Reseval Proofs Briones

Research Evaluation, volume 14, number 3, December 2005, pages 000–000, Beech Tree Publishing, 10 Watford Close, Guildford, Surrey GU1 2EP, England

Economic surplus analysis Ex ante impact assessment for research on natural resources management: methods and application to aquatic resource systems Roehlano Briones, Madan Dey, Ilona Stobutzki and Mark Prein

Under a particular representation of the impact pathway of natural resource management (NRM) research, economic surplus techniques can also be used for ex ante impact assessment. The method is applied to the case of the WorldFish Center, an international organization specializing in research on aquatic resources in developing countries. A survey of expert opinion is used to estimate productivity improvements and adoption rates for NRM research and its application. A supply–demand model for aquatic commodities is constructed to calculate the resulting change in economic surplus. Results indicate that ex ante economic impact is highest for NRM on coral reefs and inland aquatic systems.

Roehlano Briones was a postdoctoral fellow at the WorldFish Center, Jalan Batu Maung, Batu Maung, 11960 Bayan Lepas, Penang, Malaysia, at the time this study was conducted. Madan Dey, Ilona Stobutzki and Mark Prein are WorldFish Center scientists. Address correspondence to: Roehlano Briones, E-mail: [email protected]

Research Evaluation December 2005

E

CONOMIC TECHNIQUES FOR ex ante impact assessment are accepted tools for evaluating agricultural research. The need to identify high-impact activities is keenly felt in the public and nonprofit sector of developing countries, where funds for research are seriously constrained. Among these analytical techniques, the economic surplus method is the most popular and arguably the most fruitful approach for impact assessment (Alston et al, 1995). Applications of economic surplus analysis are widespread in the case of farm commodity research. Recently, with growing emphasis on sustainable development, natural resources management (NRM) has become a major category for agricultural research. Unfortunately, NRM research has so far resisted economic surplus analysis and other conventional techniques of impact assessment, whether ex post or ex ante (Izac, 1998; Pachico, 1998; Maredia et al, 2000; CGIAR, 2000; Pingali, 2001). The differences in evaluation performance may be traced to the nature of the impact pathways. For farm commodity research the impact pathway is straightforward. As impact depends on conditions in parceled, artificial systems (farms), and the decentralized choices of many individual farmers, the law of large numbers permits a reasonable approximation of productivity impacts and adoption decisions over time. However, for NRM, the impact pathway is much more complex. Impact depends on conditions in large-scale systems, for which the effects of human activity are poorly understood. Moreover,

0958-2029/05/030000-00 US$08.00  Beech Tree Publishing 2005

1

Natural resources management

adoption of research recommendations depends on policy choices, set by the state under the influence of socio-cultural and political factors. Predicting research influence and the impact of research is a difficult and highly uncertain affair. In practice, NRM research is evaluated or prioritized based on direct subjective approaches, such as scoring (Kelly et al, 1995; Randolph et al, 2001). While subjective approaches have their place, it is desirable to integrate NRM into the wider ambit of economic analysis. Reducing impacts on production and consumption to a common metric of economic value makes the choice issue more transparent. Furthermore, the evaluation exercise can avail of a wide body of evaluation techniques that extend conventional economic surplus analysis, such as incorporating modifiers for uncertainty, non-market economic values, and the like. The main contribution of this study is to demonstrate that economic surplus analysis can be used for ex ante evaluation of NRM research. The critical ingredient is an articulation of the impact pathway for NRM research, which exploits an analogy from the impact pathway for farm commodity research. The pathway suggests a procedure for eliciting expert opinion regarding the likely supply shifts resulting from research and application. These shifts are incorporated into a supply–demand model, which computes the welfare impact (a generalized version of the economic surplus measure). Combined with an estimate of the required R&D investments, one can then calculate benefit–cost ratios as an indicator of ex ante economic impact. The method is applied to impact assessment for the WorldFish Center, an international organization for research on aquatic resources in developing countries, and a member of the Consultative Group on International Agricultural Research (CGIAR). Results are presented and some implications for research and for research planning are discussed.

Framework Impact pathways Impact pathway analysis identifies causal links by which research achieves its intended benefits. Elaborating these links compels researchers to identify key stages from research to impact, the expected effects at each juncture, the indicators for measurement, and the processes securing the links (Springer-Heinze et al, 2003). In the following an impact pathway analysis is presented for farm commodity and NRM research. The two categories of research are distinguished mainly by the type of resource system being studied. Farm commodities are produced in divisible systems, for example, farms or ponds, for which rights to use are defined and enforced at an individual level. Accordingly, allocation choices of individual

