Pesticides, external costs and policy options for Thai agriculture

environmental science & policy 27 (2013) 103–113

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/envsci

Pesticides, external costs and policy options for Thai agriculture Suwanna Praneetvatakul a, Pepijn Schreinemachers b,d,*, Piyatat Pananurak c, Prasnee Tipraqsa d a

Department of Agricultural and Resource Economics, Kasetsart University, Bangkok, Thailand Department of Land Use Economics in the Tropics and Subtropics, Universita¨t Hohenheim, Stuttgart, Germany c Knowledge Network Institute of Thailand, Bangkok, Thailand d The Uplands Program, Chiang Mai University, Chiang Mai, Thailand b

abstract article info Article history: Received 4 May 2012 Received in revised form 19 October 2012 Accepted 22 October 2012 Published on line

This study addresses the questions of how to estimate the external costs of agricultural pesticide use and how to disaggregate these costs to particular chemicals and farm production systems. Using the case of Thailand—a lower-middle income country with an export-oriented agriculture and an annual growth in pesticide use of about 10%, we estimate the external costs of pesticide use for the period 1997–2010 by applying the Pesticide Environmental Accounting (PEA) tool and compare the estimates to an accounting of actual costs for two years. We also use the tool to estimate the external costs of two distinct production systems of rice and intensive horticulture. Using the PEA tool, we

Keywords:

estimate the average external costs of pesticide use in Thailand to be USD 27.1/ha of

Crop protection policy

agricultural land in 2010; yet the actual cost estimate for the same year is only USD 18.7/

Externality

ha. This difference leads us to discussing the strengths and weaknesses of the PEA

Food safety

approach. The negative externalities of pesticide use could be reduced by giving farmers

Pesticide Environmental Accounting

a financial incentive to use fewer pesticides, for instance by introducing an environmental

(PEA)

tax. We argue that for such instrument to be effective, it needs to be combined with

Thailand

supportive measures to change on-farm practices through awareness-raising about the

Southeast Asia

adverse effects of pesticides and introducing farmers to non-chemical alternatives to manage their pest problems. # 2012 Elsevier Ltd. All rights reserved.

1.

Introduction

Higher income countries use substantially more pesticides per unit of output and per unit of land than lower income countries, but the risks pesticides pose to consumers and farm workers are generally considered to be greater in many lower income countries, due to their incorrect use and due to the reliance on broad-spectrum pesticides that are more

hazardous (Carvalho, 2006; Konradsen et al., 2003). Lower income countries with strong economic and agricultural growth are also experiencing a rapid increase in the intensity of pesticide use and a concomitant increase in pesticide risk (Schreinemachers and Tipraqsa, 2012). The pace of this increase in pesticide use can be explained by a policy framework that promotes pesticide consumption, a loss of natural predators due to simplifications in field ecosystems as part of the process of agricultural intensification, the

* Corresponding author at: AVRDC - The World Vegetable Center, P.O. Box 42, Shanhua, Tainan 74199, Taiwan. Tel.: +886 6 583 7801x463; fax: +886 6 583 0009. E-mail address: [email protected] (P. Schreinemachers). 1462-9011/$ – see front matter # 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.envsci.2012.10.019

104

environmental science & policy 27 (2013) 103–113

development of pesticide resistance among pests and the fact that farmers find pesticides easy to use—providing them with a convenient way to control pests (see Xu et al., 2008 for the case of China). Increased pesticide use and the associated risks therein, pose enormous challenges for lower income countries that lack the institutional framework to effectively manage these risks, that lack the institutional capacity to enforce even the existing framework, and whose farmers have only a limited knowledge and awareness about the proper use of hazardous chemicals. These challenges are even more problematic in those countries which have shown the greatest increases in pesticide use, such as Brazil, Argentina, Mexico, Thailand and Malaysia (Schreinemachers and Tipraqsa, 2012). Each of these countries are trying to position themselves as major exporters of agricultural produce, but have to cope with increasingly strict food safety laws of importing countries (e.g. Okello and Swinton, 2010) and their increasingly affluent domestic consumers become concerned about the contamination of food with pesticide residues (e.g. Posri et al., 2006). This study illustrates this challenge using the case of Thailand—a lower-middle income country with an exportoriented agriculture and rapid growth in the level of pesticide use. Previous studies have shown that the contamination of food with pesticide residues is a serious problem in Thailand (e.g. Athisook et al., 2006; Panuwet et al., 2012; Tanabe et al., 1991). As in many other countries, the Thai policy debate on agricultural pesticides tends to focus on banning specific chemicals that are deemed highly hazardous, particularly carcinogenic pesticides. Such decision to ban should ideally be based on an analysis of costs and benefits, yet no such information currently exists in Thailand and so debates on what pesticides to ban have been prone to arguments based on ideology and commercial interests. Against this backdrop, this study addresses the questions of how to estimate the external costs of pesticides in agriculture, and how to disaggregate these costs to particular chemicals and farming systems. The only method currently available for quantifying the external costs of individual active ingredients or production methods is the Pesticide Environmental Accounting (PEA) tool developed by Leach and Mumford (2008, 2011). Yet, the tool was calibrated with data for high-income countries (Germany, UK and USA) and benchmarks are needed to assess how the tool performs if applied to lower income countries. We therefore test the use of the PEA tool by comparing it to actual cost estimates for 1996 and 2010, and use it to estimate the external costs of two distinct production systems of rice and intensive horticulture. The paper starts in the following section by describing the development of agricultural pesticide use in Thailand and how policy making has evolved from an initial focus on promoting pesticide use to more recent efforts aimed at reducing it. We then present the external cost estimation approach that was applied in this study, both at the national level and for the two distinct cultivation systems. After presenting the results, we then discuss the pros and cons of using the PEA tool as well as the policy options that give

farmers an incentive to take these externalities into account.

