__RS I__1__ _3_P POLICY RESEARCH WORKING PAPER 1835 Cost-Benefit Analysis The Global Dracunculiasis Eradication Campaign has of the Global Dracunculiasis brought about a remarkable Eradication*Campaign decline in the incidence of Eradication Campaign dracunculiasis, or Guinea worm disease, in the Aehyung Kim countries in which it was Ajay Tandon endemic. Under conservative Ernesto Ruiz-Tiben assumptions about the average incapacitation caused by Guinea worm infection (five weeks), the campaign has shown an estimated economic rate of return of 29 percent. The World Bank Africa Human Development Department October 1997 POLICY RESEARCH WORKING PAPER 1835 Summary findings Dracunculiasis (or Guinea worm disease) was endemic in Using a project horizon of 1987-98, they calculate the several African countries as well as in India, Pakistan, economic rate of return as 29 percent, under and Yemen. The past decade, however, has seen a conservative assumptions about the average remarkable decline in the incidence of dracunculiasis as a incapacitation caused by Guinea worm infection (five result of the Global Dracunculiasis Eradication weeks). Campaign. After 1998, Sudan is projected to be the only country Kim, Tandon, and Ruiz-Tiben compare spending on in which the disease will remain endemic. The authors' the eradication campaign with estimates of increased results indicate that for economic returns in Sudan to be agricultural production resulting from reductions in consistent with those in other countries where the infection-related morbidity as a result of the program. disease was endemic, eradication must he completed in Sudan at the very latest by 2001. This paper - a product of the Africa Human Development Department - is part of a larger effort in the department to use economic analysis to improve the quality of health projects. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Aehyung Kim, room J10-046, telephone 202-473-5029, fax 202-522-3157, Internet address akimEtworldbank.org. October 1997. (16 pages) The Policy Research Working Paper Se7ies disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are iess than fully polished, The papers carry the names of the authors and should be cited accordinzgly. The findings, interpretations, and contclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center COST-BENEFIT ANALYSIS OF THE GLOBAL DRACUNCULIASIS ERADICATION CAMPAIGN (GDEC) by Aehyung Kim and Ajay Tandon' The World Bank and Emesto Ruiz-Tiben The Carter Center We would like to thank Pedro Belli, Jeffrey Hammer, Donald Hopkins and the participants of the consultation meeting at the Carter Center/Global 2000 for valuable comments. Introduction In 1986, there were over 2.25 million cases of dracunculiasis (Guinea worm disease) worldwide.2 Ten years later, in 1996, the estimated worldwide incidence of dracunculiasis was close to 330,000 cases.3 This remarkable decline in the incidence of dracunculiasis has been the result of the Global Dracunculiasis Eradication Campaign (GDEC). GDEC is spearheaded by national eradication programs and is supported by a coalition of agencies, institutions, organizations, and bilateral donors. As the numbers attest, as a result of GDEC, dracunculiasis has been virtually eliminated as a major public health problem and an impediment to socioeconomic development in several African countries as well as in India, Pakistan, and Yemen.4 Dracunculiasis was eradicated from Pakistan in 1994.5 In 1996, an overwhelming majority (78%) of a provisional total of 152,185 reported cases of dracunculiasis occurred in Sudan, where civil strife has complicated the successful operation of its national Guinea worm eradication program. Dracunculiasis is caused by the nematode parasite Dracunculus medinensis. An individual becomes infected by drinking water containing tiny crustaceans (copepods) which harbor infective larvae and act as an intermediate host. After about one year, the mature female worm approaches the skin, eventually causing a blister to form. The blister ruptures (within 24-48 hours) causing the head of the worm to become exposed to the external environment. When the affected person enters a body of water the worm releases myriads of larvae which are then ingested by copepods (water fleas). Upon release of the larvae, the worm dies and must be extracted manually, usually by rolling a few centimeters of the worm each day on a stick. The extraction of the worm is a slow and intensely painful process incapacitating individuals for weeks, and often longer. In addition to the blisters and skin lesions, secondary bacterial infections exacerbate local inflammation, and often lead to sepsis, abscesses, arthritis, contracture of joints, or even tetanus. In a small number of instances, infection may be associated with permanent crippling similar to polio. There is no cure for the disease, but it is amenable to prevention through educating people about the origin of the disease and what they can do to prevent it: filtering their drinking water through a piece of cloth to remove copepods, and never entering a source of drinking water (or allowing someone else to do it) when a Guinea worm is emerging. The disease can also be prevented by providing safe sources of 2 Watts (1987) estimated the worldwide incidence of dracunculiasis to be 3.3. million cases. Review of the available data, while preparing this paper, warranted that we revise that 1986 benchmark to 2.25 million cases. 