamilton, 1994b) for a sample country in sub-Saharan Africa, Malawi. One striking aspect of this figure is that traditional net saving in Malawi was negative in 1981 and 1984–87. By standard national accounting measures, therefore, Malawi was marginally sustainable during the 1980s. When the extra 5–7 percent of GNP for resource depletion and CO2 emissions is included, Malawi was clearly unsustainable during this decade, and this excludes the value of soil erosion in this highly agricultural country.

Figure 2.Malawi: Genuine Savings Rate, 1981–90

Share of GNP

Source: data from Hamilton 1994b.

There is ample evidence therefore that many countries have been on an unsustainable development path, at least since the mid-1970s. Sub-Saharan Africa is noteworthy in this regard, and it is to this region that we now turn our attention.

Genuine Savings and Sub-Saharan African Experience

In a paper reviewing the literature on long-term development and growth in sub-Saharan Africa, Ndulu and Elbadawi, 1994, cite a number of broad conclusions: (1) sub-Saharan Africa has grown more slowly than other developing countries since the mid-1970s; (2) lower saving rates and levels of human capital have prevented it from catching up with other developing countries; (3) the policy climate in sub-Saharan Africa has not been conducive to sustained growth, characterized as it has been by disincentives to save, overvalued and variable exchange rates, high public consumption, and underdeveloped financial systems; and (4) the economies of sub-Saharan Africa have been subject to many external shocks, both economic and physical (in the form of drought and other severe weather patterns), and political instability. Analysis of genuine savings provides further insights.

Table 4 presents the rates of genuine saving for selected countries in sub-Saharan Africa in the mid- to late 1980s. The rather eclectic selection of countries reflects limitations in data availability.

Table 4.Genuine Saving Rates in Sub-Saharan Africa, 1986–90
19861987198819891990
Burundi1−23−15
Côte d’Ivoire−6−10−14−13−17
Cameroon−1−30−4−9
Congo−40−69−36−51−42
Ghana−7−4−3−2−2
Kenya77463
Mali−13−5112
Mauritania−12−18−10−3−8
Malawi−10−16−11−5−6
Niger15−56−4
Nigeria−15−24−31−38−41
Rwanda98554
Senegal−14−7−30−2
Sierra Leone−2−7−11−8−9
Chad−26−22−17−10−17
Uganda−12−7−6−15−15
South Africa5711109
Zaïre−18−21−28−30−32
Zimbabwe244106

The pattern presented in Table 4 appears to be “the curse of the mineral-rich” (Gelb, 1988). Kenya, Rwanda, Burundi, and Niger, with relatively little exports of oil and minerals, have the most promising saving performance. On the other hand, mineral-rich South Africa and Zimbabwe both exhibited positive genuine savings. As noted earlier, the figures for Nigeria are probably skewed owing to the substantial size of the deposits of crude oil. Zaïre is the other anomalously large dissaver. While these effects may have been overstated for technical reasons, having to do with the valuation of resource depletion, it is also true that the economic policy climate has been particularly unfavorable in Zaïre for many years.

Conclusions

We have provided a brief overview of the ways in which environmental degradation affects the macroeconomy in developing countries and, for that matter, in the industrial world, since the principles do not change. Our basic conclusion is that environmental degradation gives rise to major economic costs in developing countries in terms of

  • impairment of the human capital stock via premature mortality and morbidity
  • loss of marketed GNP through health effects and degradation of assets such as soil and forests
  • loss of nonmarketed GNP that could be the subject of capture via appropriately designed policies, domestic and international
  • reductions in genuine savings, which amount to the mining of the capital base on which many developing countries depend.

