There is a lot of talk about sustainability, and most people are in favor of it in a fuzzy sort of way.  The basic concern is  the sustainability (or non-sustainability) of our uses of natural resources.  Unfortunately, there's not much agreement on what sustainability really means.

What is sustainability?  The term first arose in the context of renewable resource harvesting as a management prescription: don't harvest beyond the stock's sustainable yield.  More recently, the concept has been applied to entire ecosystems and the whole world.  Most people understand sustainability as a moral obligation such as "The water, air and soil resources we inherited from our parents should be passed on to our children's generation as pure and intact as when we received them."  This is the view espoused by many environmentalists, and it's hard to argue with such noble sentiments.  Unfortunately, this definition is too vague to guide substantive policy.  People who espouse "sustainability" are often just fed up with consumerism.  The issue is further confused by some environmentalists' insistence on accounting for non-human values, e.g., animal rights.

If we really intend to pass along the same stock of resources from generation to generation, that means we can't exploit renewable resources beyond their sustainable yield levels, and we can't use any exhaustible resources at all.  If sustainability is defined as leaving the world exactly as we found it, then we can immediately reject this as impossible.  By definition, production requires the use of natural resources.  Robert Solow argues that a moral obligation to do the impossible isn't really an obligation at all.

The economist's view of sustainability is more flexible and pragmatic. Sustainability means leaving future generations the means to live as well as we do.  The "means" can include technologies and capital as well as natural resources.  The key idea here is that, over time, we can substitute technology and man-made capital for the dwindling stock of natural resources in order to maintain quality of life.

This view implies we don't owe future generations any particular levels of any resource.  Since man-made capital and natural capital are substitutable, we have to leave the world exactly the way we found it.  We don't have to maintain constant oil and gas reserves, we don't have to stop polluting entirely, we don't necessarily have to preserve every endangered species, or any particular species, from extinction.  Rather, our obligation is to invest an adequate proportion of income so that the productive capacity of the economy does not decline over the long run.

Our system of national accounts omits the opportunity costs of natural resource depletion and environmental degradation.  Gross national product (GNP) is the aggregate value of all final goods and services, or (identically) the sum of all factor payments.  Net national product (NNP) deducts depreciation from GNP, but it doesn't account for the decline in stocks of exhaustible resources, or the accumulation of pollutants in the environment.  Solow provides a compelling argument that NNP should account for these costs, and proposes a general framework for this.

He cites a theoretical investment model of John Hartwick's in which sustainability is guaranteed as long as investment at least match depreciation, including resource depreciation measured as the resource rents extracted from competitive resource markets.  So the basic prescription is to make sure we invest enough to maintain conventional capital stocks and also offset the depreciation of our resource base.   In the case of exhaustible resources, the correct valuation of a unit of resource depletion is the resource owner's marginal user cost or Hotelling rent, i.e., the market price minus the marginal cost of extraction.  Unfortunately, it is often difficult to calculate resource rents, particularly where marginal costs diverge from average costs.  If MC > AC, estimating marginal rent as P - AC overstates the depreciation of the resource.

In the case of environmental quality, the Pigovian tax on pollution provides the appropriate theoretical framework for evaluating the depreciation of the environment caused by pollution.  Unfortunately, this measurement problem is even more difficult because markets fail to price environmental quality correctly.

Solow argues that it is important to get the prices right so that the accounting is correct.  This means correcting market distortions (including market failures that lead to excess pollution) wherever possible.  If we are going to trade environmental quality or exhaustible resources for other forms of capital, we need to know the correct opportunity costs of these.

Since it is hard to determine whether we are investing enough or not, so it's probably best to err on the side of caution by providing economic incentives that promote rather than discourage investment.

This kind of sustainability argument permits us some cautious optimism, but it doesn't convince many environmentalists who worry about the irreplaceability of particular resources, the irreversibility of some kinds of environmental damage, the potential unreliability of markets and our incomplete understanding of the ecological vulnerabilities of our world.  In reality, many environmental goods don't have adequate substitutes, and many decisions have irreversible consequences that may not be fully understood when the decision is made.

Michael Toman proposes a "minimum safe standard" approach that addresses some of these concerns.  He quantifies (theoretically) ecological damages in two dimensions, damage cost and degree of irreversibility.   The minimum safe standard is represented as a society's dividing line between lower-cost, reversible ecological damages which may be acceptable as consequences of efficient market processes, and high-cost, irreversible damages which we have a moral obligation to prevent.  Ecologists who are sensitive to the irreversibilities of environmental damage might draw this line differently than economists who are sensitive to costs.

Sustainability is really a synonym for intergenerational equity.  The economic argument implies that some consumption of exhaustible resources is actually required in order to assure quality of life for future generations, since technology and capital growth depend on resource use.  Our ancestors have certainly done well by us, leaving us the technology, capital and resource stocks to live far better than they did.  They did this by saving and investing enough, creating new technologies, building our stocks of fixed capital, and  they actually did these things largely for themselves.  They damaged a lot of soil, dammed up a lot of rivers, chopped down a lot of trees and shot a lot of buffaloes, but they left us with the best standards of living any humans ever had.  So the real sustainability issue is, Are we saving and investing enough?  We're talking about investment in the broad sense: technology is an investment; knowledge is an investment; environmental protection is an investment.