Speaker:

Daniel Kaufmann, Director of Global Governance, World Bank Institute

Discussant:

Verena Fritz, Research Fellow, ODI

Chair:

Julius Court, Governance Adviser, DFID

Daniel Kaufmann

Daniel Kaufmann started by outlining that his presentation would cover:

• the evolution of the World Bank's (WB) governance and anti-corruption strategy (GAC)

• where we are now with GAC

• the main pillars of GAC

• fallacies and misconceptions about GAC

• debates about GAC

• the power of data

Outlining the evolution of GAC, Kaufmann explained that in the mid-90s there was a prohibition to mention, let alone directly to work on governance at the WB. Only when James Wolfensohn arrived as president in 1995, was corruption taken up as a major issue. Over the past ten years, work on corruption and governance more widely has grown substantially. However, as the governance indicators which started to be collected in 1996 reflect, across countries overall progress has been limited over the past ten years (some countries have seen governance improvements, while governance has deteriorated in others). This has led to calls for a new GAC strategy by the WB's new president, Paul Wolfowitz.

He explained that governance and anti-corruption are not synonymous. The set of institutions one needs to look at to examine governance is much broader than those for anti-corruption. Rather, corruption is one important aspect of poor governance.

In outlining the need for a GAC, Kaufmann explained that there were three main factors in this process. Firstly, citizens and government reformers from emerging economies have been asking the WB 'where have you been?'; secondly, the evidence that better governance is a key issue in improving aid effectiveness; and thirdly, there is a new global environment regarding governance and anti-corruption.

In outlining the main pillars of GAC, Kaufmann emphasized that it was very much in support of the wider goals of development and poverty alleviation.

There are 3 key levels on which the GAC is intended to operate:

• country (systemic issues are crucial)

• project (this involves both sectoral expertise and investigations)

• global and local partnerships (this involves donors, both bilateral and multilateral - a new agreement was reached on this systemic-level collaboration in Singapore; NGOs; a network of reform champions; and alliances of private sector groups)

In outlining the next steps for the GAC, Kaufmann explained that this was a subject for debate. The strategy paper was approved in Singapore, but the details of its implementation were not spelt out therein. It is expected that there will be a report on the strategy for the Spring meetings in April.

Kaufmann then outlined eight of the most common misconceptions and concerns relating to the GAC.

In outlining the questions, debates and challenges ahead for the GAC, Kaufmann explained that one of the major questions was whether and how to engage on the 'demand' side of governance, and also maintaining the perspective that other factors apart from poverty matter - for example, at the macro-level, lessons from aid effectiveness/failure could be brought to bear.

Verena Fritz

In her role as discussant, Verena Fritz started by stating that there are still many questions to be answered on issues of governance and anti-corruption, but that the worldwide governance indicators contribute in a major way to the state of knowledge on GAC.

She explained that she wanted to cover four main points in her discussion:

• country categorisation

• resources required for anti-corruption strategies and assessments

• conditionality and transaction costs

• lesson learning

Referring to the 4 groups of countries set out in the GAC (relatively well governed; high opportunity for governance improvements; major obstacles; exceptional risk) she stated that, for the latter two categories, there are many challenges facing such countries. Furthermore, she raised the question about criteria for deciding about how to categorise countries.

Referring to the resources required for implementing and assessing anti-corruption strategies, Fritz stated that 'high opportunity' countries in particular, can be a very tricky category as they can quickly 'fall from grace' - for example, after the Orange Revolution in the Ukraine, there was a 'high opportunity' situation, yet one and a half years on, it still lags far behind expectations. This could also apply to Kenya since the elections in 2002. There are therefore many challenges in reacting to changes within countries with regard to the different categories and to operationalise the WB's concepts.

On conditionality and transaction costs, Fritz stated that there is a commitment at the policy level not to re-increase the use of conditions, but there is also the risk that operational teams further down the line nonetheless will do so; and may even chose conditions such as the creation of AC agencies which have been recognized not to be effective. On transaction costs, there is much ODI work on aid effectiveness, the Paris Declaration, etc. There is a risk that implementing the GAC will increase transaction costs of aid - how will this issue be dealt with?

On lesson learning, Fritz stated that this should be much more hard-headed - for example, the 'demand' side needs to be strengthened - what works and what doesn't? There is a need for broader comparative evaluations - for example, anti-corruption commissions don't work. There are therefore many new challenges which remain.