2. Agricultural pesticide use and policy development in Thailand Thailand experienced a six-fold increase in the quantity of formulated pesticide products applied per hectare over the period 1987–2010 (Fig. 1). Regressing the logarithm of pesticide use on the number of years, we estimate an average growth of 8.8% per annum ( p < 0.01) over the whole period, yet since the turn of the century this growth has been close to 10% per annum ( p < 0.01). The growth in pesticide use has far outstripped the growth in agricultural output, as can be seen from the constant decline (7.4% per annum) in pesticide productivity (i.e., output per unit of pesticides): whereas Thailand produced USD 400 of agricultural output per kg of formulated pesticide products in 1987 this was only USD 100 in 2009. Most of the increase in pesticide use since 1997 can be attributed to increased herbicide use, and especially the use of glyphosate and paraquat, two controversial herbicides which use has been restricted in several countries but not in Thailand. These two herbicides accounted for 41% of all active ingredients used in 2010. Three interrelated factors are likely to have driven this increase in herbicide use: The rising costs of agricultural labor, land use change (particularly the expansion of plantation crops such as palm and rubber), and a greater liquidity among farmers as higher revenues and subsidized farm credit programs give them a chance to buy more inputs.

Fig. 1 – Agricultural pesticide use and pesticide productivity in Thailand, 1987–2010 Notes: Output based on the value added for agriculture at constant (year 2000) prices in USD. Pesticides here include insecticides, herbicides, fungicides, acaricides, rodenticides, fumigants and molluscicides. Pesticide consumption data are based on imports. This gives a reasonable estimate of pesticide use as importers are legally required to declare that chemicals are destined for agricultural use and the domestic production of synthetic pesticides is negligible. Sources: Thapinta and Hudak (2000), Office of Agricultural Regulation (2011), FAO (2011a), The World Bank (2011).

environmental science & policy 27 (2013) 103–113

Strong support for the use of agricultural pesticides characterized Thai policies from 1950 until the late 1990s. Since the establishment of a crop protection section at the Department of Agriculture in 1950, pest control was considered to be a public service provided to farmers through pest control campaigns using chemical pesticides (Praneetvatakul et al., 2007). In the 1970s, aerial spraying of organochlorine pesticides such as aldrin and dieldrin were used to control large-scale pest outbreaks in rice paddies. Heavy use of carbamate insecticides against green leafhopper eliminated the beneficial organisms present in the rice ecosystem and reduced its capacity for natural control. Together with a rice breeding strategy that emphasized genetic homogeneity, this ultimately led to the development of insecticide resistance and secondary pest outbreaks such as the brown plant hopper (BPH) (cf. Heong, 2009). The government responded by distributing chemical pesticides free of charge, dramatically expanding its outbreak budget, and abolishing taxes on pesticide imports. The introduction of Farmer Field Schools (FFS) in Thailand in 1999 marked a turning point in terms of the government’s attitude toward pesticides (Praneetvatakul et al., 2007). After the concept was endorsed by the King of Thailand, FFS were rapidly implemented in rice growing communities across the country and the USD 7 million outbreak budget was canceled with 15% of it reallocated to FFS training. However, the promotion of FFS and integrated pest management (IPM) was not sustained, and although the concept still appears in government policies, support for it is currently very low. In 2004, the Thai government tried to improve food quality and food safety by introducing a public standard for good agricultural practice, called Q-GAP. The standard has expanded rapidly with certificates issued to 212,000 farms in 2010 alone (Schreinemachers et al., in press). However, recent case studies suggests that the expansion of this scheme has been too rapid, as there is a general lack of compliance among farmers (Amekawa, 2010) and an insignificant impact on both the average quantity and toxicity of pesticides used (Schreinemachers et al., in press). On the supply side, the government has tried to rein in pesticide use through regulation. The 1992 Hazardous Substances Act harmonized the registration, licensing and monitoring of pesticides following the FAO Guidelines on the Registration and Control of Pesticides (FAO, 1985; Vapnek et al., 2007); however, these stricter regulations have proved difficult to enforce because responsibilities are distributed over many agencies, there are a large number of companies involved in the pesticide trade, plus there are millions of farmers using pesticides (Panuwet et al., 2012; Paopongsakorn et al., 1999). After years of parliamentary debate, and in spite of fierce opposition from and lobbying by a number of pesticide companies, a stricter pesticide registration system was introduced in 2011. Whereas the previous registration was valid for an unlimited period, the new rules limit the validity to six years and require detailed toxicological data to be provided as part of the registration process (Panuwet et al., 2012). In addition, Thailand has been progressively banning the most hazardous pesticides over recent years, and by 2011 had banned 98 active

105

ingredients from being used in agriculture. Activists have also called for bans on pesticides such as carbofuran, methomyl and dicrotophos to be introduced, yet opponents of such bans have argued that these chemicals are essential for Thai agriculture and food security. Without much scientific data being available on the costs and benefits of these chemicals, the debate around banning them has been prone to the influence of ideology and commercial interests. There is therefore a need to support this debate with quantitative analysis.

3.

Methods

The use of pesticides enhances crop production by limiting the damaging effects of various living organisms that feed on crops or compete with them for resources. However, being toxic by design, pesticides can also harm organisms other than pests, such as beneficial insects and soil organisms, aquatic life and humans. This potential harm brings costs to society and the environment in the form of pest resurgence and pesticide resistance, chronic and acute health problems for people taking in pesticide residues, the pollution of water resources— including drinking water, and also costs in terms of having to monitor food systems. These costs are called external costs, as they are not included in the price that farmers pay for pesticides or consumers pay for the food they consume (Pretty et al., 2000). Making the true costs of pesticides more transparent by expressing them in monetary terms helps policy makers devising economic incentives for farmers to align pesticide use with these true costs (Waibel, 2007; Zhang et al., 2007). Two approaches have commonly been used to quantify such external costs, these being contingent valuation (e.g. willingness to pay studies) and actual cost studies. Jungbluth (1996) was the first to do an actual cost study of agricultural pesticides in Thailand, and estimated it to be USD 264.6 million in 2010 prices, the majority of which (91%) she attributed to the value of food that exceeded maximum residue limits.1 Pesticide use per hectare has since increased over threefold. Our study therefore updates Jungbluth’s estimates by collecting new data on external costs from various government agencies. However, neither actual cost studies, nor willingness to pay studies, are suitable for quantifying the external costs of a particular active ingredient or a particular production system, which is what policy makers need to know when considering intervention, such as banning a chemical for use in agriculture. The only method capable of this is the Pesticide Environmental Accounting (PEA) tool developed by Leach and Mumford (2008, 2011). The PEA tool uses a set of base values for external costs (EC) associated with the application of one kg of active pesticide ingredients. 1

Jungbluth (1996) estimated the external costs to be 5539 million baht. Consumer price index 2010 = 150.3 (1996 = 100). Average 2010 exchange rate: 31.46 baht/USD (Bank of Thailand, 2011). Hence, 5539  150.3/100  1/31.46 = USD 264.6 million.

