Acreage and experience may alter the environmental impact of growers’ pesticide programs

For organic growers, two AIs, spinosad and pyrethrins, are available to target those physiological functions. The “unknown” category, which is mostly sulfur, accounted for a significant portion of treated acreage in organic agriculture. Insecticides that target the midgut, which includes Bacillus thuringiensis and several granulosis viruses, are widely applied in organic fields. Conventional growers rarely use them due to the high cost. In 2015, acreage treated with midgut targeted insecticides was 1% of total treated acreage in conventional agriculture and 24% in organic agriculture. A detailed discussion of insecticide and fungicide use by mode of action in conventional and organic production is in the appendix.Insecticides and fungicides in the two pest management programs have different modes of action and pose different levels of environmental impact. Simply comparing treated acreage or the amount of pesticide products used does not identify the differences in environmental impacts. In this context, the PURE index serves as a consistent measure across farming systems.Figure 1.3 plots PURE indices for conventional and organic fields by year. Index values for air and soil are significantly higher than those for the other environmental dimensions in both farming systems, which means that pesticide use in general has greater impacts on air and soil quality than groundwater, pollinators, and surface water. Risk indices of conventional fields are relatively stable from 1995 to 2015, with no obvious overall changes for air or soil, despite the many changes that have occurred during this 20-year period in regulations and grower portfolios. While PURE indices decreased 16% for surface water, 26% for pollinators, grow bag for blueberry plants and 7% for groundwater over the same time period, these three were much less impacted by pesticides in 1995, the beginning of the study period.

Despite the numerous regulatory actions designed to reduce environmental impacts over this 20-year period, such as the methyl bromide phase-out, large-scale substitution of pyrethroids for organophosphates, and regulations to reduce VOC emissions from non-fumigant products, the overall environmental impacts of conventional pesticide use show only limited reductions when aggregated across all crops. PURE indices for organic fields are similar to conventional fields in that the air and soil have significantly higher index vales than the others. However, the aggregate risk indices in all five dimensions are much lower in organic fields. Compared to conventional agriculture, organic agriculture has dramatically lower PURE indices for surface water , groundwater , air , soil , and pollinators . The reduction for air varies greatly across major California crops. Large reductions in the PURE index for air are observed for table grapes , wine grapes , and processing tomatoes , while others had relatively small ones such as leaf lettuce and almonds . The reduction in the PURE index for soil varies across crops as well, ranging from leaf lettuce to carrots . For surface water, groundwater, and pollinators, the differences between the PURE index in organic and conventional fields are similar across crops. A noticeable spike in PURE indices appeared in 1998 for organic agriculture caused by a single application of copper sulfate with an application rate of 150 lb/acre, which is ten times larger than the average application rate and clearly a data abnormality. The PURE index is a measure of environmental impacts on the per acre basis.One could use the yield difference between conventional and organic agriculture to adjust values in Figure 1.3 and transfer them to a measure of impacts per unit of output. Organic agriculture is found to have 10%-20% lower yields than conventional agriculture . If we use the 15% yield loss as an average to adjust the results for all crops, organic agriculture reduced the PURE index for surface water , groundwater , air , soil , and pollinators .

The impact of organic practices on pesticide use is crop specific. This aggregate result is derived based on current crop mix in California. Each crop is susceptible to a different spectrum of pests, which are managed by a distinct pesticide portfolio as part of a broader pest management program. Comparing PURE indices for individual crops shows the benefit from pesticide use in organic agriculture varies significantly. Based on value, production region, and the acreage share of organic production, four crops are selected to illustrate this point: lettuce, strawberries, wine grapes, and processing tomatoes. Lettuce, strawberries, and wine grapes are the three highest-valued organic crops in California, with organic sales values of $241, $231, and $114 million in 2016 respectively . Production of strawberries and lettuce is concentrated in the Central Coast region. Processing tomatoes are an important crop in the Central Valley. Wine grape production occurs in a number of regions across the state. In 2015, the acreage shares of organic production are 8% , 9% , 4% , and 2% for the selected crops.For my analysis, the unit of observation is a field-year, defined as a field with one or more pesticide applications in a given calendar year. In total, more than 3 million field-year observations are included in the PUR database from 1995 to 2015. Table 1.1 provides field-year summary statistics for key variables by crop. Overall, 3% of them applied only pesticides approved in organic agriculture. For all crops, conventional farms are significantly larger in size and have higher PURE indices. The average farm size in PUR is smaller than the average number in the USDA Census . One potential explanation is that one farm could have fields in different counties and apply for multiple pesticide application permits within in each county, which classifies it as multiple “farms” in the PUR. For all crops, lettuce, strawberries, and processing tomatoes, growers who operate conventional farms have significantly more experience, measured by years they are observed in the PUR. For wine grapes, conventional growers have less experience than organic growers. Ideally, farming experience is measured directly or researchers use age as a proxy. However, the PUR database does not contain any demographic information, which limited my ability to measure experience. The PUR experience is smaller than the farming experience reported in the Census, which has many reasons. . First, the PUR database I use started in 1995. Any farming experience before 1995 is not recorded. The Census is conducted every 5 years. Farms that entered and exited within the 5 year gap are included in the PUR database but not the Census, which reduce the average experience. Conventional strawberries have significantly greater impact on surface water and less impact on groundwater, measured by the PURE indices, comparing to other conventional crops. Organic strawberries, on the other hand, had a higher PURE index for air and a lower PURE index for soil than other organic crops. Pesticides used in conventional production of wine grapes have less impact on pollinators than pesticides used in other conventional crops.To identify the effect of organic agriculture on pesticide uses and associated environmental impacts, I must address the issues of selection bias at both the grower and the field levels. Compared to growers who utilize conventional practices, growers who adopt organic ones may have different underlying characteristics, such as attitudes toward environmental issues, which can also affect their pesticide use decisions directly. If grower characteristics are time-invariant, an unbiased estimation could be achieved by including a grower fixed effect in the regression.