3 This differs from the provisional number of cases reported to the World Health Organization (WHO) of 152,185. This discrepancy is largely caused by incomplete knowledge of the full extent of dracunculiasis and incomplete reporting of cases in Sudan. 4 The entire list of countries covered is: Benin, Burkina Faso, Cameroon, Chad, Cote d'Ivoire, Ethiopia, Ghana, India, Kenya, Mali, Mauritania, Niger, Nigeria, Pakistan, Senegal, Sudan, Togo, Uganda, and Yemen. See Hopkins et al. (I 995). 2 drinking water to affected comnmunities and by the application of the insecticide Abate (temephos) to selected unsafe sources of drinking water. The objective of this paper is to report on a cost-benefit analysis of GDEC: expenditure on GDEC activities is compared with the economic benefit resulting from the campaign. The emergence of the Guinea worm is a painful and debilitating process causing serious disability in infected individuals for weeks -- thereby seriously constraining their income-generating capacity. Therefore, the major benefit of GDEC is considered to be the prevention of this period of infection-related economic incapacitation.6 Based upon the resulting augmentation of agricultural production, this economic benefit is accorded a monetary value and then compared with the costs of the campaign. This "human capital" approach enables an assessment of the economic viability of large-scale public health programs such as GDEC which are demonstrably successful in controlling and/or eradicating disease incidence. The economic assessment of GDEC entails the following. First, we examine the strategies for eradication under the auspices of GDEC and the associated costs of the campaign. Second, we elaborate the projected benefits from the eradication program, focusing specifically on the prevention of economic incapacitation resulting from incidence decline. Finally, we contrast the costs with the projected economic benefits in order to estimate the economic returns of the campaign. Dracunculiasis Eradication: Strategies and Costs There are several characteristics of dracunculiasis transmission that make the disease amenable to eradication7: there is no human carrier state beyond the one-year incubation period; there is no known animal reservoir; transmission is seasonal; active detection of individuals with worms protruding from skin lesions is a sensitive means of assessing the presence of the disease in villages; and the methods for controlling transmission (as subsequently elaborated) are relatively simple. Beginning in 1986, 1987, and 1988, the Global 2000 Project of the Carter Presidential Center, in collaboration with the Center for Disease Control (CDC), assisted national Guinea worm eradication campaigns in Pakistan, Ghana, and Nigeria, respectively.8 The strategies for eradication in these three Other benefits -- difficult to quantify in economic terms -- include the alleviation of suffering and social stigmatization, improved school attendance, better child care, and elimination of infection- related expenditures. It is important to note that GDEC is an eradication -- and not a control -- campaign. Consequently, some of the operational requirements reflect the fact that the goal of an eradication campaign is to reduce the incidence of disease to zero in all affected communities by a certain target date. In contrast, the goal of a control campaign is usually the reduction of disease incidence in selected communities (e.g., those with the highest disease incidence) over a period of time to a level that is considered "tolerable" by public health authorities. India independently initiated its national Guinea worm eradication program in 1983. 3 countries later formed a model for Guinea worm eradication campaigns in other affected countries. Strategies for Dracunculiasis Eradication As elaborated in Hopkins and Ruiz-Tiben (1991), key activities of dracunculiasis eradication can be grouped into three operational phases. Phase I includes: the establishment of a national campaign coordinator, office/secretariat; completion of national baseline village-by-village surveys for case-detection and prevalence statistics; and preparation or revision of a national plan of action. Phase II incorporates implementation of interventions and includes activities such as: the identification and training of village-based workers for each endemic village; implementation of village- based surveillance using case registries; initiation of a comprehensive health education/mobilization strategy; provision of cloth filters (and training in their usage); provision of safe sources of drinking water (such as bore-hole wells) and control of copepod populations using Abate (temephos); and monitoring the coverage as well as quality of surveillance/interventions. Phase III involves implementation of case- containment strategies: as the expected case load of the village-based health worker becomes one (or below) case per worker per day, increasingly stringent surveillance and control measures are initiated -- shifting the focus from the village or community-level to the