2

Impact assessment for farm commodity research entails an estimate of potential on-farm benefit, measured by some indicator, for example, higher yields (consistent with lower per unit costs)

farmers are largely independent. However, natural resource systems are common pool resources for which subdivision into individual parcels is infeasible or uneconomical. Common examples are wild fish stocks and many upland forests. Harvesting decisions are therefore interdependent; however, this interdependence is ignored by individual harvesters, leading to resource degradation (Ostrom et al, 1994). NRM aims to regulate extraction to maintain the sustainability of harvesting activities.1 For farm commodity research, the impact pathway is relatively straightforward: research generates a specific innovation, such as a germplasm, a new form of input, a new farm practice, and so on. The innovation is then adopted by farmers (which may be accompanied by local modifications by adopters; these modification may themselves be the subject of further R&D.) Benefit from adoption most often takes the form of improved long-term productivity.2 (Other types of benefits include higher price, due to better product quality, or reduced variability of yield.) The process of adoption is a combination of natural diffusion, as farmers copy one another or share information, and deliberate dissemination through the extension and farm support system. Impact assessment for farm commodity research therefore entails an estimate of potential on-farm benefit, measured by some indicator, for example, higher yields (consistent with lower per unit costs). This is followed by an estimate of the following: • The adoption process over a particular extrapolation domain; • The average on-farm benefit upon adoption, that is, with an adjustment for yield gap, which is the shortfall between average yield and yield under best practice (Dey et al, 2000). For economic surplus analysis, these estimates are used to calculate supply shifts; simulation analysis using a supply–demand model then calculates gross benefits from the research–application continuum. Given a reasonable estimate of the required investment in research and application, one can then calculate the net economic impact of farm commodity research. Meanwhile for NRM research, the pathway to

Research Evaluation December 2005

Natural resources management

impact takes the route of resource management. A major constraint to resource management is the lack of reliable information. Often the status of a particular natural stock or ecosystem is uncertain. Furthermore, the causal relations may also be vague or weakly established, whether in terms of the impact of human activity, or the likely outcomes of various management options. Finally, the long-term economic importance of maintaining resource sustainability may also be poorly understood, undermining the argument for immediate and decisive action, particularly when protecting the resource may imply serious economic dislocation. Both biological and social science research play a vital role in management support, by assessing the status and values of the resources, and by identifying the likely impacts of human activity, management actions, and institutional arrangements. The simplified impact pathway for both farm commodity and NRM research is drawn in Figure 1. Farm commodity research generates technologies and recommendations for farmers; farmers adopt recommendations and increase the farm productivity. Similarly, NRM research produces recommendations, or an evaluation of management options; this

improves the quality of NRM decisions, which leads to an increase in resource productivity. In both cases, the results of applying research become feedback for later iterations of the research cycle (signified by the upward arrow in Figure 1). Based on this pathway analogy, two measures need to be estimated for NRM research: first is the potential yield improvement under a best-practice scenario (that is, proper or ideal application of NRM research); second is the extent to which ideal application of NRM research is adopted over time. In this study, management performance is subsumed within the estimate of adoption. For example, if the adoption of NRM research is 80%, but quality of research application is only 50% of the ideal (notionally speaking), then the adoption figure is adjusted down to 40%. This technique automatically incorporates the ‘management performance gap’, an indicator that parallels the ‘yield gap’ in the adoption of farm commodity research. Indicators and quantification procedure The column at the right-hand side of Figure 1 suggests indicators, at critical stages in the pathway, to

Research

Technology recommendation evaluation

Potential production gain per unit

Production gain per unit Disseminationadoption

Adoption rate (A10,Amax) R&D cost

Improved management

Increase in supply

Change in economic surplus Market

Net economic benefit

Impact Figure 1. Simplified impact pathway for farm commodity and NRM research Note: Expert judgment is applied to estimate production gain per unit and to calibrate the parameters of the adoption function (stated in text as adoption ceiling and adoption in ten years). R&D cost is estimated by output value shares based on literature review. Change in economic surplus is evaluated using a supply–demand model; this benefit, net of cost, yields the net economic benefit.

Research Evaluation December 2005

3

Natural resources management

be used for ex ante impact assessment. Given the paucity of ex post impact assessment for NRM, estimation of these indicators relies on expert judgment. Suppose one has defined a particular resource system for which NRM research impact is to be evaluated, and has obtained benchmark production data for that system. Two scenarios may be considered: a ‘business-as-usual’ scenario for harvesting behavior and NRM performance, and a ‘bestpractice’ scenario for research and applications on NRM. Resource experts (preferably practicing biologists) are asked to estimate production under these two scenarios, relative to the benchmark production. The difference in production between the two scenarios is the potential productivity gain from NRM research. The remaining indicator to be estimated is the adoption of the best-practice scenario. In farm commodity research, adoption dynamics are often modeled on the sigmoid curve (Alston et al, 1995). A similar logistic process is assumed to hold for NRM research. Let A denote the extent of adoption, here measured as the percentage of output produced under the best-practice application of NRM research. Let Amax denote the maximum extent of adoption, or the adoption ceiling. The logistic model assumes:

A& = A

æA - A ö÷ ÷ r ççç max ÷ è Amax ÷ ø

(1)

where r is a constant. That is, the growth of the extent of adoption is a constant multiplied by the proportional difference between the ceiling and current adoption. Here time may be measured in years, with the beginning year of the evaluation set at zero. Denote adoption at time zero as A0, and let k = r Amax . Then the time path of A is described by the following equation:

A(t ) =

Amax A0 A0 + ( Amax - A0 )e- kt

(2)

Once Amax has been estimated, the remaining problem is to parametrize k (or alternatively, r). Suppose an estimate of adoption rate in year 10 is available. Arbitrarily setting A0 = 0.005 (one-half of one percent), the parameter k can be calibrated as follows:

