Poverty-Environment Report:
Pesticide Use in the Mekong Delta, Vietnam
By
Craig Meisner, DECRG-IE
I. Background
Vietnam began using pesticides as early as the 1950s, when agricultural production was limited to cooperatives, collective farms, and state farm enterprises. Under this regime, little knowledge existed of the hazards of pesticides, and with no regulation system. Application was centrally-run by brigades of 4-5 farmers who worked in conjunction with the Plant Protection Department (PPD). The PPD supplied pesticides at subsidized prices and recommended spraying on a calendar basis, with little or no attention to field conditions. Brigade interventions generally resulted in high costs but had little effect on pests and diseases (Pincus 1995; Chung and Dung, 2000).
By 1988, farming units were recognized as autonomous production units giving households the control over all stages of production, including the pesticide use decision. As a consequence of past excess, coupled with a new lease on production, the dependency on pesticides grew and as Figure 1 shows, this growth peaked in 1998 at over 40,000 tons (FAO, 2004).[1]
Figure 1: Consumption of pesticides, 1990-1999.
In addition to the alarming increase in pesticide use, there is also significant evidence suggesting that the substances being used are harmful to human health and the environment. In a nation-wide survey conducted by the PPD in 2000, 2,500 kg of banned pesticides were found (methamidophos, DDT and other chemicals), along with 4,753 liters and 5,645 kg of illegally imported or counterfeit pesticides (PPD, 2000). Another survey in August 2000, among 480 farmers in four provinces in the South, found 96.6% of farmers used pesticides excessively and not in compliance with the instructions on the labels, and nearly 95% of farmers pour residual spray into canals and ditches or reapply on other plants or over-apply on the same crop just to get rid of it (Huan et al., 2000). In regards to the health impacts of pesticide use, hospital admission records in Vietnam trace nearly 11% of all poisonings to pesticide misuse: or approximately 840 poisonings in 53 cities and provinces in 1999 (Poison Control Center, Vietnam Ministry of Health, 2000).[2]
Given the rather dire picture of pesticide use in Vietnam, two broad questions that this section of the Poverty-Environment Nexus report attempts to answer are:
1) Are the poor using more, toxic pesticides?
2) Is pesticide use, and its attended health consequences, impacting the poor to a greater extent than the non-poor?
In developing a response to these questions, the World Bank team constructed detailed pesticide use and medical surveys to be implemented on a representative sample of poor and non-poor farmers in Vietnam. A description of the survey, sample and the results follows.
II. Description of the World Bank survey and sampling strategy
A survey of Vietnamese farmers was carried out by the Research Department of the World Bank in the winter of 2003. Structured questionnaires were used to collect information on farming systems, pesticide use and practices, applicator precautions/ averting behavior and health/ environmental effects. The survey was divided into two major sections, one dealing with the economics of pesticide use (e.g. production, pesticides used, protection, training, etc) and another strictly dealing with the health of the farmer (e.g. questions on self-reported health ailments/ health-related habits, a general physical exam, and patch-skin and blood tests for pesticides). The survey was constructed by the World Bank team in collaboration with the University of Economics in Ho Chi Minh City[3] (for the socio-economic portion of the survey) and the Centre of Occupational and Environmental Health (COEH) of the Vietnam Association of Occupational Health (VINOH) for the medical survey.[4] Each of these World Bank counterparts were responsible for the implementation of the survey. To minimize any possible reporting bias, the survey was conducted under the agreement that the team would not reveal the identity of the farms surveyed or the respondents.
Although the survey was conducted in a single season (the winter of 2003), respondents were asked detailed production and pesticide use questions that spanned the period of one year. Thus the survey contained information for the three major growing seasons of Winter-Spring 2002-2003, Spring-Summer 2003, and Summer-Autumn 2003.
II a) Definition of the poor
As the primary focus of the study was to investigate the potential inter-linkages of pesticide use and the poor, we adopted the poverty line set by the Ministry of Labor, Invalids, and Social Affairs (MOLISA) Decision No. 1143 (MOLISA, 2001). This poverty line was also re-iterated in the recent Comprehensive Poverty Reduction and Growth Strategy (CPRGS) signed by the government in 2002.[5] Applying this criteria to our sample of farmers, Table 1 presents the number of respondents who fell into the category henceforth known as the ‘poor’.
