Economic Analysis and Adaptive Capacity With

Economic Analysis and Adaptive Capacity with

Reference to an Egyptian Case Study

Gary Yohea , Kenneth Strzepekb , Tammy Paua and Courtney Yohea

April 30, 2002

a Department of Economics

238 Church Street

Public Affairs Center

Wesleyan University

Middletown, CT 06459 USA

b Civil, Environmental and Architectural Engineering

Engineering Center, OT 5-23

Campus Box 428

University of Colorado

Boulder, CO 80309-0428 USA

Contact Author:

Gary Yohe

Phone: 860-685-3658

Fax: 860-685-2781

E-mail:

Abstract

A range of “not-implausible” climate scenarios is superimposed on a range of similarly “not-implausible” socio-economic scenarios for Egypt to explore the role of adaptation in reducing vulnerability and to illustrate the utility of a methodological approach derived from the determinants of adaptive capacity. The numerical results are critically dependent upon context and model specification, but several robust qualitative insights are supported. Adaptation can make a significant difference on a macro scale, especially for pessimistic climate scenarios. Socio-economic context matters in determining adaptive capacity, and inefficient investment can diminish the capacity to adapt. The value of information that supports early differentiation between two strikingly different climate futures can be significant. Moreover, early preparation can be critically important because macro-scale adaptation can involve capital reallocation between sectors in anticipation of large future investments in adaptation infrastructure. Planning for bad news and adapting to good can be a better choice than the other way around, but working to expand the potential of some options to increase adaptive capacity can create rigidities or cause systems to under-prepare for adopting other options. These omissions can reduce the capacity to cope with more extreme climate futures because the first set of adaptations may be overwhelmed even as more efficacious alternative adaptations become less feasible.

Key Words: vulnerability, adaptive capacity, climate change, socio-economic scenario, value of information

Strzepek, et al. (2001) described a process by which nine “not implausible” climate scenarios were selected for Egypt. Their selection was the first step of a project designed ultimately to conduct detailed integrated assessments of their impacts across a range of similarly “not implausible” socio-economic scenarios. Here we report on progress in defining aggregate portraits of the requisite diverse set of socio-economic scenarios and in exercising those scenarios to investigate the economic implications of micro- and macro-scale adaptations. Impacts on agricultural and non-agricultural production of climate induced reductions in flow along the Nile will be the focus of our attention, but this is not an Egyptian case study. It is, instead, a methodological piece that uses an Egyptian context to explore how the fundamentals of economic analysis might be applied to adaptation issues. It turns out, however, that the Egyptian context is particularly well suited for this task.

A rich diversity of possible socio-economic scenarios was created by spanning a range of not-implausible representatives of a few macro-scale determinants of adaptive capacity. The first section sets the stage by reviewing those determinants within an pedagogical model that sees vulnerability to climate change and climate variability as a function of exposure, sensitivity, and adaptive capacity. Section 2 then reviews the representative climate scenarios from the earlier work by Strzepek, et al (2001) that display a range of Egypt’s exposure to climate change. Indeed, they include one scenario in which flow in the Nile actually increases, but eight less optimistic alternatives range from modest reductions to declines that would appear to be quite severe. A third section follows with a description of a Ramsey-style aggregate growth model that was designed specifically to accommodate investigations of the relative efficacy of municipal and industrial recycling, drip irrigation, and groundwater pumping in relieving climate induced stress on macroeconomic activity and food self-sufficiency. Details of the model are provided in Appendix A. As reported to us by the Minister of Water and Irrigation, macroeconomic vitality and food self-sufficiency are both explicit policy objectives of the Egyptian government; and these three adaptations are under active consideration in support of both.

A collection of representative socio-economic scenarios that extends across a wide range of “not-implausibility” defined by variation the macro scale determinants of adaptive capacity for these options is described in the Section 4. It is within these scenarios that Egypt’s sensitivity to the climate scenarios described in Section 2 are explored in Section 5 with careful attention being paid to the potential efficacy of the three possible adaptations. Section 6 then displays the utility of the conceptual methodology described in Section 1 by offering some interpretative results drawn from that structure. A seventh section responds to a hypothesis that was derived from the adaptation results of Section 5 by exploring the value of information in implementing adaptation before a closing section offers some contextual conclusions. Notwithstanding the specific context from which they were drawn, we expect that these conclusions hold considerable validity beyond the boundaries of the Egyptian illustration.

1. Adaptive Capacity.

The Intergovernmental Panel on Climate Change (IPCC) envisioned a broad relationship between vulnerability, sensitivity, and adaptive capacity in its Third Assessment Report [IPCC (2001)]. As they reviewed adaptation and adaptive capacity in this context, the authors of Chapter 18 on “Adaptation in the Context of Equity and Sustainable Development”, as well as the authors of other sector and regional chapters, came to recognize that this relationship was complex, location specific, and path dependent. Indeed, many would now contend that any subsequent analysis that did not recognize regional diversity in development trajectories, uncertainty in climate futures, and the potential for adaptation would be suspect.