é( A - A(10)) A0 ù ú k = - 0.1 *ln ê max êA(10)( Amax - A0 ) ú û ë

The WorldFish Center The WorldFish Center is one of 16 Future Harvest Centers of the CGIAR. Its mission is to reduce poverty and hunger by improving fisheries and aquaculture. To guide WorldFish in its research planning, the foregoing framework is applied to NRM research for aquatic resources in the developing world. The units of assessment are the major categories of aquatic resource systems; these resource systems are further disaggregated into the major regions of the poorer areas of the developing world. The definition of units follows the WorldFish Strategic Plan (ICLARM, 1992). For aquatic resource systems the categories are: • ‘Lakes’ — including reservoirs, are mainly freshwater bodies, whether natural or artificial, used for irrigation, power generation, and household water supply. • ‘Rivers’ — includes streams and floodplains; streams and rivers are flowing waters, while floodplains are low-lying areas adjacent to watercourses, subject to periodic or near-permanent inundation. • ‘Coasts’ — includes estuaries and lagoons, and critical habitats, such as mangroves. Coastal waters (10 m deep or lower) encompass most fishing grounds of small-scale fishers. • ‘Coral reefs’ — continental and island shelves in tropical oceans, in which reef-building corals are dominant features. In this study, the developing world is limited to Asia and Africa, which hosts the bulk of the world’s poor and undernourished. The regional divisions are: • East Asia (EA) • South Asia (SA) • Southeast Asia (SEA)–Island (Indonesia, Malaysia, and the Philippines); • SEA–Mainland (the Mekong countries, including Burma); • West Asia and North Africa (WANA); • Sub-Saharan Africa (SSA). The countries under each region are listed in Table 1. There are 24 region-resource system combinations (= four resource systems × six regions), comprising the units of assessment. Data for the resource systems and regions is partly available from the FAO fisheries statistical databases on capture production.

(3)

where ln denotes the natural logarithm. Estimates of ceiling adoption and adoption in year 10 can be elicited from policy experts and socio-economists in the field of NRM research.3

4

International research on aquatic NRM

Eliciting expert opinion The base year is set at 2001, the most recent year available during the period of the study. Three questionnaires were formulated, one each for coast, inland waters, and coral reefs. Subdivision between

Research Evaluation December 2005

Natural resources management Table 1. Country groupings by region of analysis

East Asia China Korea, Democratic Republic Mongolia South Asia Bangladesh Bhutan India Nepal Pakistan Sri Lanka Southeast Asia – Island Indonesia Malaysia Philippines Southeast Asia – Mainland Cambodia Laos Myanmar Thailand Vietnam Sub-Saharan Africa Angola Benin Botswana Burkina Faso Burundi Cameroon Central African Republic Chad Congo, Dem. Rep. Congo, Rep. Côte d'Ivoire Djibouti Equatorial Guinea Eritrea

Sub-Saharan Africa (cont.) Namibia Niger Nigeria Rwanda Reunion Senegal Sierra Leone Somalia Sudan Swaziland Tanzania Togo Uganda Western Sahara Zambia Zimbabwe West Asia and North Africa Algeria Bahrain Egypt Iran Iraq Jordan Kuwait Lebanon Libya Morocco Qatar Saudi Arabia Syrian Tunisia United Arab Emirates Yemen Oman Turkey

lakes and rivers is omitted in the questionnaires owing to absence of benchmark data. The Coasts Questionnaire is subdivided into the FAO ocean areas. For each ocean area, respondents are shown annual benchmark production figures, averaged over the period 1999 to 2001 (Table 2). Production is divided into the four major fish types (demersal fish, pelagic fish, crustaceans, and mollusks). Respondents are presented with percentage intervals above and below the benchmark figure, for example, 5% to 10% below the benchmark, 0%

Table 3. Average annual production growth for inland capture, by region

Production, Average annual growth (%) 1999–2001 (average 1991–1995 1996–2001 annual, in mt) East Asia

2,243,197

12.6

7.4

South Asia

1,960,877

3.0

6.7

Southeast Asia – Island

461,219

–0.4

–2.6

Southeast Asia – Mainland

684,205

2.4

12.4

Sub-Saharan Africa

5,325,738

0.7

1.7

West Asia and North Africa

1,303,894

7.4

2.6

World

8,665,072

3.9

4.4

Source: FAO (2003)

to 5% above the benchmark, 5% to 10% above the benchmark, etc. There are three time horizons: the short term (years 1 to 5), the medium term (years 6 to 10), and the long term (years 11 to 20). The respondents are requested to select their best guess about the appropriate interval, for each time horizon, and for each scenario (business-as-usual and bestpractice). The production estimates are then apportioned to the regions of analysis using the ratios found in the benchmark data. Once the average estimate is computed for the four fish types, a simple average of these estimates is applied for the fish types not elsewhere classified. The questionnaire for inland waters is formulated in a similar manner as for coasts, though estimates are elicited for only one fish type (inland freshwater fish). Benchmark production figures (as well as regional growth rates) are shown in Table 3. The Coral Reefs Questionnaire takes a more roundabout approach. First, a literature search was conducted to find a regionally available indicator for the state of coral reefs worldwide. The search yielded the risk category presented in the Reefs at Risk study

Table 2. Average annual production by FAO ocean area and major fish type, in mt (average for 1999–2001)

Ocean area East Central Atlantic Southeast Atlantic Southwest Atlantic West Central Atlantic East Indian Ocean West Indian Ocean Mediterranean and Black Sea East Central Pacific Northwest Pacific Southeast Pacific West Central Pacific World total

