Document(s) 5 of 10

The previous chapter provided a qualitative description of the damages suffered by people and ecosystems in the 1997 fires and haze. The Economy and Environment Program for Southeast Asia (EEPSEA) and the World Wide Fund for Nature (WWF) have attempted to calculate the monetary value of some of these damages, using a variety of methods developed in recent years by environmental economists (Freeman 1994).

In estimating the monetary value of damages from the fires and haze, this chapter was guided by several general considerations:

• Local damages were normally calculated in local currency, and converted to U.S. dollars at the following (July/August 1997) exchange rates: US$1 = S$1.4; US$1 = Rp2,500; US$1 = RM2.5.
  
  
• Damages were calculated in "present value terms" (i.e., many losses occur one time only; others recur. Income streams or environmental services that could provide recurring benefits were converted to a one-time only [present value] equivalent).
  
  
• Damages were calculated in net terms (i.e., damages are net benefits foregone. Net benefit equals gross value of the foregone good or service minus the cost of producing or extracting it. This is equivalent to value added; or profit minus normal rate of return to capital; or economic rent).
  
  
• The study attempted to approximate consumer (or producer) surplus foregone, rather than actual expenditures on prevention mitigation. The latter may substantially understate damages (e.g., in the case of health damages, some people were able to obtain medical treatment or evacuate an affected area. Other people were similarly affected but were unable to do so. Actual expenditures for treatment were therefore extrapolated to the entire affected population).
  
  
• Valuation is not appropriate or adequate for depicting the significance of some damages. For example, the magnitude of damage is felt relative to the ability to bear its cost: wealthy people can sustain larger losses than very poor people, so dollar figures are not necessarily a good measure of suffering. Valuing loss of life is difficult and controversial. In this chapter, we assume that such losses are significant but incalculable. We were also unable to place values on such things as increased risk of haze-induced illness (e.g., cancer) in the future or increased risk of species extinction.

In cases where it was not feasible to conduct new surveys, the benefit transfer (BT) approach was used. This involves the transfer of values from existing studies to the new study site, with appropriate adjustments for the size of the affected area, income levels, and other factors. Various BT values and other adjustment factors are mentioned below. These are derived from various sources, including the World Bank and the Asian Development Bank (ADB 1996). Wherever possible, BT values were "reality-checked" against local conditions.

For each type of damage, there are considerable uncertainties about both its physical extent and monetary value. In some cases, our estimation procedures produced a range of estimates. However, given the large number of impacts to be valued, many of them comprising numerous sub-components, expressing estimates as ranges was unworkable. Instead, the mid-points of ranges were used, to permit aggregation and a comprehensible presentation.

The methods used to estimate the monetary value of damages are described in step-by-step format below.

ESTIMATION METHODS FOR HAZE DAMAGES

This section outlines a common methodology prescribed for the three country studies. Methods were adapted to local conditions and data availability in each country during application; the adjustments are described in detail in the country reports in succeeding chapters.

The period covered was 1 August to 31 October 1997. In principle, the study should compare the situation with and without haze. In practice this involved a comparison of August–October 1997 to a "normal" August–October. The "normal" values used here were either:

• August–October 1996;

• average of August–October over the past five years; or

• the projected trend of August–October over the past five years, depending on what was most appropriate in a given case.

In a "normal" year, forest fires and haze still occur, though on a much smaller scale, and do result in damages to people and the environment. This chapter only estimates the damages in excess of "normal" damages; in this respect it understates the total damages that occurred in 1997.

Care was taken to separate the effects of the haze from those of the drought and the Asian financial crisis.

SHORT-TERM HEALTH COSTS: ADJUSTED "COST-OF-ILLNESS" APPROACH

The three steps used to obtain an adjusted cost of illness (COI) are outlined below:

a. Estimate Treatment Cost

1. Estimate hospital and clinic admissions for haze-related ailments per 10,000 population for August–October 1997. Use "haze-related ailments" as defined by each country's health service. If there is no such definition, use the Malaysian definition: upper respiratory tract infections (URTI), asthma, bronchitis, and conjunctivitis.
   
