Thinking about Time in the Context of Global Climate Change
The costs, effectiveness, and benefits of policy instruments to mitigate global climate change are influenced by the time at which actions are taken and at which the greenhouse gas emissions and sequestrations occur. Three dimensions are involved. The first is the straightforward matter of the timing of costs and resultant benefits with regard to the discounted present value of resources expended and benefits received. The second is the relative value decision-makers place on different beneficial effects of mitigation; goals are multidimensional, and each aspect may be met to a different degree depending on the timing of changes in greenhouse flows. The cost-benefit ratio of instruments therefore depends on the mix of goals sought. The third dimension is associated with the complex relationship between flows of emissions and sequestrations, and the resultant augmentation of the stock of greenhouse gases in the atmospheric system, inherently a time-dependent phenomenon. Each of these dimensions affects the relative attractiveness of classes of policy instruments and therefore must be taken into account in the design of an optimum system of interventions. It is not a simple matter of minimizing the dollars spent per ton reduced.
Timing of Costs and Effects (Benefits)
The issue of timing of costs and effects is straightforward. The absolute level of the discount rate to be used is a matter of great complexity and no little controversy. However, as long as it is not zero, the earlier that benefits can be received and the longer that costs can be delayed, the betterall other things held equal. The premise behind this conclusion is that resources are fungible and have alternative usesin satisfying consumption needs and augmenting future production through investment. What makes
this a matter of special concern here is that the value placed on the specific mitigation effects of instruments depends crucially on the mix of goals sought.
Goals of Mitigation of Global Climate Change
The mitigation of global climate change is not a one-dimensional phenomenon such that all possible benefits of policy actions are achieved simultaneously. This complicates the ranking of instruments on a cost per ton basis. To take different components into account in making rankings, it is first necessary to decompose the bundle of potential desired goals and then to determine how each possible policy instrument furthers or hinders the satisfaction of each. Furthermore, before the instruments can be put on a common basis it is also necessary to form some judgments about the terms of the acceptable trade-offs among mitigation goals. A similar process is required to compare the cost incurred in pursuing the use of a mitigation instrument with the cost of adaptation or of meeting non-climate-change goals such as faster economic growth or increased consumption.
Three component subgoals of global climate change mitigation can be posited. The first is to reduce the rate of change in the stock of greenhouse gases, on the twin premises that the speed of global climate change is sensitive to relatively small changes in the stock and that damage wrought is an increasing function of the rate of change. This goal would stress the avoidance of sudden increases in the flow of greenhouse gases, for example, even at the cost of giving up some reductions in the long-term level of the stock.
The second subgoal is to reduce the total amount of global climate change experienced between now and some future time, with the endpoint defined either arbitrarily or as the point where the global climate system is again in equilibrium. The presumption here is that the damage to be mitigated arises from the integral of global climate change over each year between now and the endpoint selected. This might be loosely termed the ''total damage borne" measure, and a proxy for its mitigation is the sum by years of the augmentation of the stock of greenhouse gases avoided.
The third subgoal is to reduce the ultimate level of global climate change at the chosen endpoint. Pursuit of this goal presumes that the time path of global climate change is of little consequence as long as the policy instruments result in an acceptably low level of ultimate change. As a proxy for this goal, the target is the ultimate level of the stock of greenhouse gases. Total benefits of mitigation would presumably be maximized by some optimal combination in achieving all three of these subgoals.
Figure B.1 illustrates these concepts in a schematic way. It shows the stock of greenhouse gases that is taken as a proxy for global climate change.
Line 2 at the top, business as usual (BAU), illustrates the system without policy intervention. The slope of this line (for simplicity shown as linear with time) is a proxy for the speed of global climate change. Line 1 illustrates stabilization of greenhouse gas stocks at the present level. The difference between lines 1 and 2 shows the greenhouse gas stock proxy for the cumulative exposure to global climate change due to future anthropogenic augmentation of greenhouse gases. The vertical difference between the
BAU line and line 1 represents a proxy for the level of global climate change introduced to the system at any point in time due to future anthropogenic activity.