Demersal 447,254 429,938 907,491 186,230 757,659 1,123,655 279,347 100,477 6,301,317 571,223 1,339,206 12,443,797

Pelagic 2,583,125 1,079,737 165,772 861,797 1,402,512 1,751,436 933,532 1,159,367 6,111,455 13,041,127 4,784,713 33,874,573

Crustacean

Mollusk

71,779 18,841 107,416 279,138 345,308 341,446 48,671 75,184 2,882,459 50,531 607,880 4,828,653

236,104 8,899 1,047,849 257,098 320,217 130,634 162,601 193,110 3,059,146 232,584 552,410 6,200,652

Source: FAO Fishstat (2003)

Research Evaluation December 2005

5

Natural resources management Table 4. Distribution of world coral reefs by risk status and region

Region

Coral reef area (km2)

Present risk status (%) Low

Middle East

20,000

Caribbean

Medium

High

Total

39

46

15

100

20,000

39

32

29

100

Atlantic

3,100

13

32

25

70

Indian

36,100

46

29

25

100

Southeast Asia

68,100

18

26

56

100

Pacific

108,000

59

31

10

100

World

304,260

42

31

27

100

Source: Bryant et al (1998)

(Bryant et al, 1998). The study presents the percentage distribution of reefs in global regions classified as ‘low’, ‘medium’, and ‘high’ risk (Table 4). The respondents are asked to regard these categories as proxies for extent and severity of reef damage. They are then requested to estimate the percentages of low-, medium-, and high-risk coral reefs; as with the other resource systems, the percentages are reckoned within an interval, rather than an exact number. The estimates are made by region, 20 years hence, under the business-as-usual and best-practice scenarios. Also elicited is the estimate of coral reef productivity, that is, metric tons of reef-related fish lost on average for every square kilometer loss of coral reefs. The estimates are made separately for the low-, medium-, and high-risk categories, by region. These productivity estimates may then be used to convert projections of coral reef status into projections for reef-related fish supply by region. Computing economic welfare impact The estimated productivity changes and adoption rates, obtained from using the foregoing methods of eliciting expert opinion, are then converted into supply shifts over time. The supply shifts may then be incorporated into a baseline supply–demand model. The model used in this study is patterned after the

The estimated productivity changes and adoption rates are then converted into supply shifts over time, which may then be incorporated into a baseline supply–demand model

6

AsiaFish model (Dey et al, 2004). It is a multimarket equilibrium model of the fish sector, constructed for each region under analysis. The following provides a concise model description; details are available in Briones (2004). The model consists of equations on supply, demand, exports, and imports. The model data set is constructed along the lines of Delgado et al (2003), using FAO data. Proportional supply shifts, used to represent technological change and the impact of research, are incorporated by the formalism of the ‘effective price’ (Alston et al, 1995). Changes in economic welfare (a generalization of economic surplus) are evaluated using the dual approach, which is a modified version of the formula in Martin and Alston (1994). A baseline simulation is conducted over a 20-year period. Variables such as income and population are exogenous to the model; projections on income are proxied by estimates of average annual GDP growth by region to 2015, prepared by the World Bank (2004). Population projections are based on UN estimates by region (UN, 2004). Supply shifts under the business-as-usual scenario are incorporated into the baseline simulation. Supply shifts from the bestpractice scenario, modified by the adoption rates, are incorporated into the counterfactual simulation. The measure of research benefit is the welfare impact, which is the difference in welfare between baseline and counterfactual simulations. Estimation of cost is however constrained by unavailability of data on R&D investments by resource system and region. Instead, the cost of R&D for each assessment unit is approximated as some predetermined percentage of the total value of production under the counterfactual scenario. For farm commodity research, a common rule-of-thumb is that 1% of agricultural GDP needs to be allocated to the agricultural research and extension system to reach a typical range of benefits and rates of return to research (MacIntire, 1998; Roseboom, 2004). However investment in NRM research and applications are more difficult to estimate. For coastal and coral reef systems, Balmford et al (2004) report an estimate of US$ 5 to 19 billion annual cost of placing 20% to 30% of the world’s oceans under marine protected areas. This lies within 5% to 18% of the world’s annual value of production. For developed countries (where productivity increases in the fisheries is more likely), research-management expenditures range from 5% to 30% of the value of the catch (Arnason, 2000). This is consistent for data from OECD countries for 1999 (OECD, 2003), except for a few nations with low investments (Japan, Iceland, and Mexico). These costs are associated with hypothetical management regimes comparable to the best-practice scenario assumed in the expert opinion survey. For developing countries one may put a more conservative cost ratio of 5% for coastal and inland aquatic systems, further adjusted downwards by the estimated adoption rate for NRM research.

Research Evaluation December 2005

Natural resources management

Estimates of the cost of coral reef protection would tend towards the high end of management costs. For example, White, Ross, and Flores (2000) report a cost of USD 21,000 for a marine sanctuary in the Philippines, covering one square kilometer and producing US$ 42,000 of sustainable fisheries per year. For coral reefs the percentage is therefore set at 10%, likewise adjusted by the adoption rate for NRM research on coral reefs.