individual case-level. In addition, after Phase III, post-eradication activities include the maintenance of adequate surveillance for 3 years after the last case and a request to WHO to evaluate and certify the attainment of eradication.9 Target Population Coverage GDEC eradication activities progressively reached larger sections of the target population. The initiation of control interventions reached all target countries only in 1995. Table 2 reports the years in which there were increments in target population coverage, i.e. the years in which the various countries initiated intervention activities associated with the campaign. Table 2. Target Population Coverage. Year Countries Where Control Activities Initiated 1983 India 1987 Pakistan 1989 Ghana, Nigeria 1991 Cameroon 1992 Burkina Faso, Senegal, Togo, Uganda 1993 Benin, Cote d'lvoire, Mali, Mauritania, Niger 1994 Chad, Ethiopia, Kenya, Sudan 1995 Yemen 9 See Hopkins and Ruiz-Tiben (1991) for additional details regarding strategies for dracunculiasis eradication. 4 Measuring the Costs of Eradication The estimated expenditures of GDEC for the above-mentioned eradication activities include costs incurred by Global 2000, UNICEF, as well as WHO in the time horizon 1987-1996.10 Table 1 reports the annualized expenditure figures. In nominal terms, actual and projected GDEC expenditures in the time horizon 1987-1998 equal US$87.46 million in nominal terms and US$68.46 in 1987 constant dollars. Table 1. GDEC Expenditures (1987-1998a) in US$ millions Year Global 2000b Other Sourcesb Totalb Total Constant' 1987 1.26 0.23 1.49 1.49 1988 2.81 0.26 3.07 2.96 1989 3.12 0.30 3.42 3.16 1990 3.60 2.77 6.36 5.66 1991 5.84 2.12 7.95 6.80 1992 2.50 2.93 5.43 4.46 1993 7.68 8.95 16.63 13.15 1994 6.63 3.27 9.90 7.52 1995 8.66 3.39 12.06 8.81 1996 6.67 4.20 10.87 7.64 1997 5.60 5.60 3.78 1998 4.67 4.67 3.03 Total 59.04 87.46 68.46 a. Expenditures for 1997 and 1998 are projected estimates. b. In nominal termns. c. In 1987 constant US$. Methodology for Estimating Economic Benefits Through the prevention and elimination of infection-related incapacitation and debility among productive members of the population, the consequent additional output produced is assumed to be the primary benefit of GDEC."I Several studies have underscored the adverse economic impact of dracunculiasis. For instance, Belcher et al. (1975) estimated that, in southern Ghana, adult male farmers were at greatest risk of contracting the disease and, on average, untreated adults lost 5 weeks of work time. A review of twelve published studies yielded an average duration of disability caused by Guinea worm disease to be about 8 weeks (estimates ranging from 2 to 16 weeks). Furthermore, the economic impact of dracunculiasis is aggravated by the seasonality of infection -- coinciding with peak agricultural activities -- when few alternatives exist to alleviate 10 Expenditures borne by national governments for campaign implementation and control activities are not considered in this analysis. Information regarding national campaign expenditures is difficult to obtain and, in any case, expected to be modest (in dollar terms) in comparison to total annual expenditures of GDEC from other sources such as Global 2000, UNICEF, and WHO. This is proposed as a lower-bound estimate of the economic returns accruing from GDEC. We are ignoring, for instance, the benefits in terms of reduced infection-related expenditures among cases as well as positive effects on savings and income in the long run. 5 12 labor shortages caused by infection, especially so in endemic communities. Since GDEC activities have been typically prevalent in rural areas, the marginal product of labor is estimated as the additional output gained in the agricultural sector through increasing the supply of productive labor input.'3 In order to estimate the additional output of labor accrued through prevention of Guinea worm infection, methods based on the Cobb-Douglas production function were used.'4 The following parameters are considered relevant in estimating the additional output of labor in the agricultural sector: the number of cases of Guinea worm infection prevented by GDEC (the difference between the number of cases of Guinea worm without GDEC and with GDEC) in the productive age group, the productivity loss -- in terms of percent of work-year lost -- per case of dracunculiasis, the output elasticity of labor in the agricultural sector, and the length of the effective project horizon. As in the analysis of costs, output figures are measured in 1987 constant dollars for consistency of comparison between costs and benefits. The following subsections describe how each of these parameters was estimated. Cases of Guinea Worm without GDEC Global 2000 estimates that in 1986 (i.e., prior to the initiation of GDEC activities) the annual worldwide incidence of dracunculiasis was in the vicinity of 2.2 million cases. For the benchmark case we assume that in the absence of GDEC, in the very least, the annual incidence would have remained at this level.'5 12 See Shulman (1985) for additional details regarding the coincidence of peak labor input demand with the dracunculiasis cycle and the precipitation cycle. In our paper we assume a work year of 40 weeks. 