106

environmental science & policy 27 (2013) 103–113

These base values are based on detailed actual cost studies done in Germany, the United Kingdom and the United States (Pretty et al., 2000, 2001). These actual costs, for instance, included the cost of treating contaminated tap water, of monitoring pesticides by governments, medical costs for treating pesticide poisonings, and the cost of remediating damaged habitats and losses to bee colonies. The use of these actual cost data, which can be collected from government statistics or other publicly available reports, avoids the need for subjective valuations of, say, a single bird or a human life but should be considered as a conservative estimate of the true external costs. Actual cost data were aggregated by category and divided by the total quantity of active pesticide ingredients to get the base values for the average external cost per kg of pesticides. Being a cost-transfer approach, the PEA tool then ‘transfers’ these base values to other countries by adjusting for differences in application rates, the relative toxicity of pesticides used, and economic conditions. Leach and Mumford (2008) applied the tool to estimate the external costs of pesticide use in agricultural production systems in Spain, Turkey and Israel. Yet when applying the tool to other countries there is need to find benchmarks in order to understand how the tool performs. Our study does this by comparing the external cost estimates from the PEA tool for the period 1997–2010, to the results of Jungbluth’s study and new actual cost data collected for this study. This comparison is valid as the PEA tool is also based on actual cost data. We also apply it to the farm household level to compare the external costs between rice and intensive horticulture. The PEA tool calculates the total external cost of a pesticide p (TECp) as: TEC p ¼ rate p 

8 active p X  ½ECc  Fc  ðFagemp jc 100 c¼1

¼ 1; 2Þ  Fgdppc

(1)

In which ratep represents the application rate of a pesticide p in kg of formulated product per hectare, activep is the percentage of active ingredient in the formulated product and ECc are the external cost base values taken from Leach and Mumford (2008) and converted to 2010 US dollar values. The Appendix gives a step-by-step explanation of the PEA tool with examples of the external cost calculation for atrazine, chlorpyrifos and methomyl. The PEA tool adjusts the base values for economic costs to differences in the relative toxicity of pesticides using the Environmental Impact Quotient (EIQ) tool developed by Kovach et al. (1992). The EIQ methodology comes with a database for the ecotoxicological effects of 472 active pesticide compounds on eight categories (c), including the effects on applicators and pickers (farm workers), the effects of pesticide residues on groundwater leaching and food consumption (consumers), and the effects on aquatic life, bees, birds and beneficial insects (the environment). The PEA tool converts EIQ values for each of these eight categories to external costs by multiplying the external cost base values with a factor Fc that takes three levels: 0.5 if the

chemical has a relatively low level of toxicity, 1.0 if it has a medium toxicity, and 1.5 if the chemical is highly toxic. For each category, Leach and Mumford (2008) defined the low, medium and high toxicity ranges and we show these in Appendix A. For instance, chlorpyrifos has an EIQ value of 25.00 for aquatic effects, which is in the highest range (>17) as the chemical is relatively toxic for aquatic animals. The base value for external cost for aquatic effects (USD 1.27/kg) is therefore multiplied with a factor 1.5, which means that the use of 1 kg of chlorpyrifos is associated with an external cost on aquatic life of USD 1.90. Yet, the EIQ value for bee effects is in the lower range and the external costs for bees are therefore only multiplied by a factor 0.5. The factor Fc therefore ensures that pesticides that are comparatively toxic for humans or ecosystems receive a higher external cost than pesticides that are less toxic. We will return to this issue in the discussion. The effects of pesticides on farm workers tend to be greater in lower income countries, because relatively more people are engaged in agriculture and therefore come into direct contact with pesticides. Leach and Mumford (2008) suggested using the share of the agricultural sector in the GDP as a proxy for health-related externalities, but here we use the share of agricultural labor in total employment, as this better reflects the number of people likely to come into direct contact with pesticides on farms. We therefore multiplied the external costs for farm workers (c = 1,2) by a factor Fagemp, representing a ratio of Thailand’s share of employment in agriculture to the average share of agricultural employment in Germany, the UK and the USA (weighted by GDP). We should note that although this factor captures the fact that in a lower income country more people come into direct contact with pesticides, it does not capture the fact that pesticide use in lower income countries is far more hazardous because farm workers do not sufficiently protect themselves and most spraying is done by hand rather than by tractor. On the other hand, lower income countries are likely to have lower external costs, as monitoring and clean-up is cheaper because of lower labor costs. Leach and Mumford therefore suggested adjusting the total external cost by multiplying it by a factor Fgdppc, calculated as the ratio of a country’s per capita GDP to the average per capita GDP in Germany, the UK and the USA (weighted by GDP). As the purchasing power of a dollar varies between countries, we expressed all GDP values in purchasing power parities (PPP) in current international dollars (Int. $).

4.

Data

We followed Jungbluth’s method as much as possible for the actual cost estimation, collecting data from various government agencies in Thailand. Data on the quantity and value of over 250 active ingredients used in Thai agriculture between 1997 and 2010 were obtained from the Office of Agricultural Regulation, which is part of the Ministry of Agriculture and Cooperatives (MoAC). For 77% of the active ingredients we could identify EIQ values, with the

107

environmental science & policy 27 (2013) 103–113

remainder replaced with category averages (i.e. insecticides, herbicides, fungicides).2 In order to apply the PEA method to Thailand, we converted the external costs published by Leach and Mumford (2008) to 2010 values using the average consumer price index weighted by the GDP of the three model countries (the UK, the USA and Germany), and this produced a value of 109.52 (2005 = 100) (The World Bank, 2011). The external cost estimates were adjusted to the level of income in Thailand, following Eq. (1), by multiplying the external costs by the per capita GDP of Thailand in 2010 (8554 Int. $), and then dividing it by the weighted average per capita GDP for Germany, the UK and the USA (44,315 Int. $) to produce a factor of 0.19 (The World Bank, 2011). To calculate the adjustment factor Fagemp, we used an average share of agricultural employment of 41.50% for Thailand, and 1.42% for Germany, the UK and the USA, to give an adjustment factor of 41.50/1.42 = 29.32. Farm-level data on pesticide use within rice cultivation and intensive horticultural production activities were collected using structured questionnaire surveys. Data on rice came from a 2002/2003 survey of 224 rice farmers randomly selected from five provinces in Thailand. Data on intensive horticulture activities came from a 2010 survey of 295 farmers across twelve villages in one upland watershed area in the north of Thailand, and covered the cultivation of greenhouse vegetables, cut flowers, open field flowers and vegetables, and fruit trees. Both surveys recorded the quantities of pesticide products used based on recall data, and we combined these data with a database of pesticide products to calculate the quantities of active ingredients used and the potential environmental impact using the EIQ method.