It is obvious on inspection that changes in the path of the stock through the use of policy instruments can have different effects on satisfying each of the three subgoals posited. For example, a policy that resulted in severe depression of the line through much of this time, followed by a rapid increase up to and beyond the BAU level, would reduce the total exposure to climate change, but at the cost of rapid change later and of a higher end-point value. In contrast, a rapid short-term increase of greenhouse gas stocks, in exchange for earlier stabilization and a lower endpoint, would subject the system both to the damage of rapid increase in the short run and to greater cumulative global climate change borne. The point is clear: Realization of components of the bundle of desired mitigation effects may not be achieved simultaneously by policy instruments. Trade-offs among them may be necessary. It follows that three requirements must be met before optimizing policy choices among instruments can be made. First, it is necessary to determine the damage functions associated with the three separate aspects of global climate changespeed, total quantity experienced, and ultimate level. Second, it is necessary to associate these with changes in the stock of greenhouse gases. Third, it is necessary to determine how candidate policy instruments affect the stock and how much they cost. As complex as they are important, these matters are likely to be beyond careful estimation for some time. This does not mean, however, that they are subjects that can rightfully be ignored by policymakers choosing among instruments. On a gross and intuitive level, the trade-offs involved among these goals can be incorporated usefully into decisions about the degree to which different classes of instruments should be pursued.
The Flow-Stock Relationship
Greenhouse gases are emitted through both natural and anthropogenic processes and are subsequently sequestered or serve to augment the existing stock in the atmosphere. Sequestration occurs both in unmanaged sinks such as the oceans and in sinks subject to human influence such as forests. Greenhouse gases are also transformed and lose their greenhouse property over time. (This attenuation, which differs in rate among greenhouse gases, is ignored here because it does not affect the essence of this analysis.) The stock is many multiples of the flow and, consequently, exhibits substantial inertia. Further, the portion of the flow that is subject to human management is a fraction of the flow through the system, adding to the inertia of the stock to flows notionally within human control. There are also numerous lags and feedbacks in the system that affect the flow-stock relationship
(but these are ignored here as well). Also relevant is the hypothesis that past increases in greenhouse gases have yet to be fully reflected in observed global climate change. This suggests that stabilization of the stock of greenhouse gases would still leave additional global climate change in the system; a reduction of the stock would be required to stabilize the climate itself.
Highly simplified, the basic relationships are illustrated in Figures B.1a and B.1b, which describe stocks and flows, respectively. The vertical scales differ enormously between the two diagrams; in each, the scale is broken to exaggerate the changes relative to the base. Line 2 in Figure B.1b describes the BAU trend of flows of greenhouse gases net of BAU sequestration. It is associated with the BAU trend of greenhouse gas stocks previously described in Figure B.1a. Line 1 of Figure B.1b shows net flows kept constant at current rates, just as line 1 of Figure B.1a shows constant stocks at the current level. The BAU lines of Figure B.1 are used in subsequent diagrams as the base case to which the effects of classes of instruments are compared. Actual future trends, in reality, will probably be nonlinear (i.e., curve) with respect to time, but the trend line is shown here as linear to illustrate the principles involved in characterizing different policy instruments.
Characterization of Classes of Policy Instruments
The conclusion that follows from the above discussion is that the time dimension is a useful addition to the evaluation criteria used to choose among policy instruments. Their relative attractiveness depends on more than their resource costs and the number of tons removed:
• It depends on when the costs are incurred (the later, the better) and when the benefits are felt (the sooner, the better).
• It depends on the effect of the instrument on the speed with which the climate change occurs (the slower, the better).
• It depends on the effect of the instrument on the total global climate change experienced, as summed over the years from the present to the endpoint (the smaller the total, the better).
• It depends on the effect of the instrument on the ultimate level of global climate change imposed on the future at the chosen endpoint (the lower, the better).
Major classes of policy instruments are discussed below with reference to the above time-related criteria.
Temporary Reduction in Greenhouse Gas Flows: Class 1
One class of instruments yields a temporary reduction of greenhouse gas flows. Different cases are illustrated in Figure B.2. Figure B.2b shows
schematically the net change in the flow to the atmosphere by time interval. The effect on the stock is shown (greatly exaggerated for clarity) on Figure B.2a.
The prototypical example of this class would be a public relations campaign that caused thermostat adjustments lasting 1 year (case a). The stock would be permanently reduced, but of course not by much relative to the total. This case is functionally the same as a conservation activity that requires annual operation and management expenditures (case b); the action
is simply repeated each year. (For simplicity of presentation, case a is shown as following case b; case c follows both.) In cases a and b the assumption is that contemporaneous economic consumption (defined as including diminished amenities) is foregone to achieve the reduction in greenhouse gas flows and hence stock. The costs of reducing the flows are borne in the year the reduction takes place, but the benefits are experienced (essentially) forever. This class of instruments has the following characteristics:
1. Their costs are borne as the flows of greenhouse gases are reduced.
2. Permanent reduction occurs in the stock, but not in the flow.
3. There is a once-and-for-all decline in stocks that slows the speed of growth of greenhouse gases during the time of the action.