13 We are assuming unemployment is not a major factor in the analysis. Several surveys of labor markets in Sub-Saharan Africa have indicated that unemployment is largely an urban phenomenon. In fact, as Vandemoortele (1991) indicates, unemployment in the region increases with educational and skill level. The rural labor sector (on which this study exclusively focuses) primarily comprises unskilled workers (with relatively low levels of education) as well as subsistence farmers. Therefore, increases in productive labor time are expected to result in the augmentation of agricultural output. i This methodology follows that in Kim and Benton (1995) and Kim et al. (1997). The Cobb-Douglas function relates output Y to inputs E (effective labor input in efficiency units) and K (land/capital) in the following manner: Y = AErKb, where A is a technology parameter. The parameters a and b represent the elasticity of output with respect to inputs: a, for example, is the percentage increase in output caused by a unit percentage increase in labor input. Also, a represents the labor share of output. Similarly for b. Constant returns to scale can be represented by the assumption a+b = 1, which means that if both land and labor inputs increased by a certain proportion, then output would also increase by that same proportion: for example, a doubling of both labor and land inputs would yield double the output. 15 In a subsequent section, we assess the sensitivity of our analysis to increases as well as decreases in the projected annual incidence of dracunculiasis over time in the absence of GDEC. 6 Number of Cases of Guinea Worm Prevented in Productive Age Group Based upon epidemiological studies, the estimated total number of cases of Guinea worm in the period 1986-1996 is known. A sharp reduction in the number of cases of Guinea worm infection as a result of GDEC activities has been observed. By 1996, only 329,521 cases (including those in Sudan) were estimated to have occurred." Global 2000 projections indicate that, by 1998, it is likely that dracunculiasis could be eradicated from all countries except for Sudan. The total number of cases of Guinea worm infection prevented as a result of GDEC in any given year is computed as the difference between the projected number of cases of Guinea worm without GDEC and the number of cases with GDEC. We make the following assumptions regarding the number of cases of Guinea worm infection: (a) in the absence of GDEC, the total annual number of cases of Guinea worm would have remained at 1986 levels (around 2.2 million every year in the entire region), (b) two-thirds of all cases reported (or estimated) represent those among the "economically productive" age group of 15-44, and, as mentioned in a previous subsection. Table 3 reports the total number of cases of Guinea worm infection prevented in a given year, based upon the above-mentioned information and assumptions. Table 3. Number of Cases of Dracunculiasis (1986-1998). Total Cases of Total Cases of Cases Prevented Year Dracunculiasis Dracunculiasis In Productive (with GDEC) Preventeda Age Groupb 1987 2,178,272 0 0 1988 2,118,368 59,904 39,936 1989 2,061,543 116,729 77,819 1990 1,523,540 654,732 436,488 1991 1,202,351 975,921 650,614 1992 997,016 1,181,256 787,504 1993 1,118,829 1,059,443 706,295 1994 521,365 1,656,907 1,104,605 1995 408,742 1,769,530 1,179,687 1996 329,521 1,848,751 1,232,501 1997 315,000 1,863,272 1,242,181 1998 300,000 1,878,272 1,252,181 a. Assuming the annual incidence without GDEC in the productive age group is 1,452,181 out of 2,178,272 for all age groups. b. Assuming cases in productive age group represent two-thirds of total cases. As can be seen from Table 3, by 1998, as a result of GDEC, an estimated total of almost 13 and 9 million cases of dracunculiasis would have been prevented in total and in the productive age group, respectively. This implies a unit cost of about $5 and $8 per case prevented in 1997 constant dollars for all age group and in productive group, respectively over the life of the program. 16 This is a Global 2000 estimate. 7 Productivity Loss Per Case It is important to note that the benefits accruing from the availability of more productive labor days (as a result of reduction in the number of cases of Guinea worm) is not easily estimated. Since transmission occurs seasonally, usually coinciding with the period of peak agricultural labor demand, there is a significant adverse impact on agricultural productivity. The methodology for computing the enhanced productivity of labor as a result of GDEC is as follows. We assume that production of agricultural output is represented in the form of a Cobb-Douglas production function: Y = AEaKb.l7 Denote E as the effective labor force (in efficiency units 8) without GDEC. One non-infected worker is assumed to supply one year of productive labor in any given year. An infected worker, due to disease-related incapacitation, is assumed to provide less than one year of productive labor in any given year.19 The prevention of the loss in productive labor time due to Guinea worm infection is the estimated benefit accruing from GDEC. Every case of dracunculiasis prevented is projected to add k worker-years of productive labor input time in a given year. This degree of incapacitation