An anthropologist I know once told me a great story, which may be a rural myth. It was about a remote tribe in Papua New Guinea from which two members were given the opportunity to travel outside of their homestead to see the urban world in all its ‘glory’. When they returned, they recounted their experiences to the rest of the tribe, and they set about replicating one of the more amazing things they’d seen: an airport. They cleared a runway. They built an observation tower out of wood. They even crafted headphones with little reed antennae for the ground control team to wear. When they were done, they waited for the planes to arrive.
They never did. Building the structures, the visible artifices of an airport is only symbolic. The actual meaning of what an airport is, what makes it functional, cannot be seen. It lies in the relations between people and institutions and in agreements between them.
This anecdote constantly pops into my mind when I observe technical reform processes introduced by donors (often with domestic support) in African Governments. Mark Miller and Matt have both discussed this issue in the past. What they and I have in common (apart from devastating good looks and a rapier-sharp wit) is that we have all done time as long term TAs in developing country Governments. All of us have been witness to ambitious reform programmes stalling on the road to implementation or lying dead and ineffective after implementation. Yet only sporadically have the causes of this been critically examined and learnt from.
On this note, The Roving Bandit recently linked to an exciting post from the IMF’s Public Financial Management blog (and yes, I’m aware of the depths of geekery I’ve plumbed by using the word ‘exciting’ about the IMF and PFM). In it, Richard Allen makes a series of simple, reasonable statements about how technical reform should and shouldn’t be tackled in low income countries. Three things that had me high-fiving myself:
The experience of now-developed countries suggests that the process of establishing credible and robust budgetary institutions can take many decades, or longer. There is no reason to expect LICs to be different.
Because the necessary basics are not in place, many reforms are likely to fail.
Much more attention needs to be given to the political economy constraints to reform since changing budgetary institutions is not at root a technocratic issue.
The most important point Mr. Allen makes is the last one, and it extends beyond budgeting. Very few reform processes recognize that at root, the biggest problems in Government administrations are political economy problems. They are not technical or technocratic problems. Treating them as such can simply create new problems without actually addressing the original ones.
Budgeting provides a great example of this. With the upswing in General Budget Support, donors are getting more and more interested in budget procedures in Africa. Almost inevitably, however, they have treated it as a technical issue. The approach is to make African budgets look like European or American budgets look now, not to replicate the process of development of a modern budget structure, which of course took decades in the West. This is a microcosm of the entire development paradigm. The results are rarely pretty.
Take the use of Medium Term Expenditure Frameworks (MTEFs). Through the Roving Bandit, I found Radio Ghana, apparently another long-term TA in an African Government. Ghana is a relatively high performer on PFM issues, yet his account of their MTEF is damning. Budget officers around Africa will recognize this story. Malawi uses an MTEF as well. It is also in the process of undertaking extensive technical changes to the budget, including the installation and adoption of a new software for managing budget figures.
These changes may make the budget presentation a little simpler. They may even make it easier to analyse the distribution of funding. Yet, they could also simply cloud the picture further. The MTEF is normally created out of almost thin air: often it is populated by simply dividing the remaining cost of an activity by 3. Introducing this as an aspect of legitimate budgeting may simply make it harder to analyse how resources are being allocated.
And really, is the biggest problem with budgeting the presentation of the budget? Not according to this excellent report (admission: I am friends with one of the reference team for this report, and a friend and close ex-colleague of one of the writers). In it, the argument is made that the budget is essentially a theatre: a charade in which meetings are held, numbers are discussed and reports are produced, while all of the real decisions and the most important allocation processes take place in a far more opaque and hidden manner. Things have improved in Malawi, but not because of the technical budget reforms. Change of Government did most of that. Without addressing these issues, no amount of new codes, clever databases or web-based analytics will help anyone.
Reform processes aren’t often given much chance of succeeding because they ignore such issues. This has to do with the incentives and politics of the various parties involved. Some of the key issues are:
- Allure of technology: Both donors and Governments can be seduced by the promise of a slick new system that works well in another country, and begin to think if they reproduce this at home the effects will be the same. This is the airport story I’ve mentioned above. But building the system without the right relationships at the centre of it, with right incentives to work in the right way will not achieve anything
- The time frame for interventions: Donors tend to look for fast results, even when they may not be possible. This is partly because planning cycles tend to be short – three to five years. Even within these cycles, they find it easier to use short term consultants than long term technical assistance. The result of this is that reform recommendations come based on a shallower understanding of the issues. A good long term TA can become involved in the whole process of a Government department and recommend further-reaching reforms than a short term consultant will see
- Donors often want to see quick improvements because they have vested interests: in the case of budget reforms, donors involved in budget support often need results to be visible in the near or immediate term, and thus may only address certain surface issues
- Government incentives are not always appropriate: a Government department may accept proposals for a reform it knows will not work because it provides a budget for overseas training, meetings outside of town and other sources of rents for civil servants.
It’s not impossible to make a reform successful, however. Far from it. While I was in Malawi, there was real success in increasing openness about aid receipts and improving aid reporting. I believe successful reforms are more likely if they attempt the following:
- Focus on process, not technology. While technology might help streamline a process or make it more efficient, the first concern must be to understand the real workings of the Government, including the power relations that determine them. Reform must either be compatible with or aim to change these
- Full local involvement. You may have a TA in the reform process doing a lot of the legwork, but without support and full participation by Government they will be insufficiently vested in a reform to fight for it in difficult times
- A long testing and long handover period. If the reform is implemented, it must be monitored for some time; if it’s judged to work, you need to spend time making sure its embedded in the Government. Then you should come back after a few years and see what impact it has made
- You must have local demand for the reform. If no-one wants it, it will not last. Local demand can be stimulated and maintained. As an example, within the Government in Malawi, reports on aid inflows by sector and predictability were produced and disseminated. Their value to other Government departments and donors made data collection easier, since all parties wished for the reports to continue
Reform failure is frustrating for all involved, donors, civil society and Government. A little forethought, a more open mind to the problems and a longer time frame for action can go a long way to improving their success rate. Even then, you need luck. An essential component of a lot of our work that we don’t always acknowledge.