Results Estimates of resource productivity gains The expert opinion survey covered 42 respondents. Most (29) are from WorldFish, due to their familiarity with NRM research in developing countries. The 13 external respondents are composed of both fishery biologists and resource managers in both developing country and advanced research institutions. To compute supply shifts from the survey results, outliers were eliminated and several consistency assumptions were imposed, prior to calculating the average

estimate. For example, the business-as-usual scenario should not show progressive improvement over time; resource management may initially have a negative effect on productivity, but eventually it must have a positive and nondecreasing effect over time. The average of the estimates for marine capture is shown in Tables 5 and 6. The expected declines under the baseline scenario are seen to be modest over the first five years. Meanwhile, within five to ten years, production trends are expected to deteriorate, by as much as 7.5%. Experts tend to be most optimistic for the East Indian Ocean, but largely pessimistic over the other oceans (for example, the Pacific). Finally, in the long term (within 10 to 20 years), progressive worsening under business-as-usual is evident; the decline compared to the medium term is however less severe. Overall, for most ocean areas, experts conjecture a steep drop in annual production relative to the benchmark. Meanwhile, NRM makes a modest contribution over the short term (in the order of 5% or less over the benchmark). In a few cases, ideal management is seen to cause an immediate decline in production

Table 5. Estimates of business-as-usual production trends, by FAO ocean area and fish type (% above benchmark production)

Ocean area

Crustacean

Demersal

Mollusk

Pelagic

A. Years 1 to 5

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

–2.5 –2.5 –2.5 –2.5 0.0 0.8 –2.5 –2.5 –2.5 –2.5 –2.5 0.0 0.0

–2.5 –2.5 –5.0 –2.5 0.0 –3.8 –5.0 –2.5 –2.5 –2.5 –3.8 –2.5 –3.6

–2.5 –2.5 –1.3 –2.5 –2.5 0.6 –2.5 –2.5 –2.5 –2.5 0.6 0.0 –0.6

–2.5 –2.5 –5.0 –2.5 –1.7 –3.8 –2.5 –2.5 –2.5 –2.5 –2.5 –2.5 –1.1

B. Years 5 to 10

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

–6.3 –7.5 –7.5 –7.5 –2.5 –3.3 –7.5 –7.5 –7.5 –3.8 –2.5 –2.5 –3.3

–6.3 –7.5 –7.5 –5.0 –2.5 –5.6 –7.5 –7.5 –7.5 –5.0 –3.1 –5.0 –7.9

–6.3 –7.5 –3.8 –5.0 –7.5 –1.9 –5.0 –7.5 –7.5 –3.8 –3.8 –2.5 –3.1

–3.8 –2.5 –8.8 –2.5 –2.5 –3.8 –3.8 –2.5 –2.5 –2.5 –6.7 –2.5 –1.8

C. Years 10 to 20

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

–8.8 –7.5 –7.5 –7.5 –7.5 –3.3 –11.3 –7.5 –7.5 –3.8 –2.5 –3.8 –4.0

–8.8 –7.5 –7.5 –7.5 –8.8 –8.8 –11.3 –7.5 –7.5 –5.0 –6.3 –7.5 –14.2

–8.8 –3.8 –3.8 –7.5 –7.5 –7.5 –5.0 –7.5 –7.5 –3.8 –3.8 –3.8 –3.1

–8.8 –11.3 –11.3 –7.5 –7.5 –7.5 –6.3 –7.5 –7.5 –5.0 –12.5 –5.0 –8.3

Source: Authors’ survey

Research Evaluation December 2005

7

Natural resources management Table 6. Estimates of production trends under the ‘best-practice’ scenario, by FAO ocean area and fish type (% above benchmark production)

Ocean area

Crustacean

Demersal

Mollusk

Pelagic

A. Years 1 to 5

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

7.5 2.5 2.5 2.5 –2.5 2.5 1.2 2.5 2.5 0.8 1.7 5 0.8

7.5 2.5 –6.2 5 3.3 3.1 2.5 2.5 2.5 0.8 1.9 2.5 –1.0

7.5 2.5 2.5 5 8.8 4.4 1.2 2.5 2.5 0 0.8 2.5 2.5

7.5 2.5 –6.2 2.5 0 2.5 1.2 2.5 2.5 0.8 0 2.5 1.0

B. Years 5 to 10

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

13.8 2.5 2.5 2.5 –2.5 5.0 1.3 2.5 2.5 1.3 3.3 5.0 1.7

16.3 7.5 5.0 7.5 0.0 10.0 3.8 7.5 7.5 3.8 5.6 7.5 1.7

16.3 7.5 0.0 7.5 7.5 5.8 3.8 7.5 7.5 3.8 4.4 7.5 5.0

20.0 15.0 11.3 15.0 6.3 8.8 7.5 15.0 15.0 11.3 5.0 8.8 4.3

C. Years 11 to 20

East Central Atlantic Northeastern Atlantic Southeastern Atlantic Southwestern Atlantic West Central Atlantic East Indian West Indian Mediterranean and Black Sea East Central Pacific Northwestern Pacific Southeastern Pacific Southwestern Pacific West Central Pacific