Table 1: Sample poverty distribution[6]
Poverty classification / Number / PercentPoor (≤ 1,200,000 VND/year) / 79 / 13.1
Non-poor / 524 / 86.9
Total / 603 / 100.0
II b) Geographical location of the sample
The main geographical area of interest is the Mekong Delta since it has been identified as the most intensive with respect to production and pesticide use (see Map 1 for details). As a sampling strategy, the team took the following into consideration when selecting candidate districts and communes: 1) identifying areas where the poor versus non-poor comparison would be possible (via the use of the latest poverty map from the World Bank, 2003); 2) areas where rice production is dominant; 3) selection of farms that vary in geographical location (e.g. coastal versus inland farms); and 4) focus on rural areas so as to target the rural poor (as opposed to the urban poor in the cities).
Map 1: District location and number of poor/non-poor farms in the survey sample
Note: the low representation of poor farming households is primarily due to the definition of ‘poor’ (as 1,200,000 VND/ year); where the majority of farms exceed this poverty line since many of the poor do not own land and thus do not make pesticide decisions (the primary focus of the survey).
III. Survey Results
III a) General farmer statistics
Table 2: Farmer characteristics (%).
Ownership
/Poor
/Non-poor
/Farm size
/ Poor /Non-poor
/Education
/Poor
/Non-poor
Own < 100% of land / 5.1 / 3.8 / < 0.50 ha / 36.7 / 9.4 / No schooling / 1.3 / 1.3Own 100% of land / 94.9 / 96.2 / 0.50-0.99 / 31.6 / 26.3 / Primary / 54.4 / 32.1
Total / 100.0 / 100.0 / 1.00-1.49 / 20.3 / 23.7 / Secondary / 34.2 / 45.8
1.50-1.99 / 6.3 / 17.2 / High school / 7.6 / 18.7
2.00-2.49 / 1.3 / 8.6 / Other / 2.5 / 2.1
2.50+ / 3.8 / 14.9 / Total / 100.0 / 100.0
Total / 100.0 / 100.0
Sample size (n=) / 79 / 524 / 79 / 524 / 79 / 524
A large proportion of the farmers surveyed owned 100% of their land, with 89% of poor and 60% of the non-poor farming less than 1.5 hectares of land. In terms of education, over half of the poor had at least primary education, while a larger percentage of the non-poor had secondary and high school education.
III b) Pesticide Use
Pesticide use is generally measured using a variety of key indicators such as number of applications, the absolute amount used, as well as a measure of the relative risk of the pesticide. The absolute measures of pesticide quantity are straightforward in interpretation, however, factoring in the relative risk of each pesticide requires the adoption of a methodology that can rank one pesticide as more toxic than another. The methodology we propose is elaborated below.
By simply summing all pesticides (measured as kg of active ingredient) used in crop production, this implicitly assumes that all pesticides are alike in terms of their toxicity. However, it is common knowledge that different substances have different levels of toxicity to both humans and the environment. To do this, we constructed risk-weighted measures by defining risk as the relative toxicity or lethality of each active ingredient. Risk-weighted measures place a greater weight on more toxic substances and provide a more convenient measure of comparing the use of one pesticide over another from a health-hazard perspective. To gauge the relative toxicity of each active ingredient, a measure called the LD50 (or lethal dose 50%) is used. LD50 is a statistical estimate of the number of milligrams (mg) of toxicant per kilogram (kg) of bodyweight required to kill 50% of a large population of test animals. Pesticides with a lower LD50 value are more toxic, thus in the calculations used for this study, each pesticide load was multiplied by 1/LD50 to account for its relative toxicity and giving greater weight to more toxic substances.
Figure 1: Mean pesticide application, risk-weighted amount
and number of applications by poor/non-poor (kg).