The implications for policy are

  • the need for revised accounting procedures, now well under way
  • revised macroeconomic indicators that incorporate sustainability considerations—we advocate the genuine savings concept as the first step in such a revision
  • the need for more and better economic valuation of nonmarketed resources—also now well under way
  • a wholesale rethinking of mainstream development economics at the pedagogic level to reflect the rethinking that has already taken place at the research level.
Annex
Formal Green Accounting Model

We assume a simple closed economy with a single resource producing a composite good that may be consumed, invested, or used to abate pollution, so that F(K,R) = C + K˙ + a, where R is resource use and a is pollution abatement expenditures. Pollution emissions are a function of production and abatement, e = e(F,a), and pollutants accumulate in a stock M such that M˙= e - d(M) where d is the quantity of natural dissipation of the pollution stock. The flow of environmental services B is negatively related to the size of the pollution stock, so that B˙ = −αM˙. Resource stocks S grow by an amount g and are depleted by extraction R, so that S˙ = −R + g(S), and resources are assumed to be costless to produce. The utility of consumers is assumed to be a function of consumption and environmental services, U = U(C,B).

If we assume that there is a social planner who wishes to maximize the present value of utility over an infinite time horizon, for some fixed pure rate of time preference r, then there is a close relationship between the Hamiltonian function for the optimal control problem and the measure of current welfare for the society. The Hamiltonian is given by

where γK is the shadow price of capital, in utils, γB the shadow price of environmental services, and γS is the shadow price of the resource. The first order conditions for maximizing the Hamiltonian, setting the partial derivatives with respect to the control variables C, a, and R to 0, yield the following:

Here b is the marginal cost of abating a unit of pollution, while FR is the unit resource rental rate, and beF is the effective emissions tax rate on production. The measure of economic welfare is derived by valuing each flow at its shadow price in utils, then converting to consumption units by dividing through by UC, yielding

It can be shown, for example, in Hamilton and Atkinson, 1995, that the marginal cost of abatement is equal to the marginal social cost of pollution emissions, which is in turn equal to the level of the optimal Pigouvian tax required to maximize welfare. Note that pB ≡ UB/UC is the price that a utility-maximizing consumer would be willing to pay for a unit of environmental service, and that resources are priced at their marginal product less the value of the emissions tax on production. Therefore this expression for welfare, although derived from optimal control, corresponds to what would be attained in a competitive equilibrium with a Pigouvian pollution tax.

This expression generalizes in obvious ways. Dropping the last welfare term yields sustainable income, since persistently consuming more than this implies negative genuine savings. A purely cumulative pollutant is handled by dropping the term in dissipation d. For a nonliving resource the term in growth g would be dropped.

Reference

    AndersonDenis1987The Economics of Afforestation: A Case Study in Africa (Baltimore: Johns Hopkins University).

    AndersonDenis1989“Economic Aspects of Afforestation and Soil Conservation Projects,”in Environmental Management and Economic Developmented. by GunterSchramm andJeremy J.Warford (Baltimore: Johns Hopkins University).

    • Search Google Scholar
    • Export Citation

    AsheimGeir B.1994“Net National Product as an Indicator of Sustainability,”Scandinavian Journal of EconomicsVol. 96No. 2 pp. 25765.

    • Search Google Scholar
    • Export Citation

    BalasubramanianV.N. andS.Lall1991Current Issues in Development Economics (London: Macmillan).

    BartelmusPeter1986Environment and Development (London: Allen and Unwin).

    El SerafySalah1989“The Proper Calculation of Income from Depletable Natural Resources,”in Environmental Accounting for Sustainable Developmented. by Yusuf J.AhmadSalah ElSerafy andErnstLutz (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    EröcalDenizhaned.1991Environmental Management in Developing Countries (Paris: OECD).

    EsreyS. A.1990“Health Benefits from Improvements in Water Supply and Sanitation: Survey and Analysis of the Literature on Selected Diseases,”Unpublished report to U.S. Agency for International DevelopmentWashington, D.C.

    • Search Google Scholar
    • Export Citation

    FankhauserSamuel1994“Evaluating the Social Costs of Greenhouse Gas Emissions,”The Energy JournalVol. 15No. 2 pp. 15784.

    • Search Google Scholar
    • Export Citation

    FankhauserSamuel andDavid W.Pearce1994“The Social Costs of Greenhouse Gas Emissions,”in The Economics of Climate Change (Paris: OECD).