28.8 7.5 7.5 7.5 0.0 11.3 7.5 7.5 7.5 3.8 5.0 11.3 5.6

32.5 15.0 22.5 15.0 3.8 17.5 15.0 15.0 15.0 7.5 7.5 22.5 5.0

28.8 7.5 0.0 7.5 7.5 7.5 7.5 7.5 7.5 3.8 4.4 11.3 9.4

40.0 30.0 30.0 30.0 13.8 18.8 30.0 30.0 30.0 16.3 4.2 18.8 10.4

Source: Authors’ survey

(probably as fishing pressure is reduced). In the medium term the deterioration of production (expected under the business-as-usual scenario) is almost completely reversed. The improvement is particularly sharp in the case of finfish. Interestingly, crustacean production on average is not expected to be appreciably helped by best-practice management. In the long term, ideal management leads to a range of improvement for every ocean area and fish type (except for mollusks in Southwestern Atlantic). The magnitude of the productivity gains is well below the shift implied by the global estimate of potential yield in Pauly (1996); the estimates may therefore be considered conservative, and probably more realistic given the 20-year horizon of the projections and the ecological lags of stock recovery. For inland capture, the expert survey obtained a similar pattern for the two scenarios (Table 7). Overall, the trend deterioration (under business-asusual) is less pronounced compared to the same scenario for marine capture. Apparently, the degree of excess fishing pressure and recoverability of fish stocks seem to offer worse prospects for marine 8

compared to inland capture. The worst deterioration is projected for East Asia, across all periods; the mildest is for Sub-Saharan Africa, with the longterm production decline of less than 6% relative to the benchmark. This may be attributed to the relatively low levels of exploitation of inland aquatic systems in this region, relative to other regions. Interestingly, Sub-Saharan Africa shows the second largest impact of resource management in the long run. Compared to marine capture, the impact of management intervention is greater. Some respondents explained that measures for increasing productivity, such as stock enhancement, are far more feasible for inland fisheries compared to their marine counterparts. Supply shifters (sans adoption modifiers) use this information, and allocated by a 67%:33% ratio for lakes and rivers, respectively (ICLARM, 1992, 1999). The exception is Sub-Saharan Africa, where the ratio is 50%:50% due to the relatively large contribution of the great lakes in southern Africa. The last set of shifters is computed for fisheries related to coral reefs. The relevant coral reef regions are South Asia, Southeast Asia, Sub-Saharan Africa, Research Evaluation December 2005

Natural resources management Table 7. Estimates of productivity trends and NRM shifters, inland capture, by scenario, time horizon, and region (% above benchmark production)

Scenario

Business as usual

Ideal management

Region

1 to 5

5 to 10

10 to 20

EA SA SEAI SEAM SSA WANA

–5.0 0.9 1.3 –1.4 –0.8 –2.5

–10.0 –3.3 –3.1 –6.1 –3.3 –6.7

–13.8 –5.8 –0.6 –10.7 –5.8 –7.5

1 to 5 4.4 3.8 4.4 4.3 4.2 3.5

5 to 10

10 to 20

10.6 10.5 16.3 11.4 11.1 8.5

13.8 13.8 17.5 13.2 17.2 11.0

Source: Authors’ survey

and the Middle East. Expert opinion projects a steep decline of the percentage of low-risk reefs within 20 years (Table 8), accompanied by a surge in the highrisk category. Meanwhile, ideally management does make a considerable difference, even over a relatively short period of time: a small to nonexistent extent of reefs in the high-risk zone are projected, while the low-risk category characterizes up to 65% of reefs in South Asia and the Middle East. However due to lags in the coral reef recovery, even with ideal management, less than 40% of coral reefs in Southeast Asia are classified under low risk. Table 8 also displays expert opinion on the quantity of fish per year produced, directly or indirectly, by these reefs per unit area. For relatively pristine coral reefs (that is, under ‘low risk’), the highest productivity is estimated for Southeast Asia.

tion. Gains are estimated to be very low, over the next ten years, for all regions and resource systems. In East Asia, owing to the policy environment, adoption of ideal management is seen to be very low across resource systems. Despite the low levels of development of institutions and governance in Sub-Saharan Africa, the highest potential adoption is projected for this region across all resource systems. It is matched in coral reefs and inland capture by Southeast Asia–Island, where institutions tend to be better developed and considerations of sustainability better articulated in the policy agenda. These figures are used to parameterize the logistic adoption function, which acts as the modifier for the NRM shifter under the bestpractice scenario. Impact on economic welfare

Adoption of best-practice management The average estimate based on expert opinion on the adoption of ideal management is shown in Table 9. Experts tend to be very circumspect in their projections about management practice and policy applicaTable 8. Estimates of coral reef risk status in 20 years, under alternative scenarios, in %, and reef productivity, in mt

South Asia

Southeast SubAsia Saharan Africa

Middle East

Low risk Business-asusual Best-practice

0.0 65.0

7.8 36.7

0.0 50.0

15.0 65.0

Medium risk Business-asusual Best-practice

20.0 20.0

24.0 30.8

20.0 40.0

46.5 30.0

High risk Business-asusual Best-practice

80.0 15.0

68.2 32.5

80.0 10.0

38.5 5.0

Productivity Low risk Medium risk High risk

25.8 16.7 8.3

32.5 21.0 11.6

18.8 12.5 5.0

25.8 16.7 8.3

Source: Authors’ survey

Research Evaluation December 2005

Applying the above shifters and other exogenous variables to the baseline model, one can compute the economic welfare impact from NRM research and its application. The relatively small production coverage of coral reef systems in East Asia and mainland Southeast Asia precludes the application of a price adjustment mechanism, as errors in model solution would tend to be large relative to the actual research impact. For these units, the counterfactual simulation assumes a production and price that is unaffected by supply shifts from research and application of NRM. Table 9. Estimates of adoption rates for best-practice management, in 10 years and ceiling, by developing region and resource system