4. Timing affects the total damage borne. To the extent that this consequence is a matter of concern, early conservation is to be preferred.
5. Timing of action does not influence the final level of greenhouse gas stock. It follows that to the extent that the final level of stocks is what matters, conservation now, rather than later, is a poorer bargain.
Case c is a variant of a temporary reduction. An example would be investment in establishing a forest that sequesters greenhouse gases at an increasing and then decreasing rate, with the incremental net quantity sequestered reaching zero when the forest is in long-term carbon sequestration equilibrium. The distinction between this and the previous two cases rests on the timing and the nature of the costs borne. There are investments in establishing the forest and continuing opportunity costs in sustaining the land in forests; the latter continue even after the forest reaches equilibrium. Case c represents a contingent reduction in the greenhouse gas stockdepending on a continuing resource use to secure.
Characteristic of this subclass are the following:
1. Costs are borne prior to any benefit as the flows are reduced and permanently thereafter to sustain the sequestration.
2. The reduction of stock is contingent upon continued expenditures; flow reductions are temporary.
3. The speed of growth of greenhouse gas stocks is slowed steadily as gases are sequestered, but the possibility of later escalation exists.
4. Same as cases a and b.
5. Same as cases a and b.
Permanent Reduction of Greenhouse Gas Flows: Class 2
Another class of policy instrument is one in which a one-time investment leads to a continuous reduction of flows of an equal amount over time. An example would be a change in a long-lived building's envelopes such that
less energy was used each year. This class is shown in Figure B.3, with the investment made in the present.
This class of instruments has the following characteristics:
1. Costs are borne before the flows of greenhouse gases are affected.
2. There is a permanent reduction in flows and a cumulative, permanent reduction in stocks.
3. There is a reduction in the speed of change in global climate, which starts at the time of the action.
4. Timing of action affects the total damage borne, and the quantity is an increasing function of how soon the instrument is used.
5. Timing of action affects the final level of greenhouse gas stocks. It follows that to the extent the final level of stocks is what is of consequence, action now rather than later is the better bargain.
Temporary Sequestration with Subsequent Release, and Variants: Class 3
This class of policy instruments is characterized by a cycling of greenhouse gases in and out of the atmosphere. It is one that illustrates to a striking degree the importance of different mitigation goals in comparing policy instruments.
The prototypical example of this class is the creation of a forest based on a temporary excess supply of land for agriculture, with that forest subsequently reclaimed to grow food. This example is shown in Figure B.4. The forest is shown as being established in the present. It sequesters greenhouse gases through time, as shown in case a. If the wood were then simply burned, the stock of greenhouse gases at the endpoint would not be affected; the timing of the flow alone is changed, as shown in case b. More to the point, though, the outcome is likely to be that some of the greenhouse gas will remain sequestered in lumber and some of the biomass will be burned to replace fossil fuels, which means that all of the greenhouse gases will not be returned to the atmosphere. This is shown as an alternative case c.
Variants of this cycling process abound. For example, many energy conservation efforts require initial energy-using investments that increase greenhouse gas flows in the short run. Creation of forests from scrubland initially releases greenhouse gases; it may be a substantial time before the initial augmented flow is neutralized.
Timing of flows and of changes in stocks is particularly important in evaluating this class of instruments. An instrument whose costs yielded climate change benefits only with a lag would have a further hurdle to pass if it led to an initial augmentation of greenhouse gas flows. This would be especially true if lessening the total damage borne were an element in the desired outcome. Furthermore, the endpoint against which the final level of stock is judged is crucial. If it occurred as sequestration ended, but before releases occurred, it would give a misleadingly favorable judgment of the instrument; the obverse is also true.
Characteristics of this class of instruments include the following:
1. The timing of costs with respect to effects on stocks is varied; for the forest example, there are up-front investment and continuing maintenance costs, with the latter ending only when the forest is reconverted to its original state.
2. Net flows are both positive and negative, depending on the point in the cycle. There may be no effect on the endpoint level of the stock.
3. The effect on the rate of change in greenhouse gas stocks is erratic; because sequestration is typically gradual and release rapid, spurts of increase are possible.
4. Timing of action does not affect the total damage borne over the cycle of sequestration and release.
5. Timing of action does not affect the endpoint level of the stock except insofar as it occurs in a particular part of a sequestration-release cycle.