    • Search Google Scholar
    • Export Citation

    FreemanA. Myrick1994The Measurement of Environmental and Resource Values: Theory and Methods (Washington: Resources for the Future).

    • Search Google Scholar
    • Export Citation

    GelbAlan H.1988Oil Windfalls: Blessing or Curse? (New York: Oxford University).

    HamiltonKirk1994a“Green Adjustments to GDP,”Resources PolicyVol. 20No. 3 pp. 15568.

    HamiltonKirk1994b“Estimated Rental Rates for Minerals and Crude Oil,”Unpublished report Environment Department (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    HamiltonKirk andGilesAtkinson1995“Valuing Air Pollution in the National Accounts,”Unpublished paper presented to the London Group on National Accounts and the EnvironmentWashingtonMarch15–171995.

    • Search Google Scholar
    • Export Citation

    HartwickJohn M.1990“Natural Resources, National Accounting and Economic Depreciation,”Journal of Public EconomicsVol. 43 pp. 291304.

    • Search Google Scholar
    • Export Citation

    HillP. andA.Harrison1994“Accounting for Subsoil Assets in the 1993 SNA,”Unpublished paper presented to the London Group on National Accounts and the EnvironmentLondonMarch15–181994.

    • Search Google Scholar
    • Export Citation

    HogendornJan S.1992Economic Development2nd Edition (New York: Harper Collins).

    KumariK.1994“An Environmental and Economic Assessment of Forestry Management Options: A Case Study in Peninsular Malaysia,”in Sustainable Forest Management in Malaysia Unpublished Ph.D. Thesis K. Kumari University of East Anglia.

    • Search Google Scholar
    • Export Citation

    MälerKarl-Goran1991“National Accounts and Environmental Resources,”Environmental and Resource EconomicsVol. 1 pp. 115.

    NduluB. andI. ElbadawiI.1994“Long-Term Development and Sustainable Growth in Sub-Saharan Africa,”Unpublished paper presented to the Colloquium on New Directions in Development Economics—Growth Equity and Sustainable Development SARECStockholmMarch9–111994.

    • Search Google Scholar
    • Export Citation

    PanayotouT.1994“Financing Mechanisms for Environmental Investments and Sustainable Development,”Unpublished paper Harvard Institute for International Development (Cambridge, Massachusetts: Harvard University).

    • Search Google Scholar
    • Export Citation

    PearceDavid W.1993Economic Values and the Natural World (London: Earth-scan).

    PearceDavid W.(forthcoming)“North South Transfers and Global Environmental Change,”in Global Environmental Change.

    PearceDavid W. andGilesAtkinson1993“Capital Theory and the Measurement of Sustainable Development: An Indicator of Weak Sustainability,”Ecological EconomicsVol. 8 pp. 1038.

    • Search Google Scholar
    • Export Citation

    PearceDavid W. andJeremy J.Warford1993World Without End: Economics, Environment and Sustainable Development (Oxford: Oxford University Press).

    • Search Google Scholar
    • Export Citation

    PearceDavid W.D.Whittington andS.Georgiou1994Project and Policy Appraisal: Integrating Economics and Environment (Paris: OECD).

    PezzeyJohn1994“The Optimal Sustainable Depletion of Non-Renewable Resources,”Unpublished paper University College London.

    RanisGustav andT.P.Schultz1988The State of Development Economics: Progress and Perspectives (Oxford: Blackwell).

    United Nations1993Integrated Environmental and Economic Accounting.Series F No. 61 (New York: United Nations).

    WalshJ. andK.Warren1979Selective Primary Health Care: An Interim Strategy for Disease Control in Developing CountriesNew England Journal of MedicineVol. 301No. 18 pp. 96774.

    • Search Google Scholar
    • Export Citation

    WeitzmanMartin L.1976“On the Welfare Significance of National Product in a Dynamic Economy,”Quarterly Journal of EconomicsVol. 90.

