Quantity based indicators fail to identify extreme pesticide risks.

As a matter of policy, minimizing human health and environmental risks associated with pesticide use is a major challenge but necessary for improving agricultural sustainability. Efficient and effective policies that encourage the use of less risky pesticides, such as pesticide taxes, necessitate a precise and realistic quantification of potential adverse effects. Various indicators are currently utilized in policies and they focus mainly on a purely quantitative dimension of the pesticides used, which can lead potentially to unfavorable outcomes of pesticide policies. A unique dataset applied to pesticide use by Swiss farmers in winter wheat and potato production, demonstrates that on average the two most important quantitative indicators show a significant correlation with pesticide risks as expressed by the Danish Load Indicator. However, they have almost no explanatory power for extreme risks (e.g. most intensive use patterns for pesticides with unfavorable toxicity profiles). Results remain stable over a range of aggregation levels, from application- to farm-level indicators of pesticide use. These findings render the commonly used, quantitative indicators ineffective to reduce potential environmental and human health risks of pesticides and, in the worst case, lead to misinformed market-based pesticide policies consequential to National Action Plans.