*** - difference is statistically significant at the 1% level;
** - difference is statistically significant at the 5% level.
Returning to the pesticide use indicators in Figure 1, we see that the mean application amount over the three growing seasons is higher for the non-poor, as well as in terms of the number of applications. However, when we account for the relative risk of the pesticides used, the poor are using more toxic pesticides in absolute terms.[7]
To better understand the extent of risk exposure, make use of a widely-known categorical method developed by the World Health Organization (WHO) and also based on the LD50 measure.[8] Pesticides are divided into 4 major hazard groups: Category Ia & Ib (extremely hazardous), Category II (moderately hazardous), Category III (slightly hazardous), and Category U (least hazardous or unlikely to present acute hazard under normal use). Field evidence suggests that human poisonings correlate reasonably well with these toxicity ratings (Levine and Davies, 1982). In Table 3, we provide a summation of the total amounts of pesticides under the WHO classification system. Note that over one-half of the pesticides under use in this sample are moderately hazardous. In Figure 2 we see further evidence of the earlier result that the poor are using more toxic pesticides by the higher mean use of WHO Ia & Ib pesticides.
Table 3: Total amount pesticides applied by WHO classification
Category / Total (kg a.i.) / PercentIa (extremely hazardous) / 0.0 / 0.0
Ib (extremely) / 5.4 / 0.5
II (moderately) / 594.7 / 55.5
III (slightly) / 187.3 / 17.5
U (Unlikely) / 283.8 / 26.5
Total / 1071.1 / 100.0
Figure 2: Mean application of WHO Ia & b
and common pesticide classes by poor/non-poor[9] (kg).
*** - difference is statistically significant at the 1% level;
Yet another important method of classifying pesticides is by their chemical class. The chemical class is quite simply a pesticides’ chemical structure, which will dictate the properties that a pesticide holds. Perhaps the two of the most important properties to human and environmental health are toxicity and degradability. Certain chemical classes have been identified as being highly toxic and persistent in the environment. Epidemiological studies have linked so-called carbamates, organophosphates and pyrethroids with fetal death, hormonal changes, DNA damage, birth defects, and abnormal sperm, ovaries and eggs.[10][11]
Among the sample farmers in the Mekong Delta, we see a consistent pattern in Figure 2 among the non-poor, where they are using significantly higher average amounts of carbamates, organophosphates and pyrethroids. Once again we note in the first bar of Figure 2 that it is the poor who are using the most toxic pesticides according to the WHO classification system. Thus when setting priorities, one must first select a measure upon which a policy benchmark can be formulated. For example, if one was concerned with banning only WHO Ia & Ib pesticides, then the evidence suggests that focusing on poor farmers may better address this issue than strictly focusing on all farmers in general.
Figure 3: Mean application of insecticides, herbicides,
fungicides and others by poor/non-poor (kg).
*** - difference is statistically significant at the 1% level;
Pesticides can also be classified by their intended use, for example, the use of insecticides for insect attacks, herbicides for weeds, and fungicides for the mitigation of plant mildew and root rot. Results of the survey reveal in Figure 3 that the non-poor use insecticides, herbicides and fungicides at a rate more than double that of the poor. These averages also closely coincide with those of other studies of pesticide use in the Mekong Delta, where insecticides dominated the group averages (Dung and Dung, 2000).
Figure 4: Misuse of pesticides: Do farmers use any pesticides recommended for other crops, on rice?
The misuse of pesticides, or using pesticides on a crop for which it is unintended, is not a very common practice among the farmers in the survey; although, nearly 5% of non-poor farmers mentioned misuse of some kind.[12] These results also highlight the importance of confirming self-reported information in the field.
The overuse of pesticides is also a concern from health and environmental perspectives. In the survey farmers were asked their pesticide dosage rates which were then compared to the recommended dosages on the label. A statistical analysis of the extent of pesticide overuse revealed that the poor have a significantly lower probability of overusing.[13] Other variables that contributed to the determination of overuse were farm size, income and the proportion of WHO class II pesticides used in production. Overuse was also found to be more prevalent in the provinces of An Giang and Cantho.
Table 4: Perceived risk of pesticides by poor/non-poor.
