Modern agricultural technologies have substantially increased crop production and farm productivity, but at what cost to society? The International Rice Research Institute's recent study of the health and productivity effects of pesticide use in two rice-producing regions of the Philippines reveals that rates of return to these technologies are overestimated because the adverse effects of pesticide use are not taken into account. In particular, the study shows that pesticide use can impair the health of farmers and consequently diminish farm productivity. Its statistical analysis confirms the theory that decreases in the health of farm workers are related to increases in the average cost of agricultural production. Thus it is in society's interest to reduce pesticide use. However, policies to reduce the use of all pesticides are not likely to generate as many social benefits as policies to reduce only the use of the most hazardous pesticides.
Green-revolution agricultural technologies—which make use of modern seed varieties and complementary inputs of fertilizer, pesticides, and mechanical power—have dramatically increased agricultural production and productivity. By conventional measures, investment in these technologies has yielded high rates of return.
But the estimates of these rates do not reflect the social costs of the technologies—in particular, the negative impact of insecticides and herbicides on the health of farm workers. Anecdotal evidence suggests that this impact may be significant in developing countries, but a lack of data has made it impossible to document and quantify the health effects of pesticide use in any of these countries and to value these effects in such a way that they can be compared with the productivity benefits of pesticide use.
The International Rice Research Institute (IRRI) began remedying this situation in 1989, when it initiated a four-year program to assess the health and economic impacts of applying pesticides to rice fields in two of the Philippines' principal rice-producing regions, the Laguna region of southern Luzon and the Nueva Ecija region of central Luzon. To document the health effects of pesticide use, IRRI staff monitored the input and technology use practices, including safety practices used in handling pesticides, of a group of farmers whose rice production practices had been studied by IRRI throughout the 1970s and 1980s. Medical staff also conducted detailed medical examinations of each individual in the sample group to document general health status and symptoms of pesticide exposure.
Analysis of the collected data proceeded in three phases. First, researchers used the health data to construct an index of farmers' health impairment and developed a model to explain farmers' health in relation to their actual use of pesticides, controlling for other factors that affect health. Second, researchers used a model of rice production to estimate the impact of farmers' health on rice productivity. Third, researchers simulated jointly the health model and the production model. This simulation allowed researchers to investigate the direct production benefits of pesticides as a production input and the indirect production losses that result from the adverse effects of pesticide exposure on farmers' health and productivity. In this way, researchers were able to obtain the health cost of pesticides in units of forgone production and thus to compare this cost with the production benefits of pesticide use. Using these data, researchers quantitatively assessed the health and productivity trade-offs of pesticide use.
We report here on the study's measurement of the health effects of pesticide exposure and on its relation of these effects to productivity. We then discuss the implications of the study's findings for agricultural research and the regulation of pesticide use. Although the study measured the environmental impacts of pesticide use in the Luzon and Nueva Ecija regions, these impacts are not included in the analysis below.
Health effects of pesticide exposure
A wide variety of pesticides is used in the production of rice in the Philippines. According to the hazard classification system of the World Health Organization, all the insecticides used by Philippine rice farmers are moderately or highly hazardous. (Many are either banned or unregistered for agricultural use in the United States.) By comparison, the herbicides used are relatively less hazardous.
To relate farmers' health to pesticide exposure, researchers constructed a general indicator of health impairment (the health impairment index). They measured the level of health impairment due to a particular illness in terms of treatment cost—the cost of returning the farmer to normal health—plus opportunity cost—the cost of the farmer's lost work time during recuperation. The researchers based estimates of each cost on assessments of the type and severity of illnesses experienced by each farm worker in the sample group.
To identify the relationship between pesticide exposure and health impairment, the researchers related the health impairment index to several personal characteristics of each farm worker and to each worker's exposure to pesticides. Personal characteristics included age; alcohol consumption; tobacco consumption; and nutritional status, which was measured by the ratio of weight to height. Exposure levels, for which pesticide use was a proxy, were measured by the number of times each worker applied chemicals to rice fields in the crop season preceding the medical examination and by the types of chemicals he or she applied.
The researchers then tested the hypothesis that pesticide exposure explains observed variations in the health of farmers in the sample group. They formulated a statistical model to relate the health impairment index to each farm worker's age, tobacco and alcohol consumption, weight-height ratio, and use of pesticides. In doing so, the researchers divided pesticides into high and low hazard groups, according to the World Health Organization hazard classification system.
