Responding in the comments section of Jishnu Das’ post on the WB Future Development blog, I think I went off on a bit of a tangent on ‘being a state’. Converting my comments into a blog-post here, let me try to reorganise my thoughts.
The argument for direct unconditional cash transfers have been broadly positive – but are cash transfers sufficient in a welfare state? Yes, in western Kenya, they seem to have been a huge success; in Liberia amongst drug peddlers and ex-combatants, maybe not so much. Many comparisons of UCTs and CCTs have turned out results in favour of UCTs. But what does this tell us? My reading of the arguments made by the ‘give cash directly’ proponents has mainly been in terms of its focus on external agencies (less so on governments). Given that we waste so much of the resources in delivering aid poorly or by just delivering the wrong aid, it may be better to just give cash directly to the poor and see what they do with it. Even so, the focus seems to be on what is convenient for ‘us’, not necessarily what is good for ‘them’. Extending this logic to governments is not straightforward.
The other aspect of this debate is that of administrative capacity. Delivering cash directly to people when attempted by governments call upon a certain degree of administrative capacity – even if this may be lesser than what is required to deliver services such as education and health. For many reasons that weaken the implementation chain, the transfers may not be well targeted, may not be received in full, and so on. In those cases to say ‘give cash directly’ may not be the solution at all. Without ensuring good quality supply, cash transfers that spur demand alone are not adequate instruments of welfare. Proponents of cash transfers may not be explicitly saying that the government should even withdraw from the infrastructure investments in the social sector, but isn’t there a risk that this is the message that big donors or even poor country governments may be taking out of this?
Also, governments that have access to resource rents and/or aid can consider distributing cash – what of governments that have access to no such resources, and are unable to raise sufficient taxes to redistribute? Governments need to work on improving administrative capacity – and donors and external agencies should work with government towards this.
I think we need to focus on the underlying normative argument here. There is an increasing tendency to ask the government to withdraw – and sure, the government need not make steel, but not stop from running schools – which I have difficulty agreeing with. The state in a poor country is a legitimate representative (in a democracy) and should aim to provide a comprehensive suit of welfare services, and citizens certainly have a right to expect this from their governments. There are those who may want governments to reduce their role to just dispensing cash if that is the minimum they can do well with a certain degree of efficiency – but this I think imposes the political reality of external and non-state agencies on governments, which certainly are a world apart.
I am not junking evidence-based policymaking here; but the normative argument needs to be acknowledged and challenged. Paradoxically, the thrust on rigorous impact evaluations seems to be pushing this notion along that governments should limit their role. Is this a good example of why for policymaking, an overarching theory is more important than just evaluations of small component parts; or is this just ideology over evidence? And even if the latter, is ‘ideology’ always a bad word?