5.

Results

5.1.

External cost estimates using the PEA tool

Table 1 shows the external costs calculated using the PEA tool, for 1997 and 2010. For both years, the external effect on consumers is only about 11% while the effect on farm workers is about 83%, which reflects the large share of agricultural labor among the total labor force (Fagemp in Eq. (1)). We also determined the trend in terms of average toxicity of pesticides used in Thai agriculture from 1997 to 2010 by calculating the average EIQ value weighted by quantity, and then using an ordinary least square regression. This gave an average annual change in pesticide toxicity of 1.6% per annum ( p < 0.01). Using the PEA tool we could quantify the external costs of using each individual chemical. Of the nearly 200 chemicals with known EIQ values in our sample, the three chemicals with the highest external costs per kilogram (about USD 13.4 kg) were fipronil—a synthetic insecticide classified by WHO as moderately hazardous to humans (Class II) and which has been linked to development of insecticide resistance in rice hoppers (Heong, 2009), methyl bromide (a fumigant), and ethion (an organophosphate insecticide). However, the three 2

We obtained EIQ values from http://cceeiq-lamp.cit.cornell.edu/EIQCalc/input.php (January 2011).

Table 1 – External costs of pesticide use in Thailand as based on the PEA method (USD/ha in constant 2010 prices). EIQ category

1997

2010

Total farm worker health Applicator effects Picker effects

5.56 3.43 2.13

22.42 13.30 9.13

Total consumer health Consumer effects Ground water

0.55 0.40 0.14

2.91 2.17 0.74

Total environment Aquatic effects Bird effects Bee effects Beneficial insect effects Total

0.35 0.22 0.05 0.04 0.05 6.46

1.80 1.13 0.23 0.19 0.25 27.13

Note: Area of arable land and permanent cropland was 20.1 million ha in 1996 and 19.0 million ha in 2010 (FAO, 2011a).

chemicals with the lowest external cost per kilogram (phosphonic acid, validamycin and kasugamycin) still had an external cost of USD 7 kg, suggesting that the PEA tool does not clearly separate between chemicals with a high or low environmental impact. In fact, we determined that 96% of the variation in total external costs across chemicals was because of differences in the average application rate, while only 4% was because of differences in toxicity. We will return to this issue in Section 6. Using the PEA tool, we were also able to quantify the external costs of separate production systems. We need to caution, however, that the data for rice cultivation are for the period 2002/2003 and therefore not up-to-date. As can be seen from Table 2, the average application rate for intensive horticulture in the north of Thailand was 13.3 kg of active ingredients per hectare, which is ten times the application rate for rice cultivation (1.3 kg/ha) and 3.7 times the national average (3.6 kg/ha). Fungicides and insecticides were the main pesticides used in intensive horticulture, while herbicides and insecticides were the main pesticides used for rice cultivation. We should note that of the sampled farmers in the intensive horticultural system, 97% used synthetic pesticides while 77% relied solely on synthetic pesticides for their pest management (Schreinemachers et al., 2011). Using the PEA method, we estimated the average external cost to be around USD 19.29 ha for rice cultivation and USD 105.75 ha for intensive horticulture, which compares to an average of USD 27.13 ha for Thailand as a whole. These costs compare to average pesticide expenditures of USD 60.01 ha and USD 962.64 ha for rice and horticulture, respectively. Hence, one dollar of pesticide bought by farmers creates on average USD 0.66 of external costs for rice cultivation and USD 0.23 of external costs for intensive horticulture. Internalizing these external costs would hence require a rise in the average retail price of pesticides of between 11 and 32%, depending on the price and toxicity of pesticides. This would increase the average variable cost in rice cultivation by about 6%. Because gross output per hectare is much greater for intensive horticulture than for

108

environmental science & policy 27 (2013) 103–113

Table 2 – Comparison of pesticide use and external costs for rice cultivation and intensive horticulture (at constant 2010 prices). Rice cultivationb

Variable a

Intensive horticulturec

Pesticide application rate (kg/ha) Herbicides (%) Insecticides (%) Fungicides (%)

1.3 49.0 45.6 4.6

13.3 11.4 29.4 54.4

Pesticide expenditure (USD/ha) External cost (USD/ha)

60.01 19.29

962.64 105.75

Gross output (USD/ha) External cost/gross output a b c

465.26 0.041

12,010.06 0.009

In active ingredients. 2002/2003 data for 224 rice farmers in five provinces. 2010 data for 295 farmers in a northern Thai watershed.

rice, pesticide productivity (gross output per unit of pesticide use) was much higher in the first system and the external costs per unit of output were therefore lower.

5.2.

External cost estimates based on actual costs

Using the actual cost method, we estimate that the external cost of pesticide use in Thailand was USD 353 million in 2010, with details of the data and calculations used shown in Table 3. Health costs were estimated to be USD 0.134 million, based on 8546 registered cases of acute pesticide poisoning in 2010, but because registered cases are sure to underestimate actual health costs, we followed Jungbluth’s approach of using a cost transfer function based on a detailed case study of health costs among tangerine growers (Whangthongtham, 1990). Based on this approach, we estimated total health costs to be USD 2.79 million.

