Towards better results from the World Bank

Paying attention to incentives

‘Prevailing incentives,’ says the document, ‘encourage Task Team Leaders to underreport risk and managers to manage it implicitly rather than openly’ (World Bank, 2011: 6). This is a surprisingly candid admission by the authors of the document. Translating from Bank-speak to plain English, they are saying that Bank staff do not like to admit that their projects might fail.

The significance of this admission is not just in its frankness, refreshing though this is in a world where most corporate publications are tools of public relations, but in the importance it attaches to incentives. Incentives matter a great deal in public sector reform. By this I am not referring to performance pay or other crude reward schemes, but to systemic constraints that shape the choices made by actors in reform.

For instance, one of the key choices in public sector reform is how wide or how narrow to make the reform agenda. Any student of public policy implementation will say that the more limited the agenda, the more achievable the reform programme. Overambitious reform goals can overtax both the implementation capacity of governing institutions and the political capital of those in charge of them. All too often, however, reformers over-promise, and reforms fail.

The point is not just that reformers keep making the same mistakes. The point is that there are built-in reasons why they do so. Among other things, reformers may feel compelled to over-promise in an effort to attract donor funds. To say ‘We are embarking on an experiment and we are not really sure what we can achieve’ – the unvarnished truth in most cases – is not a very good way to get funds. Project proposals are inevitably marketing tools, not just management tools.

I came to understand the importance of such basic influences on reformers’ behaviour as a result of my academic research more than a decade ago (Polidano, 2001a). My subsequent experience as a practitioner has reinforced my views. Unintentionally, large organizations can set up perverse incentives that come to exert a more profound influence on the behaviour of staff at different levels – including top managers – than the formal mechanisms which are intended to reward staff and to hold them accountable.

The World Bank’s team leaders appear to be in just this situation. The document suggests addressing the perverse incentive problem by ‘defining risk tolerances’ in the hope that this will allow team leaders ‘to be more honest about the uncertainty in a particular project’ (World Bank, 2011: 17). In other words the Bank is proposing to encourage its staff to be more open about the possibility of failure by telling them that it is OK to fail, up to a point.

It is not at all clear that such a measure will be enough. Some perverse incentives can be hard to correct because of their fundamental nature. This one may be a case in point. Ultimately, the Bank wants to reduce the percentage of ineffective projects in its public sector reform portfolio. So a team leader who admits that project x under his or her direction is at risk of failure knows that, as a result, the project might be axed and the money diverted to other initiatives. Admitting to risk may mean losing projects.

The niggling worry for team leaders who lose projects will be that instead of gaining kudos for honesty, they will come to be seen as someone who lacks resourcefulness, who does not get results, who is too quick to take problems to head office instead of solving them in the field. Being too candid about project risks may have career repercussions for staff. But this is less a matter of organizational practices than one of fundamental human nature. The worry for team leaders is likely to remain, however much the Bank may try to reassure them about declaring risks. And the Bank can only go so far in encouraging its staff to take account of risk. At the end of the day the Bank has to disburse funds, and there is no such thing as a risk-free public sector reform initiative.

As this shows, it is one thing to identify the underlying systemic incentives or constraints that shape people’s behaviour. It is another thing altogether to change those incentives. But the Bank document is surely right to explore the issue, because the incentives will certainly not be changed unless they are first understood. Changing the incentives may be one of the keys to making public sector reform less of a risky business.

‘Best fit’ versus ‘best practice’

Another way to make reform less risky is to avoid the tendency to come up with an a priori set of solutions before one has even analysed the problem. During my time as an academic, everyone seemed to be discussing whether or not the new public management was a good idea, especially for developing countries. The debate was so heavily dominated by participants’ preconceptions that it verged on the surreal. Not many people stopped to ask: To what extent is the new public management actually being taken up in developing countries? What has the outcome been? Moreover, the debate became so heavily preoccupied with the content of reform that strategic and tactical issues – how to go about reform in such a way as to limit the risks and improve the prospects of success – fell by the wayside.

The new public management debate obscured the essential truth that reforms have to be based on the conditions and circumstances particular to a given country at a given time. Evolutionary change, taking those conditions and circumstances as its jumping-off point, is more likely to succeed than context-less prescriptions that are imported lock, stock and barrel from abroad – whatever those prescriptions might be.

It may be tempting to draw a distinction between locally developed, ‘organic’ reform initiatives, and ‘alien’ initiatives imposed by aid donor agencies. This would be crude and unfair. Local actors are just as capable as anyone of coming up with proposals that fail to take account of the local context, and donors can – it should be said – play a valuable gatekeeping role against locally generated but poorly designed proposals. And yet aid donors do need to beware misusing the considerable leverage they enjoy in many developing countries. I recall seeing an evaluation report on a UNDP-sponsored civil service reform programme in a southern African country in which the author complained that reform was being undermined by a serious lack of local ownership. Yet he went on to propose an entirely new component for the project, apparently off the top of his head!

Perhaps donor agencies should not be in the business of telling recipient countries what reforms to adopt. Perhaps donors should take a back seat in public sector reform. Perhaps they should not be writing policy documents on the subject.

From this perspective, the publication of a document grandly titled The World Bank Approach to Public Sector Management 2011–2020 might not seem like a very promising development. However, the document confounds the sceptics by steering firmly away from prescriptiveness. There is a lot of humility packaged within its Bank-speak.

