Curiously, I agree with both of you. In the sense that what is being identified here is the fundamental limit to empiricism. Following any revelation of a poorly performing government program, the two statements "it's really bad, shut it down" and "it's really bad, give it more money" are logically consistent and frequently made by opposing groups. So, the key issue becomes "what are the most fundamental things that we need governments to provide", and if we can obtain consensus on this, we should fight and fight to improve administrative capacity...or, at least agree that it's the way forward. Unfortunately, over the last deacade, we seem to be going through what I call "The Great Giving Up", which is precisely that the goods that ought to be provided by the government are being increasingly privatized as we realize that poor governance is the limiting factor. Much of this is because governments are, on the other hand, providing many things that they should not be. My hope would be that if this cash debate finally opens up policy discussion on the role of the government in terms of redistribution and the provision of public goods, that would be a critical step forward.
Jishnu (jet lagged in Birbhum, West Bengal)