Since I was recently working on issues relating to Lesotho, industrial policy in particular, I spent some time trying to get some background reading done in addition to the HUGE amount of economic reports/strategies/plans/studies/policies carried out in the last few years. Just as a taster, we're talking: a Private Sector Development Strategy, a Growth Strategy, an Industrialisation Master Plan, an SME White Paper, an Investment Policy, a report on the State of Small Enterprise... And what's more, the same things get said repeatedly - even down to dealing with why implementation is so poor even when implementation matrices are detailed, institutions are given clear responsibilities etc.
On that note, Bill Easterly wrote a blog where he mentioned the Anti-Politics Machine by James Ferguson. Anti-politics because the assumptions underlying aid projects are often that the government has no political drive or incentive other than that which alligns perfectly with the goal of development. In reality this clearly is not the case and there are a whole host of other factors which affect whether or how policies are implemented. This is also something which I'm now reading abouty in Van de Walle's African Economies and the Politics of Permanent Crisis. So although I'm new to this (but probably should at least has read Van de Walle's book before now!) it seems pretty clear that we know that aid in the form of policy documents and attempts to create the capacity to implement our policy documents don't work. What's perhaps worse though is that even partial reforms have unintended consequences, with winners and losers - and given the selection of which aspects to implement are taken by politicians, presumably they or their clients are the winners, while the rest are the losers, therefore potentially worsening the situation for the intended beneficiaries of the reform (unless there is some kind of win-win situation...).
This all makes it pretty hard to continue working in this kind of area.