k may be expressed in terms of the percentage of labor days per year not worked as a result of infection. As mentioned earlier, a review of twelve published studies yielded an average duration of disability from infection of about 8 weeks, with estimates ranging from 2 to 20 16 weeks. This implies that over 20% of annual work-time is lost per case of infection. In order to be conservative, we assume that on average 5 weeks of production time is lost per case of Guinea worm: a loss of 12.5% of annual work time. We elect a conservative estimate as a benchmark to allow for the possibility that other members belonging to the households of infected individuals may pick up some of the slack in income generation during the period of disability. Furthermore, there may be attempts at long-term 17 See footnote 13 for additional details. 18 We are measuring the labor force E in terms of the labor service inputs (i.e., time worked) instead of in terms of natural units (i.e., number of workers). 19 The framework is similar to that in Cuddington (1993). The effective labor force, in the absence of GDEC, may therefore be decomposed as the sum of the labor input supplied by infected and non- infected individuals in any given year. If z is the proportion of the labor force L that is infected in any given year then the effective labor force E may be characterized as: E = (1-z) L + z (1-k) L, where (1- z) L is the labor supplied by non-infected workers and z (1-k) L is the labor supply of infected workers; k (taking a value between 0 and 1) represents the fraction of work year lost per Guinea worm-stricken worker. In the presence of GDEC, the effective labor force EG is similarly given by: EG = (I-ZG) L + ZG (1-k) L, where ZG is the fraction of the labor force infected with dracunculiasis in a given year. With GDEC, ZG is progressively lower than z. The incremental labor input available as a result of GDEC is given by (EG - E) or k(z- zG)L, where (z- zG)L is simply the number of cases of Guinea worm prevented due to GDEC in the productive segment of the population. Therefore, (z- zdL times k gives us the additional productive labor input available as a result of reducing the incidence of dracunculiasis. 20 This is assuming a 40 week work-year, given the seasonality of agricultural production and its coincidence with disease transmission. 8 intertemporal coping such that infected individuals are later able to make up (to some extent) for lost production time. In a later section, we assess the sensitivity of our results to variations in this parameter. In order to estimate the additional output produced, the percentage increase in the labor force for a given year due to GDEC is then estimated by dividing the incremental productive labor input time available by the rural labor force in the GDEC area. Output Elasticity of Labor The percentage increase in agricultural output available as a result of increased labor input time stemming from reduced infection-related incapacitation is calculated assuming an output elasticity of 0.66 in GDEC countries. This means a 1% increase in labor input 21-addi will increase output by 0.66%. Output is measured as the agricultural value-added in GDEC countries. To estimate the value of the additional output attributable to the increased labor input available as a result of GDEC, the agricultural value-added is multiplied by the computed percentage increase in output (i.e., benefits equal 0.66 times the percentage increase in labor input times the agricultural value-added). Project Horizon We assume, for the benchmark case, that the campaign has a 12 year project horizon: from 1987 through 1998.22 In a later section, we assess the sensitivity of our results to changes in the project horizon concerning operations in Sudan.23 Economic Returns from Dracunculiasis Eradication Based upon the above-mentioned methodology, a cost-benefit analysis was conducted. Table 4 reports the Net Present Value (NPV)24 and Economic Rate of Return (ERR)25 for GDEC. This follows from Kim and Benton (1995) and Over (1992). Cuddington (1993) assumes a value of 0.7 in his study of Tanzania. 22 Benefits beyond the project horizon are not considered in this analysis (benefits in later years, in any case, are progressively smaller due to discounting). Hence, our results may be taken to represent a lower bound to the projected economic returns accruing from the eradication campaign. The major unknown remains the length of time (beyond 1998) required to eradicate the disease in Sudan, where civil strife has precluded the successful operation of control activities. 24 The Net Present Value (NPV) is the present value of the stream of net returns (benefits minus costs) of a project during its economic life. The present value is determined by discounting expected future net returns at a rate which reflects opportunity costs and/or time preferences. A relatively high positive NPV is taken as an indication of the economic viability of a project. 