In this paper I argue that problems connected to good scientific practice and scientific misconduct can be fruitfully analyzed via the means of transaction costs economics. Thus, transaction costs economics is a valuable tool for the descriptive as well as normative analysis of science. I further argue that current institutional matrices impose high transaction costs on (potential) funding agencies of scientific research. This underlies many cases of bad practice in scientific research. Lowering these transaction costs would have positive effects both (...) on the adherence of researchers to the rules of good practice as well as for the amount of funding that is spend on research. In the end, I propose an institutional reform that promises to lower these transaction costs based upon an idea of Michael Rosenblatt. (shrink)