John C. O'Connor
Most definitions of sustainability imply that per capita wealth, broadly defined, should not decline. Considering how much of economics is about the creation and maintenance of wealth, it is surprising how little work has been done to estimate it. Instead, conventional economics presumes a near-perfect world in which it is enough to measure annual flows of income, production, and spending. In such a world, savers and investors always make the best choices. Wealth estimates are needed 'only' if one wonders how close they really come to that ideal.
Comprehensive estimates of the wealth of nations were first reported in 1995 in the World Bank's Monitoring Environmental Progress (MEP)1 (see Box 2K). The results were, by necessity, approximations based on available data and are being modified by the World Bank and others2 While it will take years to perfect comprehensive wealth estimates, doing so is important if growth strategies of countries are to look beyond money income. The UN System of National Accounts (SNA) has taken an important step in this direction by letting national balance sheets include natural capital as well as produced assets (see Box 3G). MEP went further, broadening the balance sheet to include human resources (investing in individuals via education, health, etc.) and social infrastructure (the value of associational processes that help people work together).
Despite its limitations, MEP reached conclusions of significance for decision- making. Produced assets, seen as the main determinant of wealth in much applied economics, account for barely a fifth of wealth in most countries, rich or poor. Some countries are well-endowed with natural capital; others are not. The bulk of wealth is in human resources and social capital which MEP commingled in a residual for practical reasons.
Wealth analysis expects that the mix of assets can change over time, although critical boundaries must be respected for (and within) each category, separately. The mix is affected by flows of income, production, and spending. However, it is also affected by changing values of various assets which can be important over generations, the time horizon for sustainable development studies.
MEP emphasized that the usual way to create wealth is the flow of genuine saving, measured as a residual of production or income less consumption, depreciation of produced (human-made) assets, and drawing down of natural resources. Underlying data work3 showed that a consistent set of 'green' net national income time series would parallel the established measure (Gross National Product, GNP) except in a handful of countries. Many countries showed marked differences in trends between orthodox saving (income less consumption) and genuine saving.
Simply put, the orthodox measure exaggerates the saving effort of many nations and can mislead decision-makers into thinking wealth is created when one asset (say standing timber) is exchanged for another (say a machine).
The second edition of MEP, in draft at this writing, refines genuine saving measures in line with the underlying concept of wealth as a sum of four kinds of capital (produced, natural, human, and social) that are sometimes substitutes and sometimes complements. It treats spending on education as investment rather than consumption and raises genuine saving by that amount. It concludes, 'Large investments in education by the most economically successful countries leads them to exceed the genuine saving rates of their counterparts in other income groups by around eight per cent.'
Considerable progress has been made, in recent decades, in recognizing the importance of human capital formation. Empirical work has focused on 'non- monetary' measures like educational attainment (the average years of schooling for individuals in each country) which are then correlated with monetary measures like GNP. MEP showed that its (residual) balance sheet measure of human resources correlates at least as well with such non-monetary measures, and suggests new directions for further study.
For example, educational attainment is the same in Turkey, Honduras, and Ghana (4.5 years of schooling) yet details behind MEP put Turkey's wealth in human resources as $24,000 per capita compared to $4,000-5,000 in comparable countries. Larger differences emerge if one compares war-torn to stable countries. Something beyond education must be affecting the possibilities people have to take advantage of their productive capacity. Data problems aside, disparities between MEP's monetary and non-monetary measures of human resources may reflect differing contributions from social infrastructure. The World Bank, with Danish support, is embarking on a major effort to define, measure, and monitor this statistically fuzzy but important concept.4
A subtle but profound methodological change in the second edition of MEP shows that, in international comparisons over time, money prices are also fuzzy concepts. Both editions measure total wealth as a 'capitalized' stream of income, using a 'green' variant of GNP as the income measure. By deducting independent estimates for produced assets and natural capital, both produce a residual that commingles human resources and social infrastructure. However, the second edition considers international price differences an aberration that can be corrected by comparing GNPs, and thus total wealth, on the basis of purchasing power parities (PPPs) rather than the so-called Atlas method used in measuring per capita income guidelines for financial preferences, etc.