To be more specific, the authors of Chapter 18 [IPCC (2001)] concluded that adaptive capacity varies significantly from system to system, sector to sector and region to region. Indeed, they noted that the determinants of adaptive capacity include a variety of system, sector, and location specific characteristics:

1. The range of available technological options for adaptation,

1. The availability of resources and their distribution across the population,

1. The structure of critical institutions, the derivative allocation of decision-making authority, and the decision criteria that would be employed (i.e., governance),

1. The stock of human capital including education and personal security,

1. The stock of social capital including the definition of property rights,

1. Access to risk spreading mechanisms,

1. The ability of decision-makers to manage information, the processes by which these decision-makers determine which information is credible, and the credibility of the decision-makers, themselves, and

1. The public’s perceived attribution of the source of stress and the significance of exposure to its local manifestations.

Many of these determinants cannot easily be quantified, but working through their content from the bottom up or from the top down can nonetheless uncover practical insights that can inform our understanding of how adaptation might diminish vulnerability.

The Egypt work reported here takes a top-down approach based on the observation that many of the determinants of adaptive capacity operate on macro-scales in which national or regional factors play the most significant role. While the set of available, applicable, and appropriate technological options (Determinant 1) for a given exposure at a particular location might be defined on a micro-scale, for example, the complete set of possible remedies for a national response should have macro roots. Determinants 2 through 6 should all have large macro components even though their micro-scale manifestations could vary from location to location or even from adaptation option to adaptation option. Resources (Determinant 2) could be distributed differently across specific locations, but adaptive capacity may be more sensitive to larger scale issues that determine the availability of resources across an entire nation. The essential questions here focus on whether sufficient funds are available to pay for adaptation and whether the people who control those funds are prepared to spend them on adaptation. Macro-scale and even international institutions (Determinant 3) could also certainly play a role in determining how decisions among various adaptation options might be made and who has access to the decision-making process. Adaptation projects will be directed towards improving well-being measured against domestically determined objectives. The stocks of human capital and social capital (Determinants 4 and 5) could be locally idiosyncratic, as well, but their local manifestations would likely be driven in large measure by macro-scale forces, such as national education programs and the efficiency of public investment. Access to risk spreading mechanisms (Determinant 6) usually evokes notions of insurance and monetary compensation after the fact; by their very nature, these are macro in scale. There are, however, many instances in which adaptation before the fact can function as a physical “insurance polity” on a micro scale – a process by which exposure and/or sensitivity might be diminished in the face of an uncertain future. Finally, Determinants 7 (informational management) and 8 (attribution of signals of change) for national adaptations must have macro-scale foundations even if their force is derived from local vulnerabilities.

1. Climate Scenarios for Egypt.

Panels A and B of Figure 1 display nine representative climate scenarios in terms of flow into Lake Nasser and the area of upstream swamps in the Sudan; they represent the primal exposure of Egypt to climate change. Each was driven by specific assumptions about greenhouse gas and sulfate emissions, climate and sulfate aerosol sensitivities, and the results of some specific global circulation model, but each was selected for its representative value. Taken together, these nine scenarios span a range of outputs produced by running COSMIC for rainfall and temperature for nine upstream countries through a hydrological model authored by Yates and Strzepek (1998). Decadal markers between 2000 and 2100 are depicted, and swamp area is included because draining that swamp could have been a macro-scale adaptation along water-scarce futures. Given the political friction that would be created by draining a Sudanese resource to sustain Egyptian economic activity, it is perhaps good news that the swamp would, under these scenarios, not be available.

3. An Economic Model for Egypt.

A modification of the classical Ramsey analysis of optimal economic growth under certainty provided the modeling context for describing how Egypt might move into the future from these “initial conditions”. It is described in some detail in Appendix A. Blanchard and Fischer (1989) and Barro and Sala-i-Martin (1995) offer general discussions of the fundamental Ramsey construction, but a non-linear formulation of the classical Ramsey model developed by Lau, et al (2001) was employed. The primal formulation was based on an explicit representation of utility for a representative household that depended on per capita consumption. The social planner maximized its present value subject to the constraint that output in period t was either consumed or invested. It is convenient to think of the production function exhibiting constant returns to scale in capital and a second factor whose supply would be exogenously specified. The capital stock in each year equaled the capital stock at the start of the previous year less depreciation plus investment in the previous period.

Several major modifications were made to this model to represent the Egyptian economy more accurately. Production was, first of all, divided into two major sectors with sector-specific capital. The agricultural sector produced only food for domestic consumption. The non-agricultural sector produced an output that could be invested to produce capital, consumed, or exported to pay for imported food. Secondly, the single household differentiated two types of consumption goods: food and a non-food consumable. The underlying accounting process noted two types of food. The first sustained a minimum caloric requirement of 2100 cal/per capita/day; it was included as a constraint in the model. The second built supplemental caloric intake up to 1100 cal/per capita/day. Supplemental calories were preferable, on a declining scale, to non-food consumables. This complication added a food balance constraint, and changed the modified objective function.

The classic Ramsey model does not allow for trade, of course, but Egypt is currently only 70% food self-sufficient. Assuming a closed economy would therefore be inconsistent with current and, in all likelihood, future realities. The model was modified to allow non-agricultural output to be exported for equal the amount of food imports subject to the terms of trade for Egyptian non-agricultural output on the world market. The terms of trade parameterwas exogenously specified so that a balance equation for production and consumption could be specified.