Statistical analyses showed that the number of times farm workers apply high-hazard pesticides was significantly related to the level of the workers' health impairment.
As expected, analysis of the data showed that the older individuals in the sample group had more health problems than younger individuals and that those with better nutrition had fewer such problems than those with worse nutrition. The analysis also revealed that smoking and alcohol consumption were not statistically significant in explaining the level of the workers' health impairment, nor was the number of times farm workers apply low-hazard pesticides.
The analysis did show, however, that the number of times farm workers apply high-hazard pesticides was significantly and positively related to the level of the workers' health impairment. Researchers found that a 10-percent increase in the use of the highly hazardous pesticides raised the health impairment index from 3.7 percent to 7.5 percent.
This health assessment suggests that workers on rice farms in the Laguna and Nueva Ecija regions face chronic health effects as a consequence of prolonged exposure to pesticides. Specifically, IRRI's statistical analyses indicate that such exposure is significantly associated with eye, skin, lung, nervous system, and kidney problems.
Farmers' health and farm productivity
Economists regard health as an important component of the farmer's human capital, because it affects the ability to do physical labor and to manage farm operations. When a farmer's health decreases, reducing effective management and field labor, overall farm productivity also should decrease. Of course, decreases in a farmer's field labor may be partially offset by such substitutions as the hiring of additional farm workers.
Researchers estimated the effects of farmers' health on farm productivity by correlating the cost of production with input prices, farm output, and the health status of farmers. According to economic theory, the average cost of production should be inversely related to the level of productivity if input prices are held constant. Thus, economic theory predicts that as a farmer's health decreases, the average cost of production increases. IRRI's analysis confirmed this theory at a high degree of statistical significance. It revealed that a 10-percent reduction in the health status of a farmer should result in a 1.4-percent to a 3.6-percent increase in the average cost of production.
No protective clothing shields this Asian farmer from the effects of the insecticide he is applying to his rice crop. A recent study of the health effects of prolonged exposure to pesticides indicates that such exposure may cancel out the productivity gains from using them
To investigate the impact of reduced pesticide use on productivity and on farmers' health, researchers combined the estimated relationships between health and pesticide use and between health and productivity in a simulation analysis.
In the simulation, the researchers restricted pesticide use through the imposition of a tax on pesticides. This choice was made as a matter of convenience; it does not imply that a tax would have to be imposed if a country implemented a policy restricting pesticide use. In fact, quantitative regulation of pesticide use is more common than taxation of pesticides in most countries. The study results hold true either way, however, because an equivalent quantitative restriction can be defined for each tax rate specified by the analysis.
Analysis confirmed that the average cost of production increases as farmers' health decreases; a 10-percent reduction in a farmer's health could result in a 3.6-percent increase in the average cost of production.
A pesticide tax should reduce pesticide use and thus improve farmers' health, but it may also reduce productivity. The trade-offs between health and productivity are not obvious, however, because a reduction in pesticide use also has a positive impact on productivity through improved health. The simulation analysis was designed to explore these trade-offs.
Once a pesticide tax has been selected as the instrument for reducing pesticide use, it can be applied in one of two ways: insecticides and herbicides can be taxed at the same rate or at different rates. Given that the insecticides used by Philippine rice farmers are more toxic than the herbicides they use, taxing the two types of pesticides at the same rate would not be efficient—that is, it would not mitigate health impacts at least cost. In order to assess the inefficiency of a uniform pesticide tax (equal tax rates for insecticides and herbicides), researchers compared the effects of this tax with those of a tax on insecticides only.
In the simulations, the uniform tax reduced insecticide use and herbicide use by about the same percentages. A 300-percent uniform tax decreased the quantity of each used by about 80 percent. The insecticide tax alone reduced the quantity of insecticides used by a somewhat smaller percentage but had almost no effect on the quantity of herbicides used. Thus both tax scenarios resulted in improvements in the health of farm workers; but because the insecticides are more toxic to humans than the herbicides, researchers concluded that the insecticide tax was about as effective as the uniform tax in reducing the adverse health effects of pesticide use.
The effects of reduced pesticide use on productivity depend not only on the indirect effect of pesticide use on farmers' health but also on the direct effect of pesticide use on crop yields. According to the statistical model of pesticide productivity, reductions in insecticide use would decrease rice yields less than reductions in herbicide use, presumably because inappropriate use of insecticides can disrupt the pest-predator balance in rice fields and lead to increased pest damage.