    • Search Google Scholar
    • Export Citation

    World Bank1995Monitoring Environmental ProgressEnvironment Department (Washington: World Bank).

Discussion

Discussant’s Comments

Kirit Parikh

Professor David Pearce has made very many valuable contributions in bringing environmental concerns into economics, and this paper is one of these contributions. I agree with the broad conclusion of the Pearce and Hamilton paper, that environmental degradation impairs human capital in developing countries, a conclusion that is usually recognized, but not adequately paid attention to. I have, however, some difficulty with the policy conclusions to be drawn, and particularly about the notion of genuine savings, on which I will elaborate.

Regarding the impact of air and water pollution, the authors have summarized estimates of mortality and morbidity costs owing to different air and water pollutants, and they rightly observe that comparison of absolute numbers may give rise to the view that one industrial country resident is worth less than one developed country resident. Unfortunately, they do not present relative numbers, but give absolute numbers, partly because for most cities estimates of gross products are not available. But I would have thought that one could have guestimated the gross products for different cities, with perhaps no more error than the error inherent in the costs of mortality and morbidity. That would have been less distorting and more clarifying of what the point is that they are trying to make.

I also have some difficulty with their policy conclusion that the focus should be on the transport sector rather than, say, on the traditional power sector. While the conclusion may be right, it does not necessarily follow from their argument. One should take into account the costs of abatement, and not just the benefits. For example, every dollar spent in the power sector may lead to a larger reduction in SO2 or particulate matter or to the reduction in cost of mortality and morbidity than a dollar spent in effectively controlling pollutants from the transport sector. One does need to worry about the costs as well as the benefits on this side.

Regarding water pollution, the authors assert that cost-benefit analyses typically favor investment in water quality treatment. Water quality and water treatment have been neglected in many developing countries and need much greater attention from governments and from international agencies (perhaps to finance it).

Turning to the nonmarket returns and joint implementation, the studies by Anderson on the value of forestation and Pearce’s studies certainly point out the difficulty of capturing nonmarket returns from environmental investment. They suggest joint implementation, whereby carbon emissions in industrial economy are offset by carbon sequestration emission reduction in developing countries. Claiming offsets by carbon emissions in joint implementation projects raises a number of issues. Let me quote from a paper (“Joint Implementation and North-South Cooperation for Climate Change,” International Environmental Affairs, Vol. 7, No. 1, 1995, p. 32) by Ms. J. Parikh:

Let us take the example of agro-forestry projects in Guatemala to offset carbon releases in the United States. A U.S. power plant proposed a plan in which 40,000 farmers in Guatemala work for 10 years for a joint implementation project for the sum of 3.8 million dollars. The lifetime of a power plant is 40 years, and so the plantation will have to exist for 40 years. In addition, the Government of Guatemala is required to pay 1.2 million dollars. Apparently, this total of 5 million dollars is to pay for saplings, land, other inputs, and continued care for 52 million trees for 10 years. This works out to a mere 10 cents per tree. Is it possible to plant and nurture a tree for 10 cents? The project seems to expect that Guatemala will be perpetually living in poverty. Does it tie 40,000 farmers to this project for 40 years? What does Guatemala get for this? What happens to the trees after 40 years? And this is an important set of issues: If they are burned, would the carbon dioxide belong to the United States? If they are not burned, would the methane generated from the decay of leaves and roots after the trees are cut down belong to the United States? What if Guatemala needs this land for something else during these 40 years? Will it tie down generations for a paltry loan by the rich? What if the new generation wants a shopping complex or any other use for remunerative activity on this area? If the U.S. company receives tax breaks for this purpose, then it did not go through any pains for this project at all. Rather, it is the Guatemalans who have paid for this project. The purpose of presenting this case is not to discourage private enterprise from taking initiatives such as this, instead it highlights the need for institutional structures in joint implementation to channel such enthusiasm in the right direction.

One appreciates the opportunity to trade, but this opportunity needs the appropriate kind of institutional mechanism and setting to make it fair and just.

