Over the last ten years there has been a quiet revolution in how the International Monetary Fund (IMF) operates in the poorest countries. IMF conditionalities have eased and are increasingly focused on good financial governance. And its macroeconomic frameworks are more accommodating to governments’ long-term development plans. As one official in an African Ministry of Finance put it, the IMF has turned from ‘dictator to partner’.  The IMF is now considering further changes, especially in relation to fragile states. This is what I’d like the IMF to be saying to its mission chiefs:

1. Focus on results – At the end of each mission the IMF should be clear on what difference the mission has made. For many countries the biggest impact is just getting the money to flow – providing the assurance donors need to release their funds or helping the government to ensure its own budget is well executed and the money reaches the intended beneficiaries.
2. Focus on the causes of fragility – In many countries there are just one or two key causes of past conflict and instability, such as large numbers of unemployed young men and lack of development outside the capital city. While the IMF can’t generate jobs or shift government investment, IMF missions need to support government efforts to address these fundamental political risks. There is no point in helping governments secure macroeconomic stability if this fails to address causes of future conflict. In some countries supporting the government to decentralise its budget should be the top priority, rather than being in the ‘nice to do if time’ category.
3. Focus less on new laws – This was one of the top requests from the fragile states at last year’s joint IMF-ODI conference. New laws look impressive in an IMF programme, but there are often alternative administrative changes that could achieve the same result more rapidly and with less government effort.
4. Go for speed and economy of effort – Government capacity is incredibly weak in fragile states so the IMF should find ways to achieve results without adding to the burden. The IMF should also find ways to achieve the same effect faster. For example, the political demands for a visible peace dividend soar after conflict. A regional solution may be the best option – drawing on the capacity of regional institutions.
5. Don’t exaggerate the fiscal deficit – In many fragile states donor grants are financing over half the budget. With most countries taking at least 10-15 years to graduate from fragility and most donors prioritising fragile states, such flows are likely to continue for years. Most donors and governments plan accordingly. But when the IMF talks about the fiscal deficit it often refers to the gap between total expenditure and the government’s own revenues, omitting all the grants. So the fiscal deficit appears huge and unsustainable. This measure only makes any sense if all grant funding were going to end the next year. The better measure includes grants as part of the revenues, resulting in a smaller and more manageable fiscal deficit – or even a small surplus.
6. Don’t block critical long-term reforms – Sorting out the pay structure for civil servants – to ensure everyone has a living wage – is critical before any long-term training and capacity-building can take place. Yet discussions of long-term pay reform rarely make it past first base. Government officials are too aware of the IMF views on keeping public sector wages ‘under control’ and hesitate to raise the issue in IMF negotiations.
7. Collect data on jobs  – Jobs are highlighted as the priority issue in fragile states, yet there is little or no data on this issue. The IMF works with national authorities to produce increasingly sophisticated estimates of gross domestic product. But if GDP growth doesn’t deliver job growth it soon becomes a deep political problem. The IMF should prioritise data collection that is aligned to partner governments’ needs – not the requests of IMF statisticians back in headquarters.
8. Publish as much as possible – Rebuilding a state means rebuilding the trust between government and citizen. In the early stages the IMF can play a direct role by publishing its analysis and its recommendations. In later stages it can shift to an indirect role – supporting national statistical authorities in publishing and validating such data. And in fragile states where even the Ministry of Finance may not have internet access – or can’t download a 2MB  file – publication is not just about putting something on the IMF’s website.
9. Help with implementation, not just advice – Fragile states have no shortage of well intentioned advice. One of the challenges for the new government in Southern Sudan will be to manage the deluge of advice it is about to receive.  Donors, foundations and freelance consultants are falling over themselves to advise the world’s newest country. But, as the lack of roads or hydro power in Liberia shows five years after the new government was elected, there is a shortage of implementation capacity. It is good that the IMF will place additional resident advisers in the Middle East to follow up and support the implementation of past IMF missions. It is bad that the IMF is not placing any of its own experienced staff in its new regional technical assistance centres, but is filling the posts with new recruits on short-term contracts.
10. Set performance criteriafor the IMF – The IMF is constantly setting deadlines for countries, but rarely reciprocates. One deadline that matters is how fast a new country can join the IMF. If Southern Sudan wants to join, could it demand membership by December 2011? The country has to be an IMF member before it can join and borrow resources from the World Bank, so the sooner the better. It would also be a nice role reversal – and a great way to show that the IMF really has changed the way it works with fragile states.