Forbidden Fruit | Illegal Pesticides in the US Food Supply: Appendix: Methodology
Concerned by this discovery, we began a systematic audit of FDA violations. To do the job properly and efficiently, however, we needed a database which contained pesticide tolerances for each crop/pesticide combination that existed in the FDA monitoring database.
Because it would be most efficient to work from the same database as the FDA, we asked the agency to provide us with the pesticide tolerance database it uses in its regional labs. This attempt, however, was derailed by the discovery that the FDA has no computerized database of pesticide food tolerances (Jones 1994), and thus the agency has no way to run its pesticide residue test results through an updated computerized system that will automatically check for violations. Instead, each laboratory relies on one or two individuals familiar with pesticide tolerances to manually check each lab result with the most recent information available to identify illegal pesticides in food. These procedures leave enormous room for human error and great latitude for discretionary enforcement of U.S. law.
Because no database existed that would enable us to determine which samples contained illegal residues, we created one. The Environmental Working Group database is limited to the 42 fruits and vegetables analyzed in this study. It contains tolerances, special exemptions, and action levels for all of the pesticides detected on these foods during the years 1992 and 1993. It is based on pesticide tolerances enumerated in the (Code of Federal Regulations (CFR) Title 40 Parts 180, 185, and 186) from 1992 and 1993, amended to reflect any changes to these tolerances that may have occurred during these years and would thus not have been published in the CFR (such as those reported each week in the Pesticide Chemical News Guide), modified to include information from the EPA on so-called emergency exemptions allowed under section 18 of the Federal Insecticide Fungicide and Rodenticide Act (FIFRA) (Fried 1995), and updated to cover canceled pesticide uses (EPA 1990). We refined the database numerous times to account, as the FDA does, for the incidental presence of industrial contaminants, allowable levels of banned pesticides which degrade slowly in the environment (i.e. DDT, endrin, heptachlor), and for metabolites (breakdown products of pesticides) (Wessel 1995).
Using this database, we compared the pesticides that the FDA identified on these 42 crops with the pesticides and levels of these pesticides legally allowed. We found 381 cases where the FDA failed to report illegal pesticides as violations.
To verify that we had simply not discovered an anomaly in the FDA database, we calculated the average violation rate for these 42 fruits and vegetables on the basis of those pesticides reported by FDA as violations. The result was a 3.1 percent violation rate, similar to the rate the FDA routinely reports to the public. This result confirms that the illegal pesticides identified, but not reported by the FDA, represent legitimate underreporting of violations on the part of the agency.
The FDA Data
The Environmental Working Group prepared this analysis by examining the latest available pesticide residue data from the routine pesticide surveillance program of the U.S. Food and Drug Administration for fiscal years 1992 and 1993 (referred to as 1992 and 1993 in the report). These data provide residue results of surveillance (random) and compliance (known violators) tests on approximately 30,000 samples of imported and domestic food that the FDA analyzed during 1992 and 1993. The commodities covered include all foods except meat, eggs, and dairy products, which are monitored by the U.S. Department of Agriculture.
For purposes of this report, we limited our analysis to the 14,923 samples of 42 different fresh fruits and vegetables which Americans eat the most. These samples are derived from routine surveillance test results for the two-year period, where produce is selected at random from the food supply. Compliance samples, those aimed at problem growers, were eliminated from our analysis.
According to the USDA, the fruits and vegetables analyzed in this report comprise 96 and 83 percent respectively of the total consumption nationwide (USDA 1993b). Most of these commodities are both imported and grown domestically. Overall, 38 percent of fruits and 7.5 percent of vegetables are imported (USDA 1993b). This report examined all samples of the 42 foods targeted, regardless of origin.
FDA's database contains a host of information specific to each individual sample. The data utilized in this report include:
- the product sampled,
- the country of origin,
- the pesticides FDA looked for,
- the pesticides found,
- the level at which the pesticide was found, and,
- whether FDA found the residues in violation of U.S. food tolerances.