The global climate change problem is precipitated by the industrial countries, and poor developing countries cannot be justly expected to worry about carbon emissions unless a fair allocation of property rights in the global carbon sink is made. This may seem somewhat peripheral to the Pearce and Hamilton paper, but it is not, and I will show this later.

Let me turn to the notion of genuine savings and sustainability. The authors have made a heroic attempt to find one simple indicator of sustainability, so it is worth looking at in some detail. Pearce and Hamilton advanced the notion of savings, which they call genuine savings, as an indicator of sustainability, and referred to the weak sustainability rule advanced by Pearce and Atkinson in their 1993 paper. In that paper they advanced the indicator and suggested that the United States, among others, was a sustainable economy. Any test that classifies the United States as a sustainable society must be suspect. Their test for sustainability suggests that if GNP minus consumption, minus depreciation of produced assets, minus depreciation of resources, including quality of environment owing to pollutants, is positive, the economy is sustainable, but if it is negative, then it is not. I find a number of conceptual difficulties and empirical pitfalls in applying this test.

The most important difficulty is that the damage to environment done by a country is not limited to the country itself. For example, the damage done by the United States is not limited to U.S. territory. Thus, the United States has very high shares in the degradation to global commons. Their contributions to global warming, to the creation and enlargement of the hole in the ozone layer, degradation of the seas owing to oil spills, water pollution, nuclear testing and nuclear and other hazardous waste disposals in the seas, resulting from the U.S. lifestyle and policies, are substantial and very large in per capita terms. Should the United States be held accountable for these damages, I would be surprised if the U.S. savings would be sufficient to restore these damages to global commons.

A second difficulty is that the measure does not involve depreciation of natural capital embodied in trade. Some authors have estimated that U.S. imports are more pollution- or natural-resource-intensive than their exports. The loss to other countries’ sustainability caused by their exports to the U.S. economy are not accounted for by the Pearce and Hamilton measure. If the United States were to satisfy its lifestyle entirely through domestic production, its savings may not be adequate to restore or substitute for degraded natural capital.

The third difficulty is that savings (or man-made capital) cannot substitute or make good natural capital. Thus, U.S. savings cannot make good the ozone layer, or Brazilian savings cannot restore virgin rain forests.

The fourth difficulty is with valuation of capital. We require the same capital stock to provide future generations with the ability to produce the same goods and services as are produced today. Even assuming that man-made capital can substitute for natural capital in the production of goods and services, such substitution need not be one dollar’s worth of man-made capital for one dollar’s worth of natural capital. This would be obvious if one looks at a production function with substitutability, such as a Cobb-Douglas function. It is possible to argue, however, that the valuation of capital and natural capital is done in a way that reflects appropriately the marginal productivity of each type of capital. Such an estimation would be complex, since it would have to account for the nonlinearities involved and would have to assume that prices are not distorted.

Finally, there is the fifth difficulty with the genuine savings estimates. Pearce and Hamilton have emphasized rightly the importance of human capital. Governments spend money on education and health services. These are in a real sense investments in human capital formation, but in an accounting sense are considered as current consumption. The same is true of private expenditure on education, which too is investment. I am afraid that the authors need to work harder to make their estimates more accurate.

One realizes that the strong sustainability rule, which is the appropriate measure, may be difficult to verify, but misleading calculations based on weak sustainability rule are worse than useless and would only lead to greater smugness in the industrial countries, whose lifestyles are thus wrongly shown to be sustainable. They would be even more reluctant to change their lifestyles.

The first four difficulties show that a positive genuine saving does not guarantee sustainability. The authors claim that their measure of genuine savings is a single indicator of sustainability. But because of the fourth and the fifth objections raised above and the way they have used it, I would argue that it is not even that. To wit, their calculations of negative genuine savings for sub-Saharan Africa cannot be taken to mean that countries with negative “genuine savings” that are depreciated by carbon emissions are on an unsustainable path. The value of carbon emissions should not be subtracted as long as there is no obligation on the part of a country to reduce its carbon emissions, or as long as its emissions are below its fair share in the global commons’ absorptive capacity. To the extent that genuine savings have become negative because of subtraction of carbon emission values, sub-Saharan Africa cannot be classified as unsustainable.