   
2. Estimate the hospital and clinic admissions for August–October 1996 or the average hospital and clinic admissions during August–October over the previous five years.
   
   
3. Subtract (ii) from (i) to get "excess" admissions.
   
   
4. Adjust for affected but untreated population. The ratio of untreated to treated case varies from country to country but is in the range of 3 or 4 to 1. The ratio for each country can be found in standard health sector studies by the World Blank or the Asian Development Bank.
   
   
5. Adjust for treatment costs beyond hospital visits (mainly medicines). As per (iv), there is a standard adjustment factor that varies by country.
   
   
6. "Shadow price", i.e., add the value of any government subsidies for treatment. Alternatively, use the price of a visit to a private clinic.
   
   
7. If necessary, extrapolate to area outside that where the hospital data was collected; use visits per 10,000 ratio in (i).
   
   
8. If possible, get cross-section data on affected and unaffected areas as a check on time series in (i).
   
   
9. Get adult/child breakdown on hospital data. This will not be used in valuation of treatment costs, but in estimating lost workdays (see below).

These steps were modified for individual countries, depending on data constraints. For Singapore, they were followed largely as outlined. In Malaysia, data on hospital and clinic admissions were matched with pollution levels to produce a dose-response function. This was extrapolated to areas of Malaysia where data on admissions were unreliable. The dose-response function was also transferred to Indonesia, where a map of cumulative haze intensity was overlaid on a population map to estimate the number of people exposed to haze pollution of various levels.

b. Estimate Workdays Lost

1. Use hospital and/or clinic visits by adults as a proxy for workdays lost. Adjust visits to workdays lost by a factor suggested by local doctors.
   
   
2. If feasible, adjust for any double counting, if people frequently go first to a clinic and then to a hospital on the same day for the same sickness episode. This would not necessarily affect treatment cost, but would affect workdays lost.
   
   
3. Multiply each workday lost by the average or minimum daily wage (depending on which is most suitable in a given country; indicate which one used). Do this for all adults, male and female. (If employees continue to receive wages while on sick leave, workdays lost are considered a loss to the employer. Firms may also be slightly overstaffed to cover absenteeism; or they might pay overtime or hire temporary help later to make up for shortfalls in production. These are also costs to employers, for which workdays are an approximation.)

c. Adjust COI for Discomfort (to Approximate Willingness to Pay)

Add (a) treatment cost + (b) workdays lost to get (c) COI.

1. The COI has been found to seriously underestimate "total" damage from an illness, as measured by an individual's willingness to pay (WTP) to avoid it. (This is because in spite of treatment and sick leave, the individual still suffers discomfort.) The ratio of WTP to COI varies with the ailment. Some ranges of values can be found in the ADB Workbook (1996, p. 188). For asthma, it is about 2:1.
   
   
2. This "adjusted COI", for lack of a better term, is the value to be used for short-term health damages only. Long-term, cumulative damages are not valued.

HAZE-RELATED PRODUCTION LOSSES

These could include rural and urban activities such as reduced crop yields resulting from reduced sunlight. In practice, the only losses measurable were:

• foregone profits in Malaysia from fishing due to reduced visibility (fishing days foregone multiplied by expected profit per day) and
  
  
• reduced industrial and commercial activity due to the ten-day state of emergency in Kuching (percentage of GNP foregone).

TOURISM LOSSES

Estimate reduced tourist arrivals from non-ASEAN sources (to control for the effect of the 1997 Asian economic crisis): compare August–October 1997 to a "normal August–October". Point-of-origin of tourists was further disaggregated in the Singapore case.

AIRLINE AND AIRPORT LOSSES

To obtain the losses incurred from airport closures due to poor visibility, one would need data on cancelled flights, expressed in mileage lost, multiplied by the airline's average profit per mile. To this should be added any profits foregone from the operation of the airports themselves.

ESTIMATION METHODS FOR FIRE DAMAGES

The estimation methodology consists essentially of multiplying the area burned in August–December 1997 and multiplying those by per hectare values for various vegetation types and land uses. The per hectare values are taken from existing data on Indonesia and, failing that, from comparable ecosystems elsewhere with appropriate adjustments. Economic damages are in net terms (i.e., profit foregone, not total revenue foregone). Discounting of future costs was done at a rate of 10 per cent.