Of course the direct effect of reduced insecticide use is offset by improvements in farmers' health and thus in productivity. However, because the uniform tax both reduces productivity and improves health more than the insecticide tax, it is unclear which tax would have a greater overall impact on productivity. The results of the simulation analysis indicate that the insecticide tax alone has a much smaller effect on productivity than the uniform tax because it reduces insecticide use almost as much as the uniform tax and is therefore almost as effective as that tax in reducing the adverse health effects of pesticide exposure.
The advantage of the insecticide tax over the uniform tax is also demonstrated by combining the productivity effects and the health effects of reduced pesticide use. Productivity effects are measured as the change in the average cost of production; health effects are measured as the costs of treatment and forgone labor per unit of rice produced. The sum of the average cost of production and the average health cost could be interpreted as a measure of the "average social cost" of production. It is important to note, however, that this measure does not include the potential environmental impacts of pesticide use.
Simulations of the combined productivity effects and health effects of reduced pesticide use show that the average social cost of production declines in the presence of the insecticide tax. In the presence of the uniform tax, this cost declines until the tax rate reaches 150 percent, and then it begins to increase. The difference in the effect each tax has on social cost is explained by the fact that the uniform tax has a larger opportunity cost in terms of forgone yield and only slightly greater health benefits than the insecticide tax. Thus, beyond a point, the yield losses outweigh the health benefits of the uniform tax. This finding again demonstrates the inefficiency of the uniform tax as compared to the insecticide tax, which is targeted at more toxic and less productive pesticides.
Implications for public policy
The IRRI study establishes that pesticide use has an adverse impact on the health of farmers and that a decline in farmers' health reduces the productivity of rice farms in two major regions of the Philippines. Therefore, policies that reduce insecticide use in Philippine rice production are likely to generate an improvement in social welfare through an improvement in farmers' health. Moreover, policies that reduce insecticide use in the rice production of other southeast Asian countries are also likely to generate an improvement in social welfare because rice production practices in these countries are similar to those in the Philippines.
The actual health benefits from a reduction in pesticide use are probably greater than those estimated in this study. The study's measure of health impairment does not, for example, account for the full social cost of illness, as it likely understates the true opportunity cost of treatment and recuperation. It also does not account for impacts on family members who are not directly involved in rice production but who may suffer impaired health as a result of incidental exposure and accidental poisonings. Taking into account these possible additional effects, as well as the off-farm impacts of agricultural pesticides (such as food or water contamination), could strengthen the general conclusion that it is socially desirable to reduce pesticide use of the most hazardous pesticides, namely insecticides, used in rice production.
The simulation analysis of the uniform tax showed that the social benefits of restricting the use of all pesticides are less than those of targeting the most hazardous ones. Thus, it can be concluded that there could be substantial social benefits from tailoring pesticide policies to a pesticide's toxicity and productivity.
The results of the simulation analysis also have implications for the rates of return to agricultural research and for the allocation of research funds among programs at national and international research centers. The discovery of a significant adverse health impact due to insecticide use suggests that ex ante estimates of the rate of return to general rice research are likely to be overstated when they are based on conventional yield and production cost data. The analysis results also suggest that estimated rates of return on technologies that reduce insecticide use, such as varieties of rice that are resistant to pests and certain integrated pest management methods, are generally understated because they do not include the health and productivity benefits associated with reductions in pesticide use.
Given that many developing countries do not have or cannot effectively enforce regulations that protect individuals from the adverse health effects of pesticides, there appear to be two policy options to reduce these effects: restricting the availability of pesticides or finding alternative methods of pest control. In the long run, the best solution is likely to be the development of effective nonchemical control methods.
But in the short term, the only viable policy may be to restrict pesticide use by imposing, for example, pesticide taxes or restrictions on the importation, production, and distribution of pesticides. The study indicates that the health benefits of pesticide regulations may be obtainable at a low cost in terms of forgone production if the regulations target the most hazardous, least productive pesticides. In the case of Philippine rice production, that means restricting the use of insecticides.
John M. Antle is an RFF university fellow and professor of agricultural economics at Montana State University. Prabhu Pingali is an agricultural economist and a program leader of the Irrigated Rice Program at the International Rice Research Institute. Antle and Pingali helped design the IRRI study and analyze the data collected in it.
A version of this article appeared in print in the January 1994 issue of Resources magazine.
A version of this article appeared in print in the January 1994 issue of Resources magazine.