9 Table 4. Net Present Value (NPV) and Economic Rates of Return (ERR) of GDEC. Project Horizon: 1987-1998. Percent of Work-Year Lost Per Case of Dracunculiasis: 12.5% (5 weeks out of 40). Discount Rate Net Present Value (NPV) 3% $7,235,744 10% $6,257,928 Economnic Rate of Return (ERR): 29% As can be seen from Table 4, even under relatively conservative assumptions of infection-related disability and limiting the calculation of benefits to the productive age group, GDEC appears to represent an efficacious use of resources. The ERR is very respectable at 29% and the NPV at a 10% discount rate is almost US$6.3 million in constant dollars. These economic returns compare very favorably with those from other health-sector projects. Typically, institutions such as the World Bank consider ERRs in excess of 10% to represent sound economic investments in "productive" sectors such as transport, energy, and agriculture.26 Switching Value We calculate the switching value (V) which is the value of a case prevented that would make the ERR equal 10%. In other terms, V times the present value (at 10% discount) of prevented cases equals the present value of costs (also at a 10% discount). Using this formulation, we find V to equal about US$10 in constant dollars. Therefore, if the value of a case prevented is at least US$ 10 in the productive age group, then the intervention can be viewed as being sound in economic terms with an ERR of 10%. Sensitivity Analysis The above analysis is based upon several crucial assumptions regarding: (a) the average degree of incapacitation caused by Guinea worm infection (5 weeks), (b) the projected annual incidence of Guinea worm in the absence of GDEC (2.2 million every year), and (c) the project horizon (1987-1998). In this section, we determine the sensitivity of our results to variances in these estimates. 25 The Economic Rate of Return (ERR) is that discount rate which makes the present value of the stream of net returns (benefits minus costs) equal to zero. If the ERR is greater than the cost of borrowing funds, the project is usually considered to be worth undertaking from an economic point of view. 26 Social sectors such as health and education are evaluated against lower "social" discount rates, reflecting the difficulties in quantifying the broader economy-wide benefits (such as benefits to future generations from disease eradication) accruing from investments in such sectors. 10 Productivity Loss Per Case We assume a range of values against the benchmark assumption of 5 weeks loss per case in order to assess the sensitivity of our results. As mentioned earlier, the average degree of incapacitation caused by Guinea worm infection -- as chronicled several studies -- is about 8 weeks. Thus, our benchmark assumption of 5 weeks of work-time loss is quite conservative in that it is more representative of the lower bound estimate of the dracunculiasis-related duration of disability reported in most studies. Table 5 reports the results of a sensitivity analysis with regard to changes in this parameter. As can be seen, the results are quite sensitive. In any case, even at a relatively low estimate of a 4 week loss of productivity per case of infection, the NPVs are substantial. Table 5. Sensitivity of NPV and ERR: Productivity Loss Per Case Project Horizon: 1987-1998. Productivity Loss 4 weeks NPV @3% $4,927,370 @10% $4,233,999 ERR 11% 5 weeks NPV (@3% S7.235,744 @10% $6,257,928 ERR 29% 6 weeks NPV @3% $9,544,119 @10% $8,281,857 ERR 44% Annual Incidence without GDEC For our benchmark analysis, we have assumed that in the absence of GDEC the annual incidence of Guinea worm would remain at the level of approximately 2.2 million over the course of the time horizon in our study. We now assess the sensitivity of our results to this assumption. We assume growth rates of -1.5% and 1.5% -- in addition to the benchmark of 0% -- in the annual incidence of dracunculiasis in the absence of GDEC.27 Values of other parameters are the same as in the benchmark. Table 6 reports the results. As can be seen, the ERRs again are quite sensitive to this parameter. Not surprisingly, any projected increases in dracunculiasis infection (due to population growth, for instance) significantly augment the economic returns from the program. 7 A negative growth in annual incidence growth is also assumed since, arguably, general socio- economic development can sometimes lead to incidence declines even in the absence of any intervention. 11 Table 6. Sensitivity of NPV and ERR: Annual Incidence without GDEC Project Horizon: 1987-1998. Percent of Work-Year Lost Per Case: 12.5% (5 weeks out of 40). Growth in Annual Incidence -1.5% NPV @3% $5,185,570 @10% $4,460,382 ERR 15% 0% NPV @,3% $7,235,744 @10% 56,257,928 ERR 29% 1.5% NPV @3% $9.617,654 @10% $8,346,331 ERR 43% Project Horizon: The Case of Sudan So far in our analysis we have assumed the project horizon to be 1987 through 1998. However, by the end of 1998, the disease is projected to be eradicated from all countries except for Sudan, given the precarious political situation there. In this subsection, we focus specifically on the eradication program in Sudan. Assuming it takes anywhere from 3 to 5 to 7 years to ensure eradication there beyond 1998, we derive the economic returns specifically for that country. Table 7 summarizes the results.28 Table 7. Sensitivity of NPV and ERR: Project Horizon Percent of Work-Year Lost Per Case: 12.5% (5 weeks out of 40). Project Horizon for Eradication in Sudan 3 years (1999-2001) NPV @3% $999,971 @10% $696,083 ERR 30% 5 years (1 999-2003) NPV @3% $83,245 @10% -$107,681 ERR 10% 7 years (1999-2005) NPV @3% -$716,339 @10% -$808,738 ERR 0% According to Table 7, the economic returns from eradication in Sudan are consistent with the returns from the campaign for other countries as long as eradication is achieved within 3 years beyond 1998 (i.e., by the year 2001). The longer it takes for eradication efforts to be successful, the lower are the projected economic returns -- going down to 0% by 2005, and negative thereafter. 