6. To the extent that the endpoint level is a matter of concern, instruments of this sort are ineffective in principle.
7. The relative standing of instruments of this class is peculiarly dependent on the endpoint chose. For example, for the forest conversion case, an endpoint sooner than the time of reconversion will give a false signal of excessive effectiveness on all groundsspeed of change, total damage borne, and maximum climate change incurred.
Lagged, Uncertain Reductions in Greenhouse Gas Flows: Class 4
Some policy instruments have effects on greenhouse gas flows substantially in the future, and those effects may be uncertain. An example would be increased research and development directed toward energy-saving technologies and practices in developing countries. These expenditures would precede (perhaps by decades) reductions in flows, but (by assumption) the reductions would then be permanent with the usual effect on stock. The research and development may, of course, be fruitless, because the technology does not work, is superseded by something better, or is not adopted for other reasons. If it were used, however, it would affect all three of the possible policy endpoints by slowing the speed of change, reducing total damage borne, and lowering the final level of global climate change.
A plausible assumption is that neither the cost of the research and development nor the gestation period before flows are reduced is affected by when it is initiated. However, the earlier the research and development is done, the greater is the burden of the costs (measured as discounted present value) and also the time period over which the reductions in greenhouse gas flows are accumulated. Greater accumulation time reduces both total damage borne and the final level of global climate change. It follows that in evaluating the wisdom of undertaking such research and development, both types of benefits should be used and that they are additive. They should, however, be calculated in expected value terms by taking into account the uncertainty of their actually coming to pass. It also follows that any research and development that might be justified in the future is an even better bargain in the present. As noted, this is because the mitigation effect is an increasing function of the time between the introduction of the new technology and the endpoint of the analysis.
A further reason for early rather than later research and development is that it moves forward in time the knowledge about what reductions are possible. This information could increase the time available to plan for needed adaptation and would indicate earlier which additional, more expensive, measures might be necessary and which could be avoided.
This class of instruments has effects on flows and stocks similar to those in Figure B.3 with the reductions displaced into the future (refer to that diagram). The characteristics of such instruments are the following:
1. Costs are borne substantially before any reductions occur, and implementation has further costs. The discount rate used in the decision process has a marked effect on evaluating the instrument.
2. Once reductions begin, there is a permanent effect on flows and a cumulative, permanent reduction in stocks.
3. There is a reduction in the rate of change in global climate, beginning when flows decline.
4. Timing of action affects the total damage borne; the quantity avoided is an increasing function of how soon the instrument is used.
5. Timing of action affects the final level of greenhouse gas stocks. Again, the sooner action is initiated, the better the bargain.
6. The uncertainty of outcomes must be factored into the decision by evaluating the mitigation effects on an expected value basis.
7. Knowledge gained about the prospects for mitigation is a component of the benefits; again, the sooner it is acquired, the greater is its value.
Accelerating Reductions in Greenhouse Gas Flows: Class 5
A class of policy instruments may induce an escalating reduction in greenhouse gas flows. Such instruments have little early consequence but exponentially increasing effects over time. A good example of this class of instrument would be one that reduced population, as described further below. Another example would be investment in infrastructure for research and development on energy conservation. Still another would be expenditures that would lead to a permanent shift in attitudes toward energy-conserving social choices, such as the use of mass transit.
Population reduction is accomplished through reduction in the fertility rate. (Delaying births would only have a once-and-for-all effect.) The immediate result in countries at or near the margin of subsistence is to increase the survival rate of those children born; so in estimating the effects, it is necessary to consider the net change in increased life expectancy. A subsequent effect is to increase per capita income. That effect is strongest in the period immediately after the decline in birthrate because during this period the ratio of employed to nonemployed persons increasespartly because women are freed from child-rearing duties and partly because there are fewer persons (children) not in work force. Another factor is that child-rearing investment and expenditures (schools, medical attention, and so on) are released to other occupations. The issue is to determine how these shifts affect the net emission of greenhouse gases over time.
Conventional wisdom suggests that a declining birthrate would have little effect on net emissions in the short run and that the effect could even be to increase emissions. The latter would occur if the investment, labor, and consumption freed by having a smaller cohort of infants emitted more greenhouse gases than the activities replaced. Over time, however, the expectation is that the effects would be positive and large and would grow exponentially as cohorts of reduced size moved through the demographic cycle. This is because the emission of greenhouse gases is expected to be more responsive to falling population than it is to rising per capita income.