In conclusion, let me emphasize that my comments are not just of academic interest. Unfortunately, joint implementation projects of dubious benefit and ambiguous offsets, together with erroneous interpretation of estimates of genuine savings, which lack conceptual clarity, seem to shift the burden of climate change adjustment on developing countries, which have not been responsible for precipitating it in the first place.

Response

David Pearce: Let me respond to Kirit Parikh. Kirit knows we have addressed the various issues that he has raised in papers that we have done over the years. I agree we should have presented the health impacts of urban pollution as a percentage of city-level incomes; however, we had no estimates of city incomes. But we have done this estimation nationally, and the proposition is valid that the health impacts as a percentage of incomes are high and are, in fact, larger in developing countries than they are in rich countries.

Second, we agree on the need to account for the cost of abatement of environmental damage owing to transport. In our opinion, many of these interventions are known to be extremely cheap, although the transport sector produces an incredible array of pollutants, many of which are extremely damaging to health. We are aware that joint implementation is controversial, but one has to recognize that the joint implementation deals are willingly entered into by two parties. We cannot be arguing that poor downtrodden peasants are being forced into these deals by some wicked multinational company or some wicked industrial-country government. Joint implementation is, in fact, one of a whole array of new financing mechanisms. If we do not exploit these mechanisms, you will not see the needed financing for environmental protection.

Kirit makes reference to our earlier paper, where the genuine savings concept was introduced and the United States came out to be sustainable, though barely. That was the whole point of the exercise. Two countries came out to be a big surprise in that paper. One was the United States, which was barely sustainable on the genuine savings rule, and the other was the United Kingdom, which was not sustainable. And if you ask why, the answer is quite simple. It fits very neatly with the popular view of the United States as a high-consumption economy. It also fits with the view in the United Kingdom that we have wasted and exploited our North Sea oil resources and have consumed them. This fact fits rather well with the nice consumption boom that the United Kingdom had in the 1980s.

Kirit says we should account for global damage. We do and that is what the CO2 is doing in the paper. There is no great difficulty in putting the ozone depletion in this write-up.

We have written extensively about the issue of substitutability between natural resources. There are several problems. One is that ecologists and ecological economists have still not proven to the world in general that a large degree of nonsubstitutability actually exists.

I am ambivalent on the education issue. I know that some people build education back into the picture and use public expenditures on education as a surrogate for positive investment in human capital.

The issue is whether the mining of resources is adequately accounted for in economic policy formulation. If the IMF advises a country to adopt a macroeconomic adjustment policy that forces a country to mine its natural resources, in what sense has sustainable adjustment taken place? The answer has to be that it has not. That is all that we are stressing. The purpose of the genuine saving concept is to try to highlight those issues.

Some people are against single indicators. Just because we offer one measure does not mean that it is the only possible measure. I am quite happy to see a whole array of sustainability indicators. Ours is only one useful way of looking at those issues.

On the IMF again, at the end of the day, if we say that the IMF is not interested in how countries adjust, then something has gone seriously wrong, because if the adjustment takes place through the overmining of natural resources, it simply cannot be sustainable.

Other Discussion

Stein Hansen: If one is just looking at the fundamental question of whether economic growth is good or bad, then one has to look at the direct as well as indirect effects. I wonder if these are represented in the Pearce-Hamilton equation. Output enters with a positive sign in their genuine saving equation, but there are the indirect effects as well from an increase in output to the other variables, particularly the depletion of natural resources and the increase in pollution. So my question to David Pearce is what, in his view, is the net effect of economic growth on the genuine savings rate? Do the negative indirect effects of GNP growth, operating on depletion and on pollution (direct and indirect), outweigh the obvious positive direct effects of GNP growth that we macroeconomists have to deal with most of the time?

