John Kelly’s 2008 report False Positives Equal False Justice sounds the siren about the nonspecificity of presumptive field tests for drugs, explaining in lawyer-friendly language how these tests function, exposing the serious flaws in the existing scientific literature that purports to validate their use, and documenting the wrongful prosecutions that result from the tests’ (mis)use. It’s a must-read.
But in the meantime, here is Just Enough Science to Make You Dangerous. While I’ll focus here on presumptive tests for marijuana, the same concepts and limitations apply across the board. Field tests for drugs are crude investigatory tools. These tests rely on the principle that certain functional groups – small clusters of atoms within the larger molecule that determine its chemical properties – will react with particular reagents to create a particular, ostensibly easily recognizable color change or series of color changes. But these functional groups are not unique to any particular controlled substance; they can be found in a number of licit and illicit molecules alike. The tests look for a structural feature of the molecule, essentially.
For example, the Duquenois-Levine (D-L) reagent, for decades the gold-standard of marijuana field test kits in the United States, does not react specifically with the complex molecule tetrahydrocannabinol (THC). Rather, it reacts with resorcinol, a functional group that is present in the THC molecule, but not exclusive to THC. A substance containing THC will react with the D-L reagents to change color because of the presence of resorcinol, but the presence of resorcinol does not indicate the presence of THC – in fact, resorcinol is a relatively common chemical building block found in Sucrets, many skin treatments, and a host of innocent plants that likely number in the thousands. A study cited by Kelly identified multiple benign, common substances that yield false positives (cypress, patchouli, eucalyptus) or inconclusive results (lavender, oregano, spearmint) when tested with the D-L reagent. In a related Marijuana Policy Project press conference, you can watch a Hershey’s chocolate bar test positive for THC. The tests simply are not specific. They cannot positively identify a substance.
Resorcinol will react with the reagents in the D-L test to produce, in sequence, “a purple color,” then “a layering with dark purple on top and light purple on bottom,” or a “dark blue or violet” followed by a “grey upper layer over a violet layer,” or a “slate-grey upper layer over a purple lower level,” or “a blue-violet or purple color” that is extracted “into the lower layer,” depending on how the manufacturer describes a positive result or depicts it on the packaging. (Compare this with this with this.) Moreover, the color the reagent actually produces in practice may vary depending upon the concentration of the drug, its form, or the presence of contaminants, as this oft-cited but Kelly-debunked study that purports to validate these tests freely admits. If the officer makes the mistake of allowing the test to overdevelop, they will likely get a false positive, as the chemicals will develop a darker purple color the longer they are in contact.
Reasonable interior decorators could disagree about what constitutes slate-grey over purple, what distinguishes blue from a proper violet, or when a certain hue is really fuchsia, not purple. How, then, can we expect police officers to consistently interpret these test results? The test results are wholly subjective and depend upon the color perception, training, and inclinations of the officer conducting the test. Further complicating the matter is the potential for operator error (overdevelopment, for example, or breaking the ampules in the wrong sequence) and the fact that field tests – conducted, as they often are, in the field, perhaps in the dark, at the side of a busy road – are frequently administered in conditions far from ideal for discerning subtle color changes. And Officer Friendly is susceptible to confirmation bias: he is testing the substance because he already believes it to be marijuana, so if it’s a close call between magenta and purple, he’s likely going to see purple.
The tests’ limitations are well-known to police officers – but as you hopefully now see, positive test results may be readily assailed by a defender armed with just a little bit of science.