The most recent data for 2006/2007 show that 15% of fruit and vegetables exceeded the maximum residue limits, giving a total residue cost of USD 228.13 million. About USD 15.77 million was spent by the government in 2010 on controlling the BPH outbreak, which, being a secondary pest, is an external cost resulting from insecticide misuse. Government research spending on pesticide research was about USD 38.85 million and a further USD 0.48 million was spent on the research and development of pesticide inputs. As the public GAP program narrowly defines food safety as being the prevention of food contamination due to pesticide residues, its entire budget of about USD 60.34 million can be considered as actual pesticide costs. In addition, the National Bureau of Agricultural Commodity and Food Standards has a budget of about USD 5.89 million for setting and monitoring food safety standards. Table 4 compares the estimates with those of Jungbluth’s 1996 study, and shows an increase in actual costs from USD

Table 3 – Estimates of actual costs related to pesticides in Thailand in 2010 (million USD). Cost category 1. Health costs due to acute pesticide poisoning a) Registered cases b) All cases 2. Pesticide contamination of: a) fruit b) vegetables 3. Costs related to the BPH outbreak in 2010 4. Budget for research related to pesticide issues 5. Budget for R&D on agricultural production inputs (related to pesticides) 6. Budget of the Q-GAP program 7. Food safety standards Totalb a b

Million USD

0.13 2.79

155.25 72.88 15.77 38.85 0.48

60.34 5.89

Source and calculation Registered cases: 8546 cases of pesticide poisoning recorded in the National Health Insurance Database in 2010 (Biothai, 2011). Average cost per case was 494.12 Baht (Jungbluth, 1996). All cases: Cost transfer approach (Jungbluth, 1996; Whangthongtham, 1990). Number of poisoning cases/kg of pesticide use  total amount of pesticide use in 2010 15% of fruit and vegetables exceeded maximum residue limits in 2006/2007 (Athisook et al., 2006). Multiplied by fruit and vegetable output valued at farm gate prices (Anonymous, 2008) Data obtained from summary of a government cabinet meeting on 1 February 2011a Budget for pesticide research at the Entomology Division. Estimated at 40% of the total budget of the DOA in 2010 (MoAC, 2011). Budget in 2010 at Agricultural Production Science Research and Development Office (DoA, 2011) (projects 4–7 related to pesticides) Annual Report, Department of Agricultural Extension, 2009–2010 (DoAE, 2010) Food safety standards set by the National Bureau of Agricultural Commodity and Food Standards (ACFS). Summary of 2010 Budget Report (ACFS, 2010)

352.70

http://www.eppo.go.th/admin/cab/cab-2554-02-01.html#19. Sum of categories 1b-7.

environmental science & policy 27 (2013) 103–113

Table 4 – External costs of pesticide use in Thailand in 1996 and 2010 as based on actual cost data (million USDestimated using constant 2010 prices). Cost Item

1996

2010

Medical Fruit and vegetables exceeding maximum residue limits Pest resurgence/secondary pest outbreaks Government agency expenditure related to pesticides

0.60 234.25

2.79 228.15

2.67 18.69

15.76 105.56

Total costs Costs/agricultural land (USD/ha)

256.21 12.78

352.70 18.71

256 million to USD 353 million, which corresponds to an average annual growth rate of 2.5%. This is much less than the average annual growth in pesticide use of nearly 10% over the same period.

5.3.

Comparison of PEA tool with actual cost estimates

Because both external cost methods are based on actual cost data, we can compare the estimates (as shown in Fig. 2). For 1996/1997, external costs estimated using the PEA tool are about two times lower than those estimated using the actual cost method, whereas for 2010, they are much higher. The trend in external costs produced from the PEA tool (14.0% per annum) closely follows growth in the quantity of pesticide use (10.6%), whereas external costs from the actual cost method grow much slower (2.5%). Another significant difference is that according to the PEA tool, more than 80% of the external costs are due to adverse health effects on farm workers (Table 1) while according to the

Fig. 2 – External cost of pesticide use in Thailand (1996– 2010), as estimated using the PEA tool and actual cost method (at constant 2010 prices).

109

actual cost method, less than 1% of the external costs are health costs.

6.

Discussion

6.1.

External cost calculation

The strength of the PEA tool is that it can be used to quantify the external cost of a particular pesticide, pest management method or farming system, which makes comparisons possible. To our knowledge it is the only tool able to do this, and such data is essential for rationalizing pesticide use and for formulating effective pesticide policies. Furthermore, the PEA tool is straightforward to apply, as application rates are relatively easy to determine and EIQ base values are known for many chemicals. Nonetheless, our study revealed several weaknesses in the PEA approach. First, by relying on the EIQ methodology, it captures toxicity but not risk exposure, which is determined by environmental factors (weather, soils and hydrology) and the way farmers handle pesticides and protect themselves. The PEA tool does not consider these factors explicitly, but simply multiplies external costs by a factor Fagemp, and assumes a linear relationship between the number of farm workers and human exposure. Feola et al. (2011) showed that a simple risk indicator such as the EIQ is no reliable proxy for more complex indicators that take into account both exposure and toxicity, and so by extension this will also be true for the PEA tool. Second, we find that the PEA tool does not clearly differentiate between highly toxic and less toxic pesticides, a weakness which is partly an inherent problem with the EIQ methodology. Studying 72 control strategies, Feola et al. (2011) found a correlation between total environmental risk and application rates that were near unity (i.e., very little of the variation in environmental risk was due to variations in toxicity) and our study confirmed this. However, this may also be due to the PEA tool itself, because by dividing EIQ values for each of the eight categories into low, medium and high values and multiplying the external costs by a factor 0.5, 1.0 and 1.5, respectively, the external cost of the most toxic chemical was at most three times that of the least toxic one. Third, the PEA method, as the actual cost estimates it is based on, does not capture external effects of pesticides for which no immediate monetary payments were made and both methods therefore underestimate the true external costs of pesticide use, which are likely to be substantially higher if including the full extent of chronic health effects, the development of pesticide resistance in crops, and yield losses due to pesticide misuse. Environmental impacts in particular, which tend to be long-term and for which it is difficult to prove that they were caused by pesticide exposure, might not be sufficiently accounted for. Fourth, we showed a discrepancy in external cost estimates between the PEA tool and the actual cost method. This is perhaps not surprising, as the estimates produced by the PEA tool reflect a trend in application rates, while the actual cost method reflects a trend in public expenditures and whatever

110

environmental science & policy 27 (2013) 103–113

priority the government gives to pesticide-related problems. The actual cost method shows the high priority the government gives to protecting consumers, but the PEA tool suggests that 83% of the external costs accrue to farm workers, which is plausible as farm workers account for 41.5% of total employment and are much more exposed to pesticides than consumers. The Q-GAP program reflects this bias as most resources go to the testing of food for pesticide residues to guarantee food safety for consumers, but relatively little is spend on training farmers or promoting IPM (Schreinemachers et al., in press). It is worth mentioning here that a crop protection policy based on economic rationale also needs data on the economic benefits of pesticides. Estimating a production function that includes pesticides as an abatement term has been the most common approach used; however, there are unresolved problems with the functional form of the abatement function, and by aggregating pesticides into kilograms or monetary terms, it is unable to attribute benefits to particular chemicals. There is therefore a great need to develop better tools to more accurately quantify the costs as well as the benefits of using pesticides (cf. Tesfamichael and Kaluarachchi, 2006).