28 Annual costs for eradication in Sudan were assumed to be between 3 to 3.5 million in real terms. 12 Conclusions This study has assessed the economic efficacy of GDEC in terms of comparing the costs of the campaign to the benefits that have resulted from the decline in the incidence of dracunculiasis. We have developed a methodology by which we are able to estimate the extent to which additional output is produced as a result of the prevention of infection- related economic incapacitation. It has been widely reported in the literature that a case of dracunculiasis results in some degree of work time lost. As a result of GDEC, this loss in productive potential is prevented. The additional output produced as a result -- derived using techniques that involve the use of the Cobb-Douglas production function -- represent the quantifiable economic benefits of the campaign. Based upon this methodology, we determine that the economic returns of GDEC are sensitive to assumptions regarding the average degree of economic incapacitation, in terms of percent of work year lost, caused by a case of dracunculiasis. Nevertheless, even under conservative assumptions we find highly respectable ERRs that range from 11% to 44%, depending upon whether an average case leads to 4 weeks or 6 weeks of economic incapacitation. A sensitivity analysis was also conducted to assess the variance caused by changes in key parameters. One key implication of the analysis is that eradication efforts in Sudan must be completed as soon as feasible in order for the economic returns there to be consistent with those obtained in other GDEC countries. 13 Appendix This appendix elaborates on the methodology utilized for the estimation of the additional output produced as a result of an augmentation of productive labor time available from dracunculiasis eradication. Basically, we are interested in estimating the marginal product of labor, or how much additional agricultural output (Y) is available from additional labor time (E), where labor input is measured in efficiency units. Assuming agricultural production may be characterized by a Cobb-Douglas production function of the form: Y = A EaKb where Y is agricultural value-added; A is a technology parameter; E is labor input in time (equals total working population for a healthy labor force); K is land/capital; and a and b are the elasticities of output with respect to the two inputs. Hence, we have: a = (dY/Y)/(dE/E) or, dYY= a (dE/E) Y The additional output (dY) available from disease control in any given year can be approximated by the output elasticity of labor a (assumed equal to 0.66) times the percentage increase in labor time (dEIE) times the agricultural value-added (Y). 14 References Adeyeba, O.A. and O.O. Kale (1991), "Epidemiology of Dracunculiasis and Its Socio- economic Impact in a Village in South-west Nigeria," West African Journal of Medicine 10(3):208-215. Aron, J.L. and P. Davis (1993), "A Comparative Review of the Economic Impact of Selected Infectious Diseases in Africa," USAID Technical Paper No. 2. Belcher, D.W., F.K. Wurapa, W.B. Ward, and I.M. Lourie (1975), "Guinea Worm in Southern Ghana: Its Epidemiology and Impact on Agricultural Productivity," American Journal of Tropical Medicine and Hygiene 24:243-249. Brieger, W.R. and J. Guyer (1990), "Farrners' Loss Due to Guinea Worm Disease: A Pilot Study," Journal of Tropical Medicine and Hygiene 93:106-111. Carter Center (1995), Let Not the Sinuous Worm Strike Me. Chippaux, J.P., A. Banzou, and K. Agbede (1992), "Social and Economic Impact of Dracunculiasis," Bulletin of the World Health Organization 70(1):73-82. Cuddington, J.T. (1993), "Modeling the Macroeconomic Effects of AIDS with an Application to Tanzania," World Bank Economic Review 7(2): 173-190. Edungbola, L.D. (1983), "Babana Parasitic Disease Project II: Prevalence and Impact of Dracontiasis in Babana District, Kwara State, Nigeria," Transactions of the Royal Society of Tropical Medicine and Hygiene 77(3):310-315. Feachem, R.G.A., T. Kjellstrom, C.J.L. Murray, M. Over, and M.A. Phillips (1992), The Health ofAdults in the Developing World, New York: Oxford University Press. Hopkins, D.R. and E. Ruiz-Tiben (1991), "Strategies for Dracunculiasis Eradication," Bulletin of the World Health Organization 69(5): 533-540. Hopkins D.R., M. Azam, E. Ruiz-Tiben, and K.D. Kappus (1995), "Eradication of Dracunculiasis from Pakistan," The Lancet 346(8975):621-624. Ilegbodu, V.A., A.E. Ilegbodu, R.A. Wise, B.L. Christiansen, and O.O. Kale (1991), "Clinical Manifestations, Disability and Use of Folk Medicine in Dracunculiasis in Nigeria," Journal of Tropical Medicine and Hygiene 94:35-41. Kale, O.O. (1977), "The Clinico-Epidemiological Profile of Guinea Worm in the Ibadan District of Nigeria," American