There is also time-variant heterogeneity that is associated with individual growers, due to factors such as farm size and experience, blueberry grow bag that simultaneously influences the adoption of organic production and pesticide use decisions. The identification concern here is that growers with more farming experience or larger farms, including both conventional and organic acreage, are more likely to operate organic fields and use less pesticides . Therefore it is not reasonable to compare environmental impacts of pesticide use for growers without considering these characteristics. For each grower, annual total acreage and experience serve as measures of time-variant heterogeneity. As shown in Table 1.1, there is a significant difference for these two variables between conventional and organic growers. There could be field-level heterogeneity as well, due to pest or disease pressure, that undermines my identification strategy. Fields with less pest or disease pressure need less pesticides and are more likely to be converted into organic production at the same time. Including field fixed effects in the estimation is one approach to address these issues. Organic fields tend to be concentrated spatially to avoid pesticide drift from nearby conventional fields . Spatial relationships are not considered here because the PUR database does not have information on the distance between fields.For all five PURE dimensions, pesticides used in organic agriculture reduced environmental impact. The reduction, captures by the variable Organic, is significant at the 1% level for five environmental dimensions. Relative to the intercept, organic practices reduced environmental impacts for surface water by 86%, for groundwater by 93%, for soil by 60%, for air by 53%, and for pollinators by 76% on a per acre basis holding other variables fixed. The relatively small impact on air is linked to the facts that natural AIs do not have less VOC emissions in general. Regulations regarding high VOC-emitting pesticide AIs also contribute to this result partially because they do no affect two systems evenly. In 2015, the sale and use of 48 pesticide products were restricted due to their VOC emissions, which accounted for 5% of treated acreage in conventional agriculture and 1% of treated acreage in organic agriculture. Although reductions in PURE index values do not translate directly into dollar values or health outcomes, results from Table 1.2 suggest that pesticide use in organic fields substantially reduced environmental impacts. The coefficient for Organic × t represents the change of the difference between two farming systems over time and is positive for all environmental dimensions, which supports the hypothesis that, comparing with conventional agriculture, the environmental impacts associated with pesticide use in organic agriculture have grown over time. Air has the largest coefficient among the five environmental dimensions, which is consistent with previous figures that environmental impacts increased the most for air across all crops. The variable t is the common time trend for all conventional fields and t is negative for surface water and groundwater, which means the environmental impacts from pesticide use decreased in conventional agriculture on those dimensions. The environmental impact on soil and air increased. The combination of variables t and Organic × t shows the time trend for organic fields alone, which is upward sloping for groundwater, soil, air, and pollinators, and downward sloping for surface water. Two variables Acreage and Exp, capture time-invariant grower heterogeneity. Although the variable Organic dominates the overall effect, coefficients for both Acreage and Exp influence the environmental impact associated with crop production. For the same grower-crop combination, a larger farm size is associated with pesticide application pro-grams that pose more negative impacts for all five environmental dimensions. Meanwhile, more experience is correlated with the environmental impacts on soil, air, and pollinators. The PURE indices for surface water and groundwater are positively correlated with experience. This is partially due to the fact that experienced farmers use less organophosphate insecticide per acre, which are more toxic to earthworms and honeybees than alternative AIs.The sub-sample estimation yields similar results . Namely, in conventional agriculture, the environmental impacts on surface water and groundwater associated with pesticide use decreased over time, pesticides used in organic agriculture significantly reduced the environmental impacts measured by the PURE index, the difference between conventional and organic pesticide use decreased.The intercept is smaller than the coefficient of Organic occasionally because the crop and time fixed effects are oftentimes positive and significant and the impacts on those dimensions in organic fields are small. For the sub-sample with fields that have transitioned between production systems, total farm acreage is no longer significantly associated with impacts on groundwater, soil, and pollinators and the environmental impact on surface water is negatively correlated with farm acreage. The main reason for this seemingly dramatic difference, comparing to the full sample estimation, is that there are more wine grape vineyards and fewer almond orchards and alfalfa fields in the sub-sample. Although the organic price premium is limited for wine grapes, the organic farming practices are associated with high quality of grapes, which encourage growers to adopt organic production .