6.2.

Options for crop protection policy

Until a few decades ago, agricultural pesticide use in Thailand, as in other lower income countries in East and Southeast Asia, was low, and so government policies were introduced with the aim of stimulating agricultural output through the promotion of pesticide use among farmers. Yet after several decades of strong pesticide use growth, the situation has reversed. The challenge is therefore to replace an institutional framework designed for pesticide promotion, with institutions that align pesticide use with its true costs and benefits. Pesticide externalities exist because pesticides create costs for society and the environment that are not transmitted to the farmers who choose to apply them. From an economic point of view, efficiency could be improved by internalizing these external costs into the price that farmers pay for pesticides, for instance through an environmental tax on pesticides. It is most practical to levy such tax on importers and producers of pesticides, which are few in number relative to retailers and farmers. Yet an environmental tax on pesticides is not enough to address the problem. Research from various countries shows that the demand for agricultural pesticides is typically inelastic and that a tax would only have a weak effect on pesticide demand, though generating considerable government revenues (Falconer and Hodge, 2000). In agreement with several other studies (Falconer and Hodge, 2000; FAO, 2011b; So¨derholm and Christiernsson, 2008), we think it is best to introduce a package of policy measures that combines an environmental tax with supportive measures to help farmers change their on-farm practices. In our opinion, these supportive measures should include research and development being carried out into IPM methods, support for FFS, awareness raising about the adverse effects of pesticides, and a more extensive farmer training and

education component as part of the public GAP program, the aim being to introduce farmers to non-chemical alternatives. The results of the PEA tool also suggest that it is farm workers, rather than consumers, who are most at risk of pesticides and it is therefore essential to put the focus on changing on-farm practices. We admit that a pesticide tax would be highly controversial in Thailand, and it seems there is no real political will to introduce it as a whole, mainly because the powerful rural electorate would disapprove of any cost increases for farmers (cf. So¨derholm and Christiernsson, 2008 for a discussion of environmental taxes in Europe).3 Yet, linking the revenues raised by a pesticide tax to various support measures would facilitate the acceptance of such a tax, and more importantly, create a more sustained level of support for IPM and FFS than is currently the case.

7.

Conclusion

Using the PEA tool, we estimate that the external costs of pesticides used in Thai agriculture were USD 27.1 ha of agricultural land in 2010. Yet, if accounting the actual costs for the same year, the external costs are only USD 18.7 ha. Both estimates are conservative as not all externalities will have been accounted for. The discrepancy between the estimates probably exists because the PEA estimates closely follow changes in application rates—we find that the tool does not clearly differentiate between pesticides of high and low toxicity; while the actual cost method reflects whatever priority the Thai government gives to pesticiderelated problems. As a result, external costs increased much faster in the PEA method than in the actual cost method. Applying the PEA tool to 2002/2003 farm survey data on rice cultivation and 2010 data on intensive horticulture, we estimated external costs of USD 19.3 ha and USD 105.8 ha, respectively. Internalizing these external costs, for instance through an environmental tax on pesticides, would raise pesticide prices by 11–32%; but to be effective, such tax would have to be combined with supportive measures to change on-farm practices by raising awareness among farmers about the adverse effects of pesticides and introducing them to non-chemical alternatives to manage their pest problems.

Acknowledgements The research was financially supported by the Thai Health Promotion Foundation and the Deutsche Forschungsgemeinschaft under project SFB-564. We would like to thank Chaniga Laitae for her skillful research assistance. Gary Morrison helped with reading through the English. We thank Raul Lejano and two anonymous reviewers of this journal for their constructive comments.

3 The problem is not unique to Thailand; also Vietnam has had to abandon its plan to levy a pesticide tax due to fierce opposition from farmers and pesticide companies (McCann, 2005).

111

environmental science & policy 27 (2013) 103–113

Appendix A. Five steps in calculating external costs in the PEA method Step 1: Redistribute external cost estimates of Pretty et al. (2001) over categories in the EIQ model. Categories in Pretty et al. (2001)

EIQ categories

EC (USD/kg a.i.)

1. Contamination of drinking water 2. Pollution incidents, fish death, monitoring 3. Biodiversity/wildlife losses 4. Cultural, land-scape, tourism, etc. 5. Bee colony losses 6. Acute effects to human health Total external cost (USD/kg a.i.)

Applicators

Pickers

Consumers

Ground water

Aquatic effects

6.44

0.10

0.10

0.60

0.10

0.10

–

–

–

0.87

–

–

–

0.50

0.50

–

–

–

0.65 1.59

– –

– –

– 0.50

– –

0.30

0.30 0.20

0.10 0.10

0.30 0.20

0.17 0.43

– 0.80

– 0.15

– 0.05

– –

– –

– –

1.00 –

– –

10.15

0.99

0.71

4.68

1.08

1.27

0.51

0.40

0.51

Birds

Bees

Benef. insects

Source: Leach and Mumford (2008). Note: EC = external cost as estimated by Pretty et al. (2001) and converted to 2010 US dollar values. The bottom row shows the external costs redistributed to the EIQ categories.

Step 2: For each pesticide, collect data on the average application rate and get EIQ values. Pesticide (examples)

Application rate (kg a.i./ha)

Atrazine Chlorpyrifos Methomyl

0.1951 0.1004 0.0368

EIQ categories Applicators

Pickers

Consumers

Ground water

Aquatic effects

Birds

Bees

Benef. insects

5.00 5.00 5.00

3.00 1.00 1.00

4.00 1.00 6.00

3.00 1.00 5.00

9.00 25.00 3.00

12.00 9.00 6.00

9.00 15.00 15.00

23.55 23.55 25.00

Sources: Pesticide quantities obtained from the Office of Agricultural Regulation (Thailand), agricultural area derived from FAO (2011), and EIQ values came from http://cceeiq-lamp.cit.cornell.edu/EIQCalc/input.php.

