December 2014 How can we measure the effectiveness of our operations and interventions? this is a question that is the focus of several papers in this edition of EDM. In the uK, the Department for International Development (DFID)’s budget has in fact been increasing in the last few years in order to meet its commitment to spend 0.7% of GDP on overseas development assistance. at the same time, other public sector budgets have been reducing and there are very considerable political pressures in the uK to reduce aid rather than to increase it. In the light of these pressures, demonstrating the effectiveness of every penny spent on overseas assistance is seen as imperative. so, the need for accountability is on the rise, and with it the whole panoply of targets, key performance indicators and impact indicators that are part of the ‘results agenda’. maria alinsunurin’s article uses such metrics to size up the Philippines microfinance sector; her approach treats loans and savings as products, staff numbers as inputs, and derives data to compare the efficiencies and outreach of a large number of mFIs using data envelopment analysis. the analysis suggests that MFIs are less efficient than banks, but stops short of examining the characteristics of borrowers, including the relative income levels of bank and mFI customers, which many organizations with a social mission might be interested in. at the same time as the emphasis on measuring performance there is a recognition that ‘development’ in its most fundamental sense is difficult to capture in our metrics; indeed, the more we rely on pre-determined measures, the more likely that deep sustainable development will have eluded our interventions. The microfinance institution that overvalues indicators such as size of portfolio and repayment rates, and incentivizes its project officers to perform well in these measures alone, runs the risk of impoverishing and creating social outcasts of women who are pushed to borrow but who struggle to repay. How we define ‘development’, or conversely ‘the problem’, is critical here. Ben Taylor’s article in this edition alludes to the contested example of insecticide-treated bednets for the prevention of malaria. strategy number one – building the local market for the production and supply of bednets – takes time; strategy number two – importing and distributing bednets for free – results in protection almost overnight. If the desired development outcome is defined as the number of women and children protected by a bednet tomorrow then the second intervention wins out; however if it is the ability to purchase a bednet easily and cheaply for years to come then strategy number one is preferable. Even if the definition of the ‘problem’ is agreed, the way to solve it requires different strategies in different areas among Editorial: Measuring development Clare Tawney