Malcolm Knight: For me, the real question is what should institutions like the IMF do and what criteria should we use in looking at the macroeconomic elements of stabilization programs? We have had two very interesting approaches today—the one of Pearce and Hamilton and the other of Stein Hansen—and both are actually quite different in their intellectual underpinnings.

Stein Hansen’s approach is to have an environmental group establish the environmental priorities that he then took as given, to look at their economic impacts, and to experiment with the set of economic policies that would allow achievement of those goals along with their implications on the level and structure of consumption. That is a sort of war strategy—you have something you absolutely have to achieve and you run the macroeconomy subject to that constraint.

The Pearce and Hamilton approach appears to be more incremental, looking at the effect of macroeconomic policies on some index (and there has been a lot of discussion about whether the genuine savings measure is the proper index or not), but it is an index of incremental modifications that prompt changes in macroeconomic policies aimed at changing the level of economic growth or consumption. Now these seem to be very different approaches and perhaps the authors can comment on the relative merits of the two approaches.

Stein Hansen: As a response to this particular question, the dominant purpose of what we set out to do was to bring the NGOs and the Ministry of Finance (and the government) together and help establish mutual trust between these parties. As I said this morning, we did not just do computable general equilibrium work, we also did a lot of other analytic and fairly basic research work in parallel. Some of this work has addressed fundamental sustainability issues and, drawing on works that David Pearce, and others, and we ourselves had done, we come to the not-all-that-surprising conclusion that there were no simple sustainability rules. For that reason we decided to keep all possible indicators in mind and put them into practice within the operational framework already used routinely by the government, and also to help, along the way, bridge gaps of misunderstanding between the various camps.

Melting a number of things into one single indicator, whether it is green or brown or whatever color, helps alert people and create awareness about issues. But once you have done that, I think, you should move on to the more multidimensional picture and put it into operation, which, of course, requires data for the variables you have chosen.

David Pearce: I certainly would prefer to get an economic handle on resource-depreciation and on environmental damage. I know that these things are controversial, but they are now pretty standard fare in the World Bank, in national policies, and certainly in the European Community. So, my preference would be to begin that way and to try to establish environmental priorities. One of the exercises we are doing at the moment is in fact for the European Commission—trying to identify the environmental priorities of Europe itself. I am more than aware, after 30 years in this business, that you do not come up with a fully quantified risk-assessment ranking of all those goals. Those exercises have been done here in the United States by the EPA. They are fascinating, but also show the pitfalls of trying to come up with a ranking. But my argument is that this process is extremely important, even if at the end of the day you might not believe the actual numbers that emerge. We find in our European work that the process leads you to focus not so much on environmental issues but on selected economic sectors, such as transport, the policies of which can have a lot of environmental payoff.

I do not know how IMF adjustment policies are formulated. Perhaps the sectoral focus is not suited to the operational work of the IMF, but you must look at how adjustment takes place. One wants to know how economies adjust. If, for example, you are adjusting by running out of fuelwood and you have to import kerosene, then you are obviously affecting the balance of payments. As I understand it, this is of fundamental relevance to the IMF. If you react to economic difficulties, say, by mining your tourist sector, as some Caribbean Islands are doing, you may be creating a perfectly unsustainable situation as regards future growth and prosperity. The loss of the coral reefs and the disappearance of ground-water supplies are illustrations of that. So it seems to me that how people adjust is extremely important. And it is therefore important to quantify, as far as possible and as far as credible, the way in which countries adjust.

So, it does not matter whether it is genuine savings, green GNP, or physical indicators, so long as you start looking at the sustainability issue. The fact is that you have to do something and at the moment there is no evidence that the Fund is doing much by way of examining how those adjustments are taking place.