Coast

EA SA SEAI SEAM SSA WANA

Coral

Inland capture

10 years

Ceiling

10 years

Ceiling

10 years

Ceiling

5.0 4.3 7.5 10.6 10.0 7.5

10.0 15.0 20.0 18.0 25.0 13.8

4.3 4.0 12.6 7.5 15.0 11.8

8.3 15.0 24.0 13.8 30.0 18.8

6.7 7.5 12.0 7.0 11.0 7.5

11.7 18.8 24.0 19.0 24.0 15.0

Source: Authors’ survey

9

Natural resources management

The economic welfare impact (under the Benefit column) and estimated costs are shown in Table 10. The figures correspond to the net present value over a 20-year period, using a discount rate of 5%, a level recommended by Alston et al (1995). Also shown in Table 10 are the relative benefit–cost ratios, which serve as the basis for ranking the R&D units in terms of net economic impact. Results indicate that coastal NRM in SEA–Island, SEA–Mainland, SSA, and coral reef NRM in SEA– Island generate the greatest gross benefits. However, coral reef systems in Sub-Saharan Africa, South Asia, and mainland Southeast Asia offer the highest benefit relative to cost. This holds despite the higher cost percentage assumed for coral reef systems, and may therefore be seen as a fairly robust feature in the impact ranking. Interestingly, the large coral reef system in island Southeast Asia ranks fifth, preceded by rivers in East Asia. Other inland water systems complete the top ten places; the first coastal system ranks 11th, for East Asia. Apparently, the larger systems are accompanied by lower supply shifts, relative to the price adjustment, which together vitiate the role of size in ranking the potential impact of research.

to conventional economic techniques of evaluation. This study demonstrates that, on the contrary, NRM research can be evaluated largely along conventional lines. The key is to articulate the impact pathway of NRM research, in a manner parallel to farm commodity research. With this framework, the usual tools of expert opinion and economic surplus analysis can be applied for computing measures such as the welfare impact, the cost of R&D, as well as the benefit–cost ratio. The method is applied to international research on aquatic resources, in the case of the WorldFish Center. The results indicate that research and application of NRM yields the highest net benefits for the smaller aquatic systems, particularly for coral reefs. The larger systems (particularly coasts) are assessed as a lower economic impact once price adjustments are incorporated into the evaluation of benefits and costs. Further research is needed on achieving greater accuracy in estimating supply shifts and adoption rates, particularly as the current set of assumptions is constrained by the sparseness of data, relevant case studies, and related information. This study may be seen as offering a workable method for quantitative economic studies of ex ante impact of NRM research.

Conclusion Notes

In recent years, NRM has gained prominence alongside farm commodities as a area of agricultural research. However the ex ante impact of NRM research is seldom evaluated, with the literature assessing this type of research as particularly resistant

Table 10. Benefit–cost ratio for NRM research, by region– resource system

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

SSA – Coral SA – Coral SEAM – Coral EA – Rivers SEAI – Coral SEAM – Rivers SA – Rivers EA – Lakes SSA – Lakes SSA – Rivers EA – Coast SEAI – Rivers SEAI – Coast SSA – Coast SA – Lakes WANA – Coast SEAI – Lakes SEAM – Lakes SA – Coast WANA – Rivers WANA – Lakes SEAM – Coast WANA – Coral EA – Coral

Benefit

Cost

BCR

157,733 396,959 132,935 31,057 1,313,675 85,930 81,517 39,093 206,078 206,078 559,247 33,440 1,340,028 1,207,402 111,136 463,866 51,770 107,363 739,313 149,707 205,008 1,312,545 24,574 20,683

23,075 67,731 23,959 11,358 510,756 44,304 43,138 23,074 121,747 121,747 349,555 21,148 880,224 857,771 87,627 382,117 43,053 90,047 648,271 137,868 280,845 2,010,982 48,531 226,186

100.00 85.74 81.17 40.00 37.63 28.37 27.64 24.78 24.76 24.76 23.40 23.13 22.27 20.59 18.55 17.76 17.59 17.44 16.68 15.89 10.68 9.55 7.41 1.34

Source: Authors’ calculations

10

1.

2.

3.

NRM also acts to regulate externalities from various forms of resource use, not necessarily confined to harvesting natural stocks. For example, farms may use the waste disposal services of natural systems, or expand production area by conversion of natural systems (for example, mangroves). Regulation of externality is a critical function of NRM; however this treats it as a separate quantification problem that lies outside the scope of this study. Research on innovative farming systems that promote longterm resource sustainability, such as IPM (integrated pest management) or IAA (integrated aquaculture-agriculture), have also been classified under NRM research. While the ambiguous nature of this research activity is recognized, this study classifies them under farm commodity research as the technologies generated are applied on divisible resource systems. Alternatively, one may calculate a discrete-time version of the logistic equation, consistent with the estimate based on years. However the continuous time formulation of equation (3) is also perfectly acceptable.