Step 3: For each EIQ category, determine if a pesticide is of relatively low, medium or high toxicity. Range of EIQ values

Factor (Fc)

Low risk Medium risk High risk Pesticide (examples) -Atrazine -Chlorpyrifos -Methomyl

EIQ categories Applicators

0.5 1.0 1.5

<25 25–85 >85

0.5 0.5 0.5

Pickers

Consumers

<14 14–76 >76

0.5 0.5 0.5

<16 16–55 >55

0.5 0.5 0.5

Ground water <2 2–4 >4

1.0 0.5 1.5

1.0 1.5 0.5

Aquatic <5 5–17 >17

Birds

Bees

Benef. insects

<15 15–51 >51

<15 15–51 >51

<25 25–85 >85

0.5 0.5 0.5

0.5 1.0 1.0

0.5 0.5 0.5

Source: Ranges for EIQ values come from Leach and Mumford (2008).

Step 4: Calculate the economic adjustment factors. Adjustment factor

People coming into contact with pesticides (Fagemp) Cost of labor (per capita income levels) (Fgdppc)

EIQ categories Applicators

Pickers

Consumers

Ground water

Aquatic

Birds

Bees

Benef. insects

29.32

29.32

1.00

1.00

1.00

1.00

1.00

1.00

0.19

0.19

0.19

0.19

0.19

0.19

0.19

0.19

Source: Adjustment factors calculated using data from The World Bank (2011).

112

environmental science & policy 27 (2013) 103–113

Step 5: Estimate the total external cost of a pesticide through multiplication (application rate  external cost base value  toxicity factor  economic adjustment factors) and then add up all EIQ categories. Aggregating the external cost of each pesticide in use gives the total external cost over all pesticides. Pesticide (examples)

Total external cost (USD/ha)

Atrazine Chlorpyrifos Methomyl

1.12 0.58 0.21

EIQ categories Applicators

Pickers

Consumers

Ground water

Aquatic

Birds

Bees

Benef. insects

0.54 0.28 0.10

0.38 0.20 0.07

0.09 0.04 0.02

0.04 0.01 0.01

0.05 0.04 0.00

0.01 0.00 0.00

0.01 0.01 0.00

0.01 0.00 0.00

references

ACFS, 2010. 2010 Budget Plan. National Bureau of Agriculture Commodity and Food Standards, Bangkok. Amekawa, Y., 2010. Towards sustainable agriculture in the developing world: Theoretical perspectives and empirical insights.Graduate Theses and Dissertations, Paper 11294, Ohio State University. Anonymous, 2008. Thailand in Figures 2006/07. Alpha Research Co. and Manager Information Service Co., Bangkok, Thailand. Athisook, K., Lertreungdej, Y., Santadrop, J., Jongmeevasana, P., Payanant, T., Pattanakijjarak, A., Chaipol-ngam, R., 2006. Pesticide residue in foods monitored in Thailand, 2000–2003. Bulletin of the Department of Medical Sciences 48, 108–120. Bank of Thailand, 2011. Daily foreign exchange rates. Available online at: http://www.bot.or.th/English/Statistics/ FinancialMarkets/ExchangeRate/_layouts/Application/ ExchangeRate/ExchangeRate.aspx (accessed February 2012). Biothai, 2011. Brief paper ‘‘Effects of chemical pesticides on Thai health’’, Available online at: www.biothai.net/sites/ default/files/BriefingPesticideHEALTH.pdf (accessed December 2011). Carvalho, F.P., 2006. Agriculture, pesticides, food security and food safety. Environmental Science & Policy 9 (7–8), 685–692. DoA, 2011. Research and Development Program on Agricultural Production Inputs. Department of Agriculture, Ministry of Agriculture and Cooperatives, Bangkok. DoAE, 2010. Annual Report 2009. Department of Agricultural Extension. Ministry of Agriculture and Cooperatives, Bangkok. Falconer, K., Hodge, I., 2000. Using economic incentives for pesticide usage reductions: responsiveness to input taxation and agricultural systems. Agricultural Systems 63 (3), 175– 194. FAO, 1985. Guidelines for the registration and control of pesticides. Food and Agriculture Organization of The United Nations, Rome. FAO, 2011a. FAOSTAT Database on Production. FAO Statistics Division, Food and Agriculture Organization of the United Nations, Rome. Available online at: http://faostat.fao.org (accessed March 2011). FAO, 2011b. Save and Grow. A Policymaker’s Guide to the Sustainable Intensification of Smallholder Crop Production. FAO, Rome. Feola, G., Rahn, E., Binder, C.R., 2011. Suitability of pesticide risk indicators for Less Developed Countries: A comparison. Agriculture, Ecosystems and Environment 142 (3–4), 238–245. Heong, K.L., 2009. Are planthopper problems caused by a breakdown in ecosystem services? In: K.L.H.B.H. (Eds.), Planthoppers: New Threats to the Sustainability of Intensive Rice Production Systems in Asia. International Rice Research Institute, Los Ban˜os, Philippines, pp. 221–231.