The conclusion that declining population leads to lowered greenhouse gas emissions depends crucially on the amount by which per capita incomes rise. The latter is a matter of fact and depends on particular circumstances. In the poorest countries, for example, declining population may free sufficient resources from basic consumption to allow substantial increases in highly productive investment. This investment could more than compensate for a declining labor force, especially in the early years after the decline in birthrate. Total output could consequently increase, not fall, and added output-related emissions could overshadow the decline in emissions associated with the population drop per se. This is especially true if emissions per person rise more than proportionately with income, which is likely to be the case in a poor country due to shifts in the mix of consumption goods toward those that use more energy. In the long run, after incomes rose sufficiently, countervailing forces would likely dominate, but that long run may be far in the future. This possibility should be taken into account in interpreting the results of the example given below.
The class of instruments that might lead to an accelerated reduction in greenhouse gas flows is illustrated in Figure B.5. The change in BAU flows from a permanent downward shift in fertility rates is shown as positive for a period after the change occurs. It then turns negative and builds in waves over time. The waves are occasioned by the movement of the diminished childbearing cohort through time. The increasing (to some limit) reduction in flows leads to an exponential decline in greenhouse gas stocks as compared to BAU.
The characteristics of instruments of this class (as illustrated in the population example) follow:
1. Expenditures occur well before greenhouse gas emission flows decrease and continue as needed to sustain the drop in fertility rates.
2. The initial impact may be to increase flows and stocks, but this is temporarythough for really poor countries successfully launched on development, it can last a long time.
3. Reductions in flows are permanent because there is a shift downward in population, including that of childbearing age, even if fertility reduction expenditures cease.
4. Once begun, there is a cumulative, permanent reduction in greenhouse gas stocks at an exponential rate.
5. The speed of the change in stocks is reduced (at a growing rate) from the time that lower fertility rates result in lower emissions. To the extent that steady movement toward stabilization is a goal, early use of this instrument is more desirable than later.
6. Timing of actions affects the total damage born; thus to the extent that lessening total damage is a goal, the sooner the instrument is used, the better.
7. Timing of actions affects the endpoint level of greenhouse gas stocks as an increasing function of the interval between initiation and endpoint, making early action relatively more valuable than later, all other things held equal.
Lessons, Implications, and Further Work
What has come before suggests that instruments can be divided into classes on the basis of the time at or during which they affect greenhouse gas emissions and stock. Division of this sort is crucial if instruments are to be reasonably compared with each other. This is true on three counts.
First, the relative desirability of instruments depends on their costs related to their benefits. Because the costs of greenhouse gas emission changes may occur at different times and their benefits extend into the future, costs and benefits of different instruments must be put on the same scale with respect to time for their comparison to be meaningful. This is done through the use of a discount rate as discussed in Chapter 20.
Second, the goal of reducing global climate change has at least three dimensions. Instruments can have different effects on the satisfaction of each, depending on the time that effects on greenhouse gas stocks are observed. This is tied to the third matter: the complex relationship between flows (which instruments affect) and stocks (which are the result of changes in flows). It is stocks and changes in them that affect the issue of policy interestglobal climate change.
As an illustration of the interconnectedness of these matters, if the final greenhouse gas stock is the outcome of interest, instruments should be measured against that, and those having the same effect on greenhouse gas flows during one period may have very different impacts on greenhouse gas stocks at the endpoint selected. This suggests that judgments of the relative cost-effectiveness of different instruments can be very dependent on the temporal endpoint chosen. A ton of flow reduced may be a bad bargain, no matter how cheaply achieved, if it is reinserted into the atmosphere before the endpoint of interest.
Proper treatment of time has another use as well. It can have important consequences for determining the optimal timing and quantity of any intervention. This is true if that decision is based on what society gets in the form of lesser global climate change vis-à-vis what it gives up in terms of current satisfaction and the enhanced ability to accommodate future adaptation. In this, fully accounting for all the positive aspects of mitigationreduced speed of change, reduced total exposure to damage, and final level of global climate changeis important. Each has separate effects on the consequences of societal interest such as rise in sea level, agricultural productivity,
and changes in ecological systems. They can also differ in their effects on the distribution of consequences over time and geography.
The purpose of this appendix is to indicate the role of the time dimension in formulating a global climate change mitigation strategy. It illustrates ways in which different instruments may lead to outcomes that diverge from those expected when only tons reduced and costs are considered. Application of the relationships discussed here requires an understanding of the physical relationships among flows, stock, and global climate change that lies beyond current knowledge. It also requires complex judgments about the trade-offs among sometimes competing policy goals. In illuminating what information is needed to formulate an efficient and effective policy, it suggests potentially fruitful areas for further research. Even before that research is done, however, policymakers can use some of these insights to select the mix of instruments that appears to have the greatest prospect for improving total welfare.