Mohan Munasinghe: We in the World bank have started asking the question: How do we engage those working on the environmental side and on the economic management side in the developing world to work together? A helpful tactic we have used is to get the environmental group to identify, say, the five most important environmental issues, not a hundred, but the five most important, and then ask them the question: Which of the current set of economic policies affect the selected environmental issues the most? At the same time you go to the Ministry of Finance or the Central Bank and you put the reverse question: Of the reform program you have in mind, which policies are likely to affect some aspect of the environment? And then in the next step you put these groups together and develop a composite, which we have called an Action Impact Matrix. This matrix has proven to be an extremely useful exercise for us in developing a consensus and also establishing priorities. So, in my view, you have to work from both sides and you have to ensure that the country gets the ownership of final decisions in order to make headway.

Salah El Serafy: Much more important than pursuing the impact of structural adjustment on the environment is actually to do the ground work of finding out what is harming the environment so as to rebut this with good microeconomic and macroeconomic policy.

Kirit Parikh: I completely agree with David Pearce that you use green accounting or natural resource accounting essentially to focus attention on the damage that we are doing to the environment. My only objection, however, is regarding the use of a single concept as an indicator of sustainability, especially when you use it in comparing countries, as the World Bank and the Fund often have to.

Benedicte Christensen: I want to come back to the issue that Malcolm Knight raised, namely the implications for the Fund. I do not want to leave the impression that the Fund is not concerned with environmental matters. I very much share the view that the key word is sustainability. What we are concerned with in the Fund are not environmental objectives per se, but the sustainability of economic growth and of external balance. And I think both of these involve important environmental concerns and issues. Even though we do not measure them by genuine savings rate or by other concepts, we do measure them indirectly.

When we have medium-term balance of payments projections for countries that rely on mining or natural resource exports, we are very much concerned with excessive depletion of such resources. Similarly, when we have countries relying on logging exports, we take them into account, both in our growth projections and export projections and discuss to some extent even the regulatory framework, especially, if it has an important bearing on the macroeconomic situation. I think the key to our work on the environment is to ask if the lack of an environmental policy has an important bearing on the macroeconomic outcome.

David Pearce: I think this is an extremely important comment. Now let us go a bit further and start looking at measures of soil erosion as well, because that is also extremely important for the longer-term viability of exports. There is a whole range of natural resources and their depletion and depreciation become relevant to macroeconomic exercise.

Vito Tanzi: As regards sustainability, it is unclear to me what we need to focus on—sustainability over what period? Over 2 years, 5 years, 100 years, 500 years, what period? It is easy to conceive of a situation in which something is not sustainable for 100 years, but might be acceptable for two or three years and helps you bring about other changes in the economy. So there might be a situation in which you might want to accelerate the exploitation of natural resources for a while, which can help you bring about other adjustments in the economy, after which you go back to a more sustainable path. Unless we define the time horizon, I am not sure what sustainability really means.

One more comment. As I was listening to the discussion, I was reminded of the discussion that took place in our Executive Board on the issue of the environment. One Executive Director noted that today you come with the environment, tomorrow you might come with the right of women, the next week with the aged, and still the next week with children. In other words, where do we draw the line? Why should the Fund get involved in the environment and not in the rights of women, or the rights of children, or something as important. If we say, because the environment has something to do with sustainability, that leads me to my first question: What does sustainability mean and over what length of time?

David Nellor: To follow up on what Benedicte Christensen said, in the cases where the Fund has dealt with resource-based economies, the staff have looked for advice from the World Bank and others on what sustainability means. However, we still require a quantitative target initially in order to be able to design the macroeconomic program.

The other issue is the time frame over which adjustment takes place. If you attempt to achieve sustainability over a very short time frame, the dislocation caused may in some cases just be as damaging as allowing the continuation of current practices. Maintaining a balance in the rate of adjustment and the rate of environmental protection is needed, which we have not really focused on in the discussion thus far.

Note: The authors are respectively Director and Senior Fellow, Centre for Social and Economic Research on the Global Environment, University College London and University of East Anglia, United Kingdom.
1The African figures are influenced by the performance of the oil states, especially Nigeria and Gabon. The earlier caveat about discount factors applies here as well.
Note: The discussion was chaired and moderated by Malcolm Knight.

    Other Resources Citing This Publication