1. Area burned: Estimates are based on a total area burned of five million hectares, distributed as follows: 20 per cent forest, 50 per cent agriculture/plantation, 30 per cent others (unproductive). These figures are derived primarily from satellite mapping studies of Sumatra and Kalimantan by the National University of Singapore's Centre for Remote Imaging, Sensing and Processing (CRISP), with adjustments by EEPSEA and the WWF for areas burned outside those provinces.
   
   
2. Timber: Timber values take into account estimates of timber stock by the government of Indonesia, as well as growth estimates of forests and net international prices. A net price of US$50 per cubic metre was used. This was cross-checked with an alternative estimation method based on land values and found to yield consistent results.
   
   
3. Agriculture: Agricultural losses were estimated on lost production in terms of years of output. Differences in productivity between plantations and smallholdings were factored in, and agricultural land productivity estimates used in this chapter were generally corroborated with observed agricultural land prices; such prices would be expected to capitalize future production values. We have assumed that, after burning, full agricultural productivity would be re-established in three years, with partial productivity being re-established in years one and two after the burning. This is consistent with the average productive cycles of mixed crops (a combination of annuals and perennials and tree crops).
   
   
4. Direct forest services: A benefit transfer (BT) approach was used, drawing on average world values of tropical rain forest ecosystems, applying them only to the forest area in the sample (i.e., 1 million hectares). The principal source was Costanza et al. (1997). Figures provided in that source are probably less precise than stated values for culture, timber, and climate control/regulation and genetic resources were removed to avoid double counting with independent estimates described elsewhere. This yielded a net value lost of US$530 per hectare per year. It was assumed that non-timber forest products would be re-established over a period of five years.
   
   
5. Indirect forest services: A similar procedure to that described for direct forest services was applied and yielded a net value lost of US$1,481 per hectare per year. It was further assumed that the losses applied only to the area "effectively burned" of forest which, consistent with the "combustion factor" in CRISP estimates, was 50 per cent of actual forested area. It was assumed that indirect forest services would be re-established over two years.
   
   
6. Biodiversity losses: The approach used here is to value "capturable biodioversity" from Indonesia's perspective. It is not the full value of international value of biodiversity. The figure takes a value of US$300 per square km. per year as an average of values found from various studies of willingness to pay (WTP) to preserve tropical rain forest of various qualities.
   
   
7. Fire-fighting costs: This includes all documented costs for fire-fighting beyond "normal year" expenses. It includes the contributions of personnel and cash from within and outside Indonesia.
   
   
8. Carbon release: Carbon dioxide and methane emission estimates in the CRISP study were increased by the ratio of total area burned (five million hectares) to area assessed by CRISP (4.56 million hectares). Such emissions increase global warming, which in turn is assumed to cause economic damage. Previous studies for the Intergovernmental Panel on Climate Change (IPCC) have put a value of up to US$30 on the damage caused by a tonne of carbon emitted (Watson et al. 1996); figures up to this amount are commonly used in international negotiations (Pearce 1998). In this chapter, a conservative figure of US$10 per tonne was used.

TABLE 2.1
Summary of Research and Valuation Methods, by Country

REFERENCES

Costanza, R., R. d'Arge, R. de Groot, S. Farber, M. Grasso, B. Hannon, K. Limburg, S. Naeem, R.V. O'Neill, J. Paruelo, R.G. Raskin, P. Sutton, and M. van den Belt. "The Value of the World's Ecosystem Services and Natural Capital". Nature 387 (1997): 253.

Freeman, M. The Measurement of Environmental and Resource Values: Theory and Methods. Resources for the Future. Washington, DC: Resources for the Future, 1994.

Pearce, F. "Growing Pains". New Scientist, 24 October 1998.

Watson, R., M. Zinyowera, and R. Moss, eds. Intergovernmental Panel on Climate Change: Technologies, Policies and Measures for Mitigating Climate Change. Intergovernmental Panel on Climate Change (IPCC), November 1996. 

Document(s) 5 of 10