Of the many excellent notions Dean Karlan brought up during his USAID lecture last week, one had particular resonance.  Prof. Karlan expressed his wish that practitioners, who likely made up the majority of his audience at the USAID seminar, view impact evaluations (& RCTs) as investments rather than just accountability measures. 

For instance, an impact evaluation finds that some microfinance clients use credit to start new businesses, while others use it to dig themselves into debt with non-productive activities.  In this hypothetical the average impact of credit is just slightly positive, but a microfinance institution that sees results as an investment might recognize the heterogeneity in credit usage as an opportunity for operational tweaks (e.g. if the MFI's objective is to encourage entrepreneurship, they might consider additional credit scoring/training/monitoring of clients). Here the research results open up new operational possibilities instead of shutting the door on a particular intervention. 

MFIs and development organizations with whom research teams partner should never feel that an impact evaluation is being, "done to them," but rather with them and ultimately, for their benefit.  Here the onus is partially on the research team to convey both the importance of rigorous evaluation as well as how evaluation results can be unfurled for improved practice.