This methodological change has little effect on estimates for advanced countries since PPPs effectively presume that prices in developing countries would be the same except for imperfections, such as being less developed. Thus, per capita wealth in the United States is unchanged (about $400,000 per capita) but that the per capita wealth of Turkey rises (from $34,000 to $78,000) if 'non-tradables' like human resources are priced as they are in the US. This narrows the numbers gap between rich and poor, much as expressing temperature on a Celsius scale narrows it relative to Fahrenheit. It also suggests how the national balance sheet would look if, say, teachers were paid the same in Turkey and the US. If development means emulating today's high income countries, the difference provides a hint of which assets will rise most in relative price in a sustainable future.
MEP analysis of natural capital recognized that valuation is a very complicated exercise even when physical quantities are reasonably clear, like the amount of land and what is on and under it. Most economists would agree that the value of natural capital goes beyond direct marketed benefits to humans, like crops and ores. One solution is to 'monetize' indirect benefits like watershed protection for woodlands; assign 'options' values for as yet undiscovered uses of nature; and impute some- thing extra for its 'intrinsic' worth. However, there is no consensus about which approaches make sense in which contexts; economic theory provides no guidance on such matters.
Rather than advocating specific techniques, MEP compared monetary and non- monetary 'weights' much as was done for human resources. Physical quantities were added first by a crude set of market prices and then in a Natural Capital Indicator (NCI) developed by the World Resources Institute for the Global Environment Facility.5 This showed how the monetary value of natural capital depends on the value of land, which varies between high income and developing countries in accordance with vast disparities in their per capita incomes and imperfect international markets for land; while non-monetary valuation schemes, like NCI, reflect the far greater importance of developing countries as the home for biodiversity.
The main premise of wealth estimates is that sustainability means giving future generations as many opportunities as, if not more than, we have enjoyed. In accounting terms, that means leaving them as much or more wealth per capita. This is partly about the quantities of various assets willed to future generations. It is also, and perhaps more importantly, about the values they will assign to assets which may differ from current money prices.
Beyond inflation (increases in the general level of money prices), relative prices change with technological innovation, rising scarcity, level of development, and evolving human preferences (future generations may place greater value on a natural environment). Since MEP only reports on a single year (1990) and relies heavily on prevailing market prices, it provides only a guidepost, and a system for evaluating alternative views of a sustainable future. In this sense, more precise benchmarks may be less important than devising wealth time series and methodologies capable of looking beyond today's imperfect markets.
Economic theory holds the promise that global wealth can be more evenly distributed in a sustainable future by reducing obvious imperfections in markets, like those revealed by comparing Atlas and PPP estimates. Comparing monetary and non-monetary measures for natural capital and human resources suggests the possibility that future generations may value something beyond a mechanistic extrapolation of past and present market prices. Large changes in wealth and the way it is distributed internationally could result from such forces over time; they would not be detected with conventional measures of income, production, and spending. The question for today's decision-makers is whether future generations would rather have us discuss and monitor such issues now, however imperfectly, or ignore such factors because of the difficulty of measurement.
NOTES
1 Monitoring Environmental Progress: A Report on Work in Progress, Chapter 8; John C. O'Connor, et al; The World Bank, 1995; and Sustainability and the Wealth of Nations: First Steps in an Ongoing Journey, Ismail Serageldin, World Bank, 1995. Further results appear in the 1996 edition of MEP.
2 An interesting variant, an endowment-based framework, is being developed by the US Government's Inter-Agency Working Group on Sustainable Development Indicators (see www.hq.gov/iwgsdi).
3 See Internet: http:llwwwesd. worldbank.org/html/esd/env/publicat/mep/gaeatoc.htm
4 See Chapter 6 of the 1996 edition of MEP for a detailed discussion of underlying issues.
5 E. Rodenburg, D. Tunstall and F. van Bolhuis, Environmental Indicators for Global Cooperation. GEF Working Paper no.11 (Washington, DC: World Bank, 1995).