Journal of Tropical Medicine and Hygiene 26(2):208-214. Khan, H.D., M. Aminuddin, and C.H. Shah (1986), "Epidemiology and Socio- Economic Implications of Dracunculiasis in Eleven Rural Communities of District Bannu (Pakistan)," Journal of the Pakistani Medical Association 36(9):233-239. Kim, A. and B. Benton (1995), "Cost-Benefit Analysis of the Onchocerciasis Control Program (OCP)," World Bank Technical Paper No. 282. Kim, A., B. Benton, and A. Tandon (1997), "Economic Assessment of the African Programme for Onchocerciasis Control (APOC)," Mimeo, The World Bank, Washington, DC. 15 Nwosu, A.B.C., E.O. Ifezulike, and A.O. Anya (1982), "Endemic Dracontiasis in Anambra State of Nigeria: Geographical Distribution, Clinical Features, Epidemiology and Socio-Economic Impact of the Disease," Annals of Tropical Medicine and Parasitology 76(2): 187-200. Over, M. (1992), "The Macroeconomic Impact of AIDS in Sub-Saharan Africa," World Bank Technical Paper No. 3. Over, M., R.P. Ellis, J.H. Huber, and 0. Solon (1992), "The Consequences of Adult Ill-Health." In Feachem, R.G.A. et al. (1992). Rohde, J.E., B.L. Sharma, H. Patto, C. Deegan, and J.M. Sherry (1993), "Surgical Extraction of Guinea Wormn: Disability Reduction and Contribution to Disease Control," American Journal of Tropical Medicine and Hygiene 48(1):71-76. Ruiz-Tiben, E. (1992), "Dracunculiasis and Its Impact on Agriculture," Global 2000, Inc., The Carter Center. Shulman, M.H. (1982), "On Correlations Between Dracunculiasis and Malnutrition in West Africa," Workshop on Opportunities for Control of Dracunculiasis, Washington, DC. Smith, G.S., S.R.A. Huttly, D. Blum, N. Okeke, B.R. Kirkwood, and R.G. Feachem (1989), "Disability from Dracunculiasis: Effect on Mobility," Annals of Tropical Medicine andParasitology 83(2):151-158. Suleiman, M.M. and K. Abdullahi (1989-90), "Guinea Worm Infection in Kiri-Manai Village, Sokoto," The Nigerian Journal of Parasitology 9-11(13-16). Vandemoortele, J. (1991), "Employment Issues in Sub-Saharan Africa," Africa Economic Research Consortium Special Paper No. 14. Watts, S.J. (1987), "Dracunculiasis in Africa in 1986: Its Geographic Extent, Incidence, and At-Risk Population," American Journal of Tropical Medicine and Hygiene 37(l):119-125. Watts, S.J. (1989), "Guinea Worm: An In Depth Study of What Happens to Mothers, Families, and Communities," Social Science and Medicine 29(9):1043-1049. World Health Organization (1996), Weekly Epidemiological Record, 10 May. 16 Policy Research Working Paper Series Contact Title Author Date for paper WPS1815 Unfair Trade? Empirical Evidence in Jacques Morisset August 1997 N. Busjeet World Commodity Markets Over 33997 the Past 25 Years WPS1816 Returns to Regionalism: An Raquel Fernandez August 1997 J. Ngaine Evaluation of Nontraditional Gains 37947 from Regional Trade Agreements WPS1817 Should Core Labor Standards Be Keith E. Maskus August 1997 J. Ngaine Imposed through International Trade 37947 Policy? WPS1818 What Affects the Russian Regional Lev Freinkman August 1997 N. Campos Governments' Propensity to Michael Haney 38541 Subsidize? WPS1819 The Argentine Pension Reform and Dimitri Vittas August 1997 P. Infante Its Relevance for Eastern Europe 37642 WPS1820 Private Pension Funds in Argentina's Dimitri Vittas August 1997 P. Infante New Integrated Pension System 37642 WPS1821 The 'IPO-Plus': A New Approach to Itzhak Goldberg August 1997 I. Goldberg Privatization Gregory Jedrzejczak 36289 Michael Fuchs WPS1822 Intergovernmental Fiscal Transfers Jun Ma September 1997 C. Ima in Nine Countries: Lessons for 35856 Developing Countries WPS1823 Antidumping in Law and Practice Raj Krishna September 1997 A. Bobbio 81518 WPS1824 Winners and Losers from Utility Omar Chisari September 1997 T. Malone Privatization in Argentina: Lessons Antonio Estache 37198 from a General Equilibrium Model Carlos Romero WPS1825 Current Accounts in Debtor and Aart Kraay September 1997 R. Martin Creditor Countries Jaume Ventura 39026 WPS1826 Standards and Conformity Sherry M. Stephenson September 1997 M. Patena Assessment as Nontariff 39515 Barriers to Trade WPS1 827 The Determinants of Agricultural Yair Mundlak September 1997 P. Kokila Production: A Cross-Country Don Larson 33716 Analysis Ritz Butzer Policy Research Working Paper Series Contact Title Author Date for paper WPS1828 The Determinants of Banking Crises: Asli Demirguc-Kunt September 1997 P. Sintim-Aboagye Evidence from Developed and Enrica Detragiache 38526 Developing Countries WPS1829 Economic Reform and progress in Norman Loayza September 1997 E. Khine Latin America and the Caribbean Luisa Palacios 37471 WPS1830 Private Ownership and Corporate Roman Frydman September 1997 B. Moore Performance: Some Lessons from Cheryl W. Gray 38526 Transition Economies Marek Hessel Andrzej Rapaczynski WPS1831 How Trade PaKterns and Technology Wolfgang Keller September 1997 J. Ngaine Flows Affect Productivity Growth 37947 WPS1832 Pension Reform in Bolivia: Innovative Hermann von Gersdorff September 1997 C. Pavlak Solutions to Common Problems 82099 WPS1833 Cost Recovery and Pricing of David B. Humphrey October 1997 T. Ishibe Payment Services Robert H. Keppler 38968 Fernando Montes-Negret WPS1834 The Comparative Advantage of Pedro Belli October 1997 L. Schunk Government: A Review 31779