Finally, Fritz added that the findings from the governance indicators indicated that there had been improvements in freedom and democracy but not necessarily in terms of government effectiveness. She observed that it is surprising that there isn't more of a visible improvement in Africa, given the effort that has gone into capacity building on that continent over the last ten years. She also warned of the dangers of 'feckless pluralism' - there may be a wider choice of parties, but this does not mean that new governments may be more effective. In addition, the first indicator, 'voice and accountability', actually measures civil freedom and there is no good accountability indicator to measure whether governments are actually accountable.

Discussion

Questions and comments raised in the discussion included:

• The WB and other institutions tend to talk only to the elites in countries, not citizens - this is a very narrow conversation. Does this not limit the WB's ability to measure governance?

• Is one indicator more important than the others - i.e. government effectiveness? Have there been problems in bringing the different indicators together? Is it valuable to do so?

• With regard to the issue of DFID withholding £50million from the WB recently - does this indicate a lack of trust with regard to President Wolfowitz?

• In terms of measuring 'accountability' - this term only really exists in the English language and does not translate easily into other languages. Instead it is necessary to describe 3 or 4 different concepts. Would it therefore be more useful to break this concept down into several more easily understandable pieces of analytical work?

• With regard to aid effectiveness, shouldn't we be doing less, but doing it differently?

• There are different 'means' of accountability - for example, the UK and France have very divergent views. The meaning of the term also varies between the public, private and voluntary sectors. In addition, the concept of 'demanders' and 'suppliers' in terms of accountability is also flawed because the voluntary sector could be perceived as both.

• A 2004 WB report into the political and economic state of Uganda found that it was doing well with regard to GAC, but it now appears to be slipping, lubricated by WB unconditional budgetary support. Is this accountability and/or conditionality? How does the WB reconcile this if the funding they deliver appears to be fuelling such a decline?

• Has the WB considered using a classification which is dependent on the size of a territory? For example, applying the indicators at a provincial level might be more useful for large countries.

In his response, Daniel Kaufmann stated:

With regard to the categorisation of countries, Kaufmann agreed that most developing countries fall into the middle of the four categories and have major challenges and constraints which the WB programme will need to be tailored to. There is a need to engage differently and more smartly with countries in selected, focused areas. With regard to the Ukraine and Kenya: political analysis has not been one of the strengths of the WB - this is a problem which needs to be addressed. Country leaders also need support to implement their actions. The process should also be de-centralised and there should be a framework of mutual accountability rather than conditionality.

On the indicators, these are aggregates from existing indicators, and rely on a set of responses from a broad cross-section of people. These responses do show some progress. For example, Singapore does very well on 4-5 of the 6 indicators, but it is an outlier with regard to voice and accountability. With regard to the aggregation of the indicators, the WB prefers not to do this since there would be no clear methodological base for doing so, but some other researchers have taken the data and aggregated it in various ways.

With regard to issue of trust in President Wolfowitz, the DFID WB funding issue is already close to being resolved. The argument on this issue must be professionalised so there is clarity on the way forward. It is important to note that the DFID White Paper on governance does not differ substantially from the WB's approach.

With regard to the WB report on Uganda, Mr Kaufmann stated that an approach is needed which furthers both development and poverty alleviation. There have been accusations of arbitrariness levelled at the WB and there is a need for a paper on how to conceive these ideas and also on the power of the data. It is always necessary to ask tough questions but preferably without disengaging from countries.

With regard to territory size, working at the provincial level in large countries does makes sense and the WB is already doing this in Mexico - and there has been a request for the same from India.

There would be definite advantages of GAC work being independent from the WB, but if this was the case, no internal influence could be brought to bear. All the data is on the website - there is no WB influence over this, so in that sense, the data itself is independent. This is a core competence of the WB. Though the WB is already engaged with many colleagues and experts in these sectors, but there is a challenge of getting relevant anti-corruption and governance expertise into the WB.

In her response, Verena Fritz stated that the WB may not be particularly well equipped to become a major player on governance and anti-corruption issues; since its fundamental expertise is around economic issues. However, given its importance as a development institution and 'agenda setter' it is welcome that it seeks to become more engaged; and now we have to see how this can best be done.

In conclusion, the chair, Julius Court, summed up by stating that this area is very much still a work in progress for both the Bank and for DFID. He asked for people to stay engaged with the issues and to hold Daniel Kaufmann accountable for what he has said about GAC at the WB. He added that it was important to clarify that DFID withheld £50million of funding from the WB was in fact due to concerns over conditionality and not the GAC strategy.