Donald Kornfeld and Sandra Titus contend that implementing known remedies for scientific misconduct is better than calling for further research (Nature 548, 31; 2017). I disagree: we need research to determine whether such measures are working.

Just as the presence of more police officers leads to more arrests, the expansion of efforts to counter scientific misconduct is exposing more cases. Institutions already implement punitive measures, operate offices for research integrity and run ombudsman systems. Journals retract papers more frequently than they used to, and scandalous cases are widely publicized online (see F. Hesselmann et al. Curr. Sociol. 65, 814–845; 2017). Prevention tactics include setting up university commissions for good scientific practice, and introducing regulations for responsible research.

Still, the detection of more cases will stimulate calls for ever-harsher countermeasures. However, criminological research has shown that such law-and-order policies have devastating effects — for example, by discriminating against people who are already disadvantaged (see M. Cavadino and J. Dignan The Penal System; Sage, 2013). Although such policies have political appeal, they can be ineffective and unjust.

What is needed is a measured response to the problem of scientific misconduct, which can stem only from more research.