Jungbluth, F., 1996. Crop Protection Policy in Thailand: Economic and Political Factors Influencing Pesticide Use, Pesticide Policy Project Publication Series. Institute of Horticultural Economics, University of Hannover, Hannover. Konradsen, F., van der Hoek, W., Cole, D.C., Hutchinson, G., Daisley, H., Singh, S., Eddleston, M., 2003. Reducing acute poisoning in developing countries – options for restricting the availability of pesticides. Toxicology 192 (2–3), 249–261. Kovach, J., Petzoldt, C., Degni, J., Tette, J., 1992. A Method to Measure the Environmental Impact of Pesticides. New York’s Food and Life Science Bulletin 139, New York Agricultural Experiment Station. Cornell University, Ithaca, NY, 8 pp. Leach, A.W., Mumford, J.D., 2008. Pesticide environmental accounting: A method for assessing the external costs of individual pesticide applications. Environmental Pollution 151 (1), 139–147. Leach, A., Mumford, J., 2011. Pesticide environmental accounting: a decision-making tool estimating external costs of pesticides. Journal fu¨r Verbraucherschutz und Lebensmittelsicherheit 6, 21–26. McCann, L., 2005. Transaction costs of pesticide policies in Vietnam. Society & Natural Resources 18 (8), 759–766. MoAC, 2011. Annual Report 2010: Working Performance and Expenditure of the Fiscal Year 2010. Ministry of Agriculture and Cooperatives, Bangkok, Thailand. Office of Agricultural Regulation, 2011. Data on the Import of Pesticides for Use in Agriculture. Ministry of Agriculture and Cooperatives, Bangkok. Okello, J.J., Swinton, S.M., 2010. From circle of poison to circle of virtue: pesticides, export standards and Kenya’s Green Bean Farmers. Journal of Agricultural Economics 61 (2), 209–224. Panuwet, P., Siriwong, W., Prapamontol, T., Ryan, P.B., Fiedler, N., Robson, M.G., Barr, D.B., 2012. Agricultural pesticide management in Thailand: status and population health risk. Environmental Science & Policy 17 (0), 72–81. Paopongsakorn, N., Meenakanit, L., Jungbluth, F., Waibel, H., 1999. Approaches to Pesticide Policy Reform. Institute of Development and Agricultural Economics, Faculty of Economics and Management, Leibniz University of Hannover, Hannover. Posri, W., Shankar, B., Chadbunchachai, S., 2006. Consumer attitudes towards and willingness to pay for pesticide residue limit compliant ‘‘safe’’ vegetables in Northeast Thailand. Journal of International Food & Agribusiness Marketing 19 (1), 81–101. Praneetvatakul, S., Waibel, H., Meenakanit, L., 2007. Farmer Field Schools in Thailand: History, Economics and Policy. Pesticide Policy Project Publication Series, No.12. Institute of Development and Agricultural Economics, Faculty of Economics and Management, Leibniz University of Hannover, Germany. Pretty, J., Brett, C., Gee, D., Hine, R., Mason, C., Morison, J., Rayment, M., Bijl, G.V.D., Dobbs, T., 2001. Policy and practice: policy challenges and priorities for internalizing the externalities of modern agriculture. Journal of Environmental Planning and Management 44, 263–283.

environmental science & policy 27 (2013) 103–113

Pretty, J.N., Brett, C., Gee, D., Hine, R.E., Mason, C.F., Morison, J.I.L., Raven, H., Rayment, M.D., van der Bijl, G., 2000. An assessment of the total external costs of UK agriculture. Agricultural Systems 65 (2), 113–136. Schreinemachers, P., Schad, I., Tipraqsa, P., Williams, P.M., Neef, A., Riwthong, S., Sangchan, W., Grovermann, C. Can public gap standards reduce agricultural pesticide use? The case of fruit and vegetable farming in northern Thailand. Agriculture and Human Values, http://dx.doi.org/10.1007/ s10460-012-9378-6, in press. Schreinemachers, P., Sringarm, S., Sirijinda, A., 2011. The role of synthetic pesticides in the intensification of highland agriculture in Thailand. Crop Protection 30 (11), 1430–1437. Schreinemachers, P., Tipraqsa, P., 2012. Agricultural pesticides and land use intensification in high middle and low income countries. Food Policy 37 (6), 616–626. So¨derholm, P., Christiernsson, A., 2008. Policy effectiveness and acceptance in the taxation of environmentally damaging chemical compounds. Environmental Science & Policy 11 (3), 240–252. Tanabe, S., Kannan, K., Tabucanon, M.S., Siriwong, C., Ambe, Y., Tatsukawa, R., 1991. Organochlorine pesticide and polychlorinated biphenyl residues in foodstuffs from Bangkok, Thailand. Environmental Pollution 72 (3), 191–203. Tesfamichael, A.A., Kaluarachchi, J.J., 2006. A methodology to assess the risk of an existing pesticide and potential future pesticides for regulatory decision-making. Environmental Science & Policy 9 (3), 275–290. Thapinta, A., Hudak, P.F., 2000. Pesticide use and residual occurrence in Thailand. Environmental Monitoring and Assessment 60 (1), 103–114. The World Bank, 2011. World Development Indicators. The World Bank, Washington, DC. Available online at: http:// databank.worldbank.org (accessed February 2011). Vapnek, J., Pagotto, I., Kwoka, M., 2007. Designing national pesticide legislation. FAO Legislative Study 97, Food and Agriculture Organization of the United Nations, Rome. Waibel, H., 2007. Externalities and the costs of pesticide dependence. In: Kunstadter, P. (Ed.), Pesticides in Southeast Asia Environmental, Biomedical and Economic Uses and Effects. Silkworm Books, Chiang Mai, Thailand, pp. 181–195. Whangthongtham, S., 1990. Economic and Environmental Implications of Two Alternative citrus Production Systems – A Case Study from Pathum Thani Province. Asian Institute of Technology, Bangkok, Thailand.

113

Xu, R., Kuang, R., Pay, E., Dou, H., de Snoo, G.R., 2008. Factors contributing to overuse of pesticides in western China. Environmental Sciences 5 (4), 235–249. Zhang, W., Ricketts, T.H., Kremen, C., Carney, K., Swinton, S.M., 2007. Ecosystem services and dis-services to agriculture. Ecological Economics 64 (2), 253–260. Suwanna Praneetvatakul is an Associate Professor at the Department of Agricultural and Resource Economics, Faculty of Economics, Kasetsart University, Bangkok, Thailand. She holds a PhD in Agricultural Economics from the University of Hohenheim, Germany and an MSc in Agricultural Systems from Asian Institute of Technology, Thailand. Her main research interests include sustainability of agriculture, natural resources and environmental evaluation and management, and the impact assessment of agricultural research. Pepijn Schreinemachers is Agricultural Economist at AVRDC The World Vegetable Center. At the time of this study he was senior researcher at the University of Hohenheim. He holds a PhD in Agricultural Economics from the University of Bonn, Germany, and an MSc in Rural Development Studies from Wageningen University, the Netherlands. His current research focuses on integrated modeling, pest management and impact assessment. Piyatat Pananurak is a researcher at the Knowledge Network Institute of Thailand (KNIT). She has been working at KNIT since 2010 after obtaining her PhD in Economics from the Institute of Development and Agricultural Economics, Gottfried Wilhelm Leibniz Universita¨t Hannover, Germany. She also holds an MSc in Agricultural Economics from Kasetsart University, Thailand. Her main research interests include agricultural policy, agricultural development, and the impact assessment of agricultural research. Prasnee Tipraqsa is Thailand Project Manager for the USAID funded project ‘‘Lowering Emissions in Asia’s Forests’’. She holds a PhD in Natural Resource Management from the Heidelberg University, Germany and an MSc in Environmental Science from Kasetsart University, Thailand. Her main research interests are the impact of land use change and climate change on environmental services, agricultural development, and the use of agrochemicals.