Freakonomics, a book cowritten by the U Chicago economist Steve Levitt, is a treatise on the economics of ethics. I am in the school of belief that morality and ethics are separate but not mutually exclusive sets of… er…. lifestyle choices. Ethics, I think, can be explained in economic terms where as morality is ethics beyond economics (matters of the soul and whatnot) – but that’s a whole different blog.

I purchased a copy of Freakonomics on Friday (the extended edition) and had the whole thing read in about 4 hours. It’s a simple, if not disjointed read, and is packed full of analyses of people engaged in unethical behavior as explained in economic terms. It seems as though, at least on occasion, it is more economical (in terms of risk to benefit, cost analysis, penalties of social isolation, etc) to cheat and be “unethical” than it is to be ethical and not cheat.  I.E. cheating is not the result of errant random human behavior, but calculated actions performed (rather predictably) by large groups of people. It is reasonable, for instance, for a teacher to cheat the No Child Left Behind act by doctoring her student’s answers to make them look higher. This way, she won’t get fired for being a shitty teacher. While the idea behind No Child Left Behind is to generate better teachers, the easiest fix is obviously to make the students look smarter. A quick query to Mrs. Finchsigmate, a teacher herself, led me to believe that teacher cheating is ubiquitous – teaching to the test, giving more than the alloted time, spoon feeding answers to kids… shit Levitt would be able to detect through simple regression analysis but for which there is no smoking gun.

This leads me, as you might imagine, to the economics of ethics in science. It’s far more economical for a researcher to self plagiarize if the currency is number of articles in a tenure decision, especially when the amount of grant money is low and the act of self plagiarization is already rampant and the consequences of being caught are small. Cheating is likely as ubiquitous in science as it is in education, but the small, venial infractions do not cause massive mistrust by the public and serve only to inconvenience those of us in the field (if they ever see the light of day to inconvenience anyone). It is the huge, cardinal infractions that cause public misgivings toward science and, in return, public skepticism and potential decreased funding for promising areas of research. Somewhere along the line, science went from being a route of curious exploration by upper middle class men and clergy to being a full blown, trillion dollar a year industry (most likely the late 19th and early 20th century in Germany with the dye and aniline boom). Suddenly science wasn’t seen just as an intellectual exercise intended to impress a small population but an industry where people could go from rags to riches. Indeed, it wasn’t until recently that the notion that it is possible for one to simply make a single, small organic molecule that would in turn generate over 13 billion dollars a year. Discoveries made in the lab by a team of underpaid graduate researchers can net PIs millions (a quick walk to MIT, where a certain 100+ man group comes to mind). The economics of honesty aren’t what they used to be, clearly. People haven’t changed (they never do) – their incentives change.

The only way to insure the integrity of data, now that the stakes are higher, is by assurances of data audits. Not infrequent data audits, but random high percentage data audits with access to FIDs, reflection images from crystallography measurements, frozen samples of cell lines, a minimum of 50 different images in a single well for microscopy, samples of final products, mandate inclusion of full NMR spectra from -2 to 14ppm, etc. This sort of shit will keep unscrupulous drug companies from ghost writing and publishing cherry picked data. It will also keep overly ambitious but intellectually lazy researchers from doctoring data at their leisure. It will keep the embryo cloners and CH activators of the world on guard. A police state. Something none of us want. Having just finished up an Org. Lett. myself, I found the process of collecting every spectra tedious. Not hard, mind you, just a pain in the ass.

On the flip, if scientists are going to be expected to be so rigorous in their trials, the journals must do their part. Not just in terms of running computer algorithms on figures, but also performing double blind reviews of papers to ensure fairness as well as provide (anonymous) reviewer commentary with the article so everyone can appreciate the points a reviewer brings up. Embarrassing? Possibly. Sometimes reviewers hit on really smart things. Other times, they’re retarded assholes who didn’t bother to read the paper. Why shouldn’t those documents be available? They are valuable critiques of the work by leaders in the field.

Fear of being caught is a powerful motivator and, regardless of what naïve notion you have of the “nobility of science” it is chalked full of intelligent people, motivated by things that aren’t necessarily in your best interest. Cheating will happen, some of it we must admit to, but with the appropriate measures in place, we can be assured (to some extent) that we are doing the best we can in this high stakes game.