Comparing the content of the first and second generation of PRSs highlights the following:

·   There is a much more prominent role for growth in second generation PRSs. Optimistic growth targets feature as much in the second round as they did in the   first.

·   Increased use is made of sectoral modelling in growth targets in the second generation.

·   Growth strategies in both rounds of PRSs are characterised by a stronger emphasis on measures to address vertical (within sector) challenges thanhorizontal (cross-sectoral) ones.

·   There is increased attention to growth-poverty linkages in the second generation through distributional analysis and attempts to adjust growth policies in ways that benefit poor people and vulnerable groups.

·   Analysis of constraints to growth within sectors and links between this analysis and policy actions has improved. Analysis of constraints across sectors remains weak.

·   Trade has a more prominent place in second generation PRSs but they retain the earlier emphasis on export promotion rather than market access issues.

·   Discussion of trade policy options and trade-offs for poverty reduction  continues to be weak.

·   Both first and second generation PRSs focus much more on supply-side constraints to trade than demand-side constraints, although there is still a lack of clear connections between supply side factors and trade opportunities.

Early explanations for the relative neglect of growth and trade policies in early PRSs centred on the role of donors in prioritising social sector expenditure, but three other factors also appear to play a significant role in shaping constraints and opportunities at country level:

1.     Pro-poor reforms to growth and trade are politically more difficult,especially in Africa’s low-income, aid-dependent countries which have some of the characteristics of neo-patrimonial states. The PRS approach has sought to address patronage politics by encouraging a more open and consultative form of policy-making and this has led to some improvements. However, more fundamentalpolitical reform is not always suitable for external initiatives and is unlikely to be achieved over a short timeframe.

This points to a need for greater realism about the prospects for promoting pro-poor growth and trade in Africa. Expectations of immediate results would be better replaced by a medium- to long-term trajectory. Aspirations for wholesale policy transformation need to be countered by acceptance of the incremental nature of reform and a focus on the limited number of policies that it is feasible to reform in favour of poor people.

2.     Lessons from East Asia suggest that the key to economic development maylie not in the broader dimensions of good governance but in state capacity to develop effective policies for growth and trade in a limited and specific number of areas. Unfortunately, current donor efforts to support the development of effective growth and trade policies are not focusing on specificareas of state governance.

An alternative approach would involve extending lessons about aid effectiveness inthe social sectors to national systems and institutions directly affecting growth and trade. The role of aid in determining growth and trade outcomes isnecessarily limited, while programmatic instruments such as general budget support and SWAPs are proving more effective for social causes such as health and education.

3.     In squaring the circle between political constraints and policy gaps onpro-poor growth, there continues to be a clear role for bringing evidence to bear in the decision-making process at country level. Yet donors themselves have historically provided to be notoriously weak at drawing on evidence as a basisfor their own policy positions on growth and trade. Instead of advocating policies tailored to country context and based on quality research, their efforts have all too often been driven by political dogma and nationalself-interest. This has contributed to a climate of mistrust between governments and donors in dialogue around growth and trade.

More effective donor support for pro-poor growth and trade in Africa is likely todepend on reforms at a number of levels.

·   First, there is a need to reform trade diagnostics initiatives such asthe Integrated Framework to ensure they are much more closely sequenced withnational political, planning and budget cycles, and allow space for national decision-making.

·   Second, where appropriate, there may be a case for extending programmatic aid modalities beyond the social sectors to include neglected sectors and cross-cutting issues directly affecting growth and trade (e.g.through SWAPs).

·   Third, there is a need to support further research into what works for pro-poor growth and trade at country level, ideally in collaboration with otherdonors.

·   Beyond these immediate challenges lie more fundamental obstacles toimproving donor support for pro-poor growth and trade in Africa:

·   Perceived pressure from domestic taxpayers to deliver tangible and immediate results from aid spending is currently creating a climate of expectation which is particularly unrealistic in the context of politically contentious growth and trade reforms. More effort is needed to educate citizensin donor countries about the incremental and long-term nature of development efforts in Africa.

·   Divisions between development and trade and industry ministries in donorgovernments can create inconsistencies between the policy positions taken atcountry level and those pursued through international fora such as the WTO.More work is needed to ‘mainstream’ development thinking across governments and ensure greater coherence between aid and trade policies.

·  The internal organisational structure of most donor agencies,particularly at country level, is highly conducive to creating sector andinstrument ‘silos’ which discourage joint working between those responsible for supporting broader national planning and budget processes and professional specialists such as trade advisers. DFID’s recent internal reforms aimed at overcoming these divisions. On the other hand, purely national development programmes are unlikely to be fully consistent with cross-border issues related to trade and Aid for Trade.