Incentives, yes, but cultural and social norms matter as much as well

From At least five interesting things to start your week (#38) by Noah Smith.  The subheading is WW2 production; "Neopopulism"; fake science; desire modification; deregulation ideas.

And yet the incentives of modern science seem almost designed to encourage widespread fraud. The key metrics for success in science are 1) publications in peer-reviewed journals, and 2) citations of those publications. And as Goodhart’s Law tells us, all metrics will eventually be gamed. There are many ways to game the publish-or-perish system — p-hacking, specification search, citation rings, etc. Fraud might not be the easiest or safest of these. But it’s almost certainly the most damaging.

Peer review is a poor defense against fraud; even careful, diligent reviewers who might spot bad statistics or weak inferences will struggle to identify outright fakery. That leaves culture — a general feeling of pride in the integrity of the scientific system — as the front-line defense against fraud. There was never any assurance that this culture would remain intact as larger and larger sums of money got thrown at science.

 

There is a hierarchy of social control.  The most onerous, expensive and least desirable is Law.  Absolutely necessary but not easily tuned to the pace and micro evolution of human actions.  

Next are cultural norms, usually arising from a mix of religion and history.  Cultural norms ensure, to a degree, an alignment of goals, values, and behaviors.  Where it is in effect, it tends to be effective at preventing conflict in the first place.

Then there are social norms within a community, class or extended family.  One can include expected social manners and etiquette in this category.  These tend to be very effective at smoothing interactions between and among individuals.  They involve little explicit governance cost but are good both at prevention, justice and restitution.  It is emergent in form and far more dynamic than cultural norms.

Fraud in science is a legal matter but, given the importance of science in our Age of Enlightenment culture, it is also a matter of cultural norms.  Those who undermine the reliability of and therefore trust in the scientific process are very directly attacking the Age of Enlightenment culture.

Simultaneously, the absence of cultural and social norms within the university environment has facilitated the expansion of scientific fraud to the point of profound and extensive distrust.  The appeal to "experts" and desire to "stand with the science" become cynical efforts at coercion rather than a commitment to discovery and progress.

Incentives indeed matter but cultural and social norms matter as much as well.  Probably more so.