References J Alston, G Norton and P Pardey (1995), Science under Scarcity: Principles and Practice for Agricultural Research Evaluation and Priority Setting (Cornell University Press, Ithaca, USA). R Arnason (2000), ‘Costs of fisheries management: theoretical and practical implications’, 11th Biennial Conference of the International Institute for Fisheries Economics and Trade 2000 Proceedings. A Balmford, P Gravestock, N Hockley et al (2004), ‘The worldwide costs of marine protected areas’, Proceedings of the National Academy of Sciences, 101(26), pages 9694–9697. R Briones (2004), ‘Research priorities for the WorldFish Center’, Final Report submitted to the WorldFish Center, 3 December 2004. D Bryant, L Burke, J McManus et al (1998), Reefs at Risk: A Mapbased Indicator of Threats to the World’s Coral Reefs, World Resources Institute (WRI), ICLARM, World Conservation

Research Evaluation December 2005

Natural resources management Monitoring Centre, and United Nations Environment Programme. CGIAR (2000), ‘Impact assessment of agricultural research: context and state of the art’, ECART, ASARECA, CTA and GTZ Workshop on impact assessment of agricultural research in Eastern and Central Africa, 16–19 November 1999, Entebbe, Uganda. C Delgado, N Wada, M Rosegrant et al (2003), Fish to 2020: Supply and Demand in Changing Global Markets (International Food Policy Research Institute, Washington DC, USA, and WorldFish Center, Penang, Malaysia). M Dey, R Briones and M Ahmed (2004), ‘Disaggregated projections of fish supply, demand, and trade: baseline model and estimation strategy’, forthcoming in Aquaculture Economics and Management. M Dey, A Eknath, L Sifa et al (2000), ‘Performance and nature of genetically improved farmed tilapia: a bioeconomic analysis’, Aquaculture Economics and Management, 4(1), pages 83–106. Food and Agriculture Organization (2003), FAO Fishstat, downloaded from: , January 2004. ICLARM (1992), ICLARM’s Strategy for International Research on Living Aquatic Resources Management. ICLARM (1999), Aquatic Resources Research in Developing Countries: Data and Evaluation by Region and Resource System, Supplement to the ICLARM Strategic Plan 2000–2020, ICLARM Working Document 4. A-N Izac (1998), ‘Assessing the impact of research in natural resources management’, Synthesis of an international workshop, 27–29 April 1998, International Center for Research in Agroforestry, Nairobi, Kenya. T Kelly, J Ryan and B Patel (1995), ‘Applied participatory priority setting in international agricultural research: making trade-offs transparent and explicit’, Agricultural Systems, 49, pages 177–216. W Martin and J Alston (1994), ‘A dual approach to evaluating research benefits in the presence of trade distortions’, American Journal of Agricultural Economics, 76(1), pages 26–35. M Maredia, D Byerlee and J Anderson (2000), ‘Ex post evaluation of economic impacts of agricultural research programs: a tour of good practice’, paper presented at the workshop on ‘The Future of Impact Assessment in CGIAR: Needs, Constraints, and Options’, Standing Panel on Impact Assessment (SPIA) of the Technical Advisory Committee, 3–5 May, Rome, Italy. J McIntire (1998), ‘Coping with fiscal stress in developing-country agricultural research’, in S Tabor, W Janssen and H Bruneau

Research Evaluation December 2005

(editors), Financing Agricultural Research: A Sourcebook (International Service for National Agricultural Research, The Hague, Netherlands), pages 81–96. Organization for Economic Cooperation and Development (2004), The Costs of Managing Fisheries. (downloaded January 2004). E Ostrom, R Gardner and J Walker (1994), Rules, Games and Common-pool Resources (Westview Press, Boulder, Colorado, USA). D Pachico (1998), ‘Conceptual framework for natural resource management research and basic methodological issues in impact assessment’, paper presented at the international workshop on ‘Assessing the Impact of Research in Natural Resource Management’, 27–29 April, Nairobi, Kenya. D Pauly (1996), ‘One hundred million tonnes of fish, and fisheries research’, Fisheries Research, 25(1), pages 25–38. P Pingali (2001), Milestones in Impact Assessment Research in the CGIAR, 1970–1999. With an Annotated Bibliography of Impact Assessment Studies Conducted in the CGIAR, 1970– 1999, prepared by Matthew P Feldmann, Mexico, DF: Standing Panel on Impact Assessment, Technical Advisory Committee of the Consultative Group on International Agricultural Research. T Randolph, P Kristjanson, S Omamo et al (2001), ‘A framework for priority setting in international livestock research’, Research Evaluation, 10(3), December, pages 142–160. J Roseboom (2004), ‘Agricultural research and extension funding levels required to meet the Anti-Hunger Programme objectives’, paper written for the FAO, Rome, downloaded from <www.fao.org/sd/dim_kn4/docs/kn4_040401a1_en.pdf>, November 2004. A Springer-Heinze, F Hartwich, J Henderson et al (2003), ‘Impact pathway analysis: an approach to strengthening the impact orientation of agricultural research’, Agricultural Systems, 78(1), pages 267–285. UN (2004), World Urbanization Prospects: The 2003 Revision Population Database, United Nations Population Division, Economic and Social Affairs Department. World Bank (2004), Global Prospects Realizing the Development Promise of the Doha Agenda, Washington, DC (downloadable from <www.worldbank.org>). A White, M Ross and M Flores (2000), ‘Benefits and costs of coral reef and wetland management, Olango Island, Philippines’, in H Cesar (editor), Collected Essays on the Economics of Coral Reefs (CORDIO, Sweden), pages 215–227.

11