The document draws a distinction between ‘best fit’ (whatever works best in the local context) and ‘best practice’ (one-size-fits-all prescriptions, notably the new public management). The document states that the Bank is now no longer preoccupied solely with the content of reform (what should be done), but also with the context (where it should be done) and the process (how it should be done). ‘Accordingly, there has been a strong move away from “Washington consensus”-style theorizing about PSM [public sector management] reform … towards the idea that “what works” in PSM reform is highly context-contingent’ (World Bank, 2011: 11). Having myself argued in favour of an eclectic approach (Polidano, 2001b: 60–64), this is music to my ears.

The document does not stop there. It adds: ‘“What works” to improve PSM results may be hard to know at the design stage, but may only emerge during project implementation through stakeholder engagement and experimentation’ (World Bank, 2011: 16). This is a revolutionary statement. Traditionally, donors disburse money on the basis of firm project plans which are signed off by both the donors and recipient-country governments. Now here is the World Bank, the donors’ doyen, saying that textbook project management does not apply to public sector reform. The Bank is recognizing that sometimes you do not know what you are looking for until you have found it: ‘ready, fire, aim’ rather than ‘ready, aim, fire’.

The question is whether the ‘ready, fire, aim’ approach is compatible with project-based lending. As far back as the 1990s, proposals were floated within the Bank to ‘de-projectize’ Bank lending for public sector reform (Nunberg, 1996). The document notes that, indeed, Bank lending instruments have evolved to permit greater flexibility in project management, but it says that further progress is needed in this direction.

The document also acknowledges, however, that the shift in thinking has still to fully reflect itself in the Bank’s public sector management work on the ground, where Bank staff themselves have often expressed concerns that ‘best fit’ has yet to fully displace ‘best practice’ thinking (World Bank, 2011: 6–7).

The Bank puts forward a number of reasons why best practice persists. First, governments themselves adopt best practice orthodoxies in the belief that this will add to the credibility of their reform initiatives and help them to gain support – ‘including sometimes from the World Bank’ (World Bank, 2011: 7). Second, there is an entire consultancy industry devoted to packaging and selling best practice orthodoxies. Third, advisers cannot start with a blank slate in every new project: best practice provides a convenient starting-point. Finally, the absence of a well-developed, explicit body of knowledge derived from practical experience about ‘what works’ makes it harder to argue against best practice orthodoxies.

Two points seem particularly worth noting here. First, recognition is again being given to systemic factors which shape the behaviour of the players in public sector reform, on the part of both host-country governments and other actors including donor agency staff. Second, it is surprising that an intellectual powerhouse like the World Bank, with its extensive experience in public sector reform, has not documented that experience through research into ‘what works’ (and under what circumstances).

Speaking with one voice

Can the Bank deliver on the document’s promise to eschew ‘best practice’ orthodoxy in favour of a more eclectic, experimental approach? The document gives rise to some doubts on this count in its discussion of the need for consistency in the solutions the Bank proposes to client countries, particularly where it has more than one team working on interlinked issues in the same country. ‘Clients’, says the document, ‘expect the Bank to provide consistent solutions to their problems and not to speak with different voices about the same problem’ (World Bank, 2011: 14). So the document proposes to establish ‘a common disciplinary background for staff working on PSM issues, enshrined in robust technical competences’ (World Bank, 2011: 19).

Here, perhaps, is another instance of the Bank being subject to a perverse systemic constraint. Client governments want highly-paid foreign experts to come up with answers, and if the experts cannot agree among themselves it will dent their credibility. But the Bank’s response to this fundamental expectation may be at odds with the direction mapped out elsewhere in the document. Will the ‘single voice’ amount to a new orthodoxy? The risk is there, because learning ‘what works’ in a given context is a messy business with plenty of scope for disagreement among the Bank staff who are doing the learning. One can hardly expect them all to spontaneously come to the same conclusions. If staff are to be expected to speak with one voice, then the message has to be put in their mouths beforehand.

This is one instance where the Bank document itself appears to be speaking with more than one voice. Indeed, it is not always clear who the document is speaking for. When it discusses the possibility of developing data to measure the impact of public sector reform, the document says that:

… such data may be part of the response to increasing pressure for measurable results at the Bank’s corporate level. While there are undoubted risks in any attempt to respond to such pressures, in long term PSM projects (where the trajectory of feasible reform will only emerge over time), some response will be a pragmatic necessity. (World Bank, 2011: 17–18)

The authors of the document here assume an identity as a distinct group within the World Bank. In a remarkable statement, they seem to be saying: ‘We don’t agree with our bosses, but we will have to keep them happy.’

But if this document does not – in parts at least – speak for the corporate leadership of the World Bank, it is all the more praiseworthy for that corporate leadership to permit the document to be made public. This document represents a major exercise in self-appraisal on the Bank’s part. In this day and age when so much importance is given to public relations and ‘spin’, the document is most remarkable for its frankness and honesty. It openly discusses issues which, in many other organizations, would be brought up only in quiet whispers over coffee. There are areas where it may be appropriate to sound a note of caution about this document, but in overall terms it is encouraging to see the evolution of the Bank’s views on public sector reform. The document is a positive sign for the future of public sector reform in developing countries, and I dare say that in the years to come it will be seen as a landmark publication.