Water is, of course, an essential factor of production in the fully irrigated Egyptian agricultural sector as well as an important factor in hydroelectricity production, some industries and transportation. Water was therefore included in the model in a dynamic Leontief fashion. Two different by constant rates of technological progress were specified for each of the two sectors so that water use per unit output in either sector would decline over time. Water was also demanded for domestic consumption, and this use received the highest priority. Domestic water use was modeled as a function of income expressed by GDP/capita. Domestic water use typically increases at a very steep rate until incomes reaches approximate $2000 per capita then it becomes almost constant. Taking these three demands together, a water balance equation was imposed on the model.

Finally, it should be noted that specifying exogenous target levels of food self-sufficiency frequently resulted in infeasible solutions when the potential impacts of climate change on Nile flows and agricultural water requirements were extreme. To avoid this problem, the level of food self-sufficiency was modeled as an endogenous variable to be determined by the model. Published reports and private communications with the authors reveal that maximizing food self-sufficiency is policy goal of the Egyptian government. To model such a policy, the objective function was modified to reflect disutility derived from importing food equal to consumption of non-agricultural good. This structure placed additional weight on minimizing food imports because each unit of food imports caused a double loss in utility.

4. Selecting the Socioeconomic Scenarios for Detailed Analyses of Adaptations.

The selection process for socio-economic scenarios began by running the Ramsey model for more than 600 combinations of nine climate scenarios, two alternative population futures, and two alternative specifications of four underlying socioeconomic parameters. Table 1 summarizes the variables that determined the range of possible socioeconomic context. Two population scenarios allowed population to stabilize by the middle of the 22nd century at 1.5 to 2.5 times current levels. The resulting range spanned projections offered by various sources using different assumptions about family size, longevity, and cultural perspective.

The determinants of adaptive capacity described in Section 1 highlight the potential significance of resource availability (and distribution) as well as the ability of decision-makers to allocate those resources effectively. Strzepek, et al. (2001) reinforced this message by noting different capacities to adapt under high and low capital futures. Table 1 shows that variation across high or low paths for three critical parameters reflected the import of these insights with considerable richness. Nonagricultural productivity growth, for example, assumed high or low values of 2% or 1% per year. Agricultural yields were similarly given high or low trajectories with rates of 1.5% or 0.5% per year. Finally, the efficiency of investment was assumed to be high (normal) or low by assuming values of 1.0 or 0.8, respectively. The low value in this case meant that one unit of output devoted to investment would, by virtue of misallocation by government planners and/or associated second-best allocations by private investors, increase the “effective” capital stock by only 0.8 units. Finally, Egypt’s future trading position in the world market will be a critical determinant of the availability of resources. The terms of trade were therefore assumed to be favorable or unfavorable by assuming high or low values of 1.0 or 0.8, respectively. Low terms of trade meant, for example, that one unit of consumable goods traded on the world market would produce 0.8 units of imported food.

Figure 2 displays the results of these runs in terms of an index of Egyptian food self-sufficiency and total food plus consumable good consumption in the year 2067 drawn from more than 600 different combinations of variables identified in Table 1. The year 2067 was chosen because it reflects a point in the relatively distant future by which time the nine climate scenarios had, for the most part, diverged. The implications of climate and socioeconomic circumstances were therefore fully represented. The food self-sufficiency coefficient reflects the proportion of total food consumption supported by domestic food production. It was chosen as an important indicator of Egypt’s future because of the importance placed on food security by the Egyptian government; this policy objective was identified in Strzepek, et al. (2001) as a critical differentiating characteristic of future economic vitality. The sum of food and consumable consumption was chosen, as well, because it reflects critical components of the determinants of domestic welfare; it is reflected as a multiple of the level achieved in the year 2000. Finally, there is nothing special about range of results produced for the year 2067. Other years, from roughly 2030 through 2100, produced ranges that were entirely comparable to those depicted in Figure 2.

Linear patterns are clear in Figure 2, and investigating their sources made it relatively easy to select a manageable number of representative scenarios. The first step sorted the results by climate regime, and six groupings emerged. The patterns for climate scenarios 1, 5, and 6 seemed to be unique, for example, but points indicating various socioeconomic futures for scenarios 2 and 3 tended to bunch together. Climate scenario 3 was selected to reflect these possibilities. Scenarios 4 and 8 also displayed similar patterns, so scenario 8 was selected. Finally, scenarios 7 and 9 displayed results with little diversity; and scenario 9 was selected to carry those futures forward. Placing the runs from these climate scenarios into 6 groups also produced patterns – different clusters for different population scenarios. Each showed varying ranges of food self-sufficiency and total consumption for across the various socioeconomic specifications. Some of these ranges were large; others were quite small. The limits of each, though, could be captured by the same combination of parameterizations; they are identified in Table 2 as (socio-economic) scenarios A through F. Figure 3 places these 36 representations into the context of Figure 2 to show the considerable degree to which, taken together, they span the diversity of the original larger set of possible futures.