Perceived risk / Poor / Non-poorHigh risk / 30.4 / 21.0
Medium risk / 25.3 / 26.3
Small risk / 27.8 / 31.1
No risk / 13.9 / 20.4
Don’t know / 2.5 / 1.1
Total / 100.0 / 100.0
Sample size (n=) / 79 / 524
Increasing a farmer’s awareness of pesticide risk is an important first step in designing any program on safer alternatives, however, do farmers think that the pesticides they are using pose any significant risks to their health and the environment? According to the study’s results in Table 4, only 42% of the poor and 52% of the non-poor think the pesticides they are using pose small or no risks. In addition, the low level of uncertainty (“Don’t know”) leads us to believe that a high percentage of farmers regard their pesticide use as hazardous, however, it may be the case that they are unaware of the exact health consequences.
III c) Pesticide control measures and averting behavior
As an alternative to pesticide use, other pest management methods are possible and appear to be gaining traction in Vietnam. The main appeal to alternative methods is that the farmer decreases his/her use of pesticides and it is therefore considered a more “safer” alternative. Examples of other methods include Integrated Pest Management (IPM: an ecologically-based approach to control of harmful insects and weeds[14]), biological control (use of natural or modified organisms, genes, or gene products to reduce the effects of pests and diseases), or organic methods (farming that avoids the use of synthetic chemicals and genetically modified organisms (GMOs)).
Among the farmers sampled in the Mekong Delta, approximately 32% of the non-poor practice alternative pest control methods and 21% among the poor (Figure 4).[15] The most popular control method across both the poor and non-poor is Integrated Pest Management (IPM), then mechanical-physical techniques (such as manual weeding) and biological control measures. The level of adoption appears to be higher for the non-poor leading to a possible explanation that the non-poor may have better access to information and training on alternative methods. This was indeed verified when 54% of the non-poor (versus 37% for the poor) stated that they had received formal training on IPM methods. Thus if the policy goal is to train that segment of the population that is relative more unaware, one could either include more poor people in the program or have a program that targets them solely.
Figure 5: Prevalence of pest control measures by poor/non-poor.
The education of farmers on safer alternatives is one piece of a two-part initiative on raising awareness and improving farmer productivity. The second part is to increase education on the safe handling and application of pesticides. As was earlier discussed, even simple knowledge of the relative toxicity may induce the farmer to take remedial or more protective actions in the short-term. From our survey evidence, in Figure 6 we see that only about 29% of the poor and 34% of the non-poor had any formal training on pesticide safety. This difference was not significantly different leading us to believe that the poor have the same access to information as the non-poor.
Figure 6: Prevalence of basic training on safe handling
and application of pesticides by poor/non-poor.
The potential suppliers of pesticide information may enter at different stages of the pesticide use decision. As mentioned above, one stage may be ex-post, or after the farmer has chosen what to use and is now learning more about the pesticide they were using in the past. Another is at the purchasing stage or retailer level. When purchasing pesticides, farmers may also have the opportunity to learn more about the pesticide through pamphlets, brochures and other media. This information may influence the pesticide decision and would be particularly helpful to the farmer in terms of understanding what is needed to adequately prepare for the application (e.g. how much to apply, what the pesticide can be used for or on, what protective measures are necessary for avoiding exposure, etc).
Among the sample of farmers in this study, only 42% of the non-poor reported receiving safety information and procedures on pesticide use versus only 39% for poor farmers. Of the farmers who received information, the main source of the information was from the public media, pesticide companies and agricultural officials (Table 5). The significant role of the public media for the poor clearly indicates an opportunity for extension agents and local community groups to foster greater communication and awareness among poorer farmers in the region.
Table 5: Main sources of pesticide information
by poor/non-poor (%).
Source /Poor
/ Non-poorPublic media / 38.8 / 40.9
Pesticide companies / 17.8 / 18.5
Other / 14.4 / 12.8
Agricultural official / 17.7 / 11.6
Pesticide retailer / 5.5 / 9.1
No information available / 5.5 / 7.1
NGO / 0.2 / 0.1
Total / 100.0 / 100.0
Sample size (n=) / 79 / 524
In addition to formal training and access to information on the health risks posed by pesticides, protection habits are probably one of the most effective short-term measures that a farmer may take to better ensure safety. In the survey, farmers were asked about the protective clothing they wear during pesticide application. As Figure 7 shows, a high percentage of farmers use basic protective measures such as shirts, and trousers, and about half use hats and masks, even among the poor. However, with respect to increased protective measures such as gloves, glasses and shoes, the average falls, especially among the poor.[16]