An initiative of the National Academy of Public Administration, and the Woodrow Wilson School of Public and International Affairs and the Bobst Center for Peace and Justice, Princeton University Oral History Program Series: Civil Service Interview no.: N1 Interviewee: Gregory Ellis Interviewer: David Hausman Date of Interview: 20 July 2009 Location: World Bank Washington, DC U.S.A. Innovations for Successful Societies, Bobst Center for Peace and Justice Princeton University, 83 Prospect Avenue, Princeton, New Jersey, 08544, USA www.princeton.edu/successfulsocieties HAUSMAN: This is David Hausman, and I'm here with Gregory Keith Ellis at the World Bank in Washington, D.C. Mr. Ellis, have you consented to this interview? ELLIS: I have. HAUSMAN: Thanks. So I'd like to begin this conversation by talking about the role that you played in civil service reform here at the World Bank and elsewhere. So that our listeners can get to know a bit about you, could you just describe your current role and the jobs that brought you to this position? ELLIS: OK. Currently I am a senior operations officer with the Fragile and Conflict-Affected Countries Group within the World Bank. I am actually a secondee from the Australian aid agency AusAid; I'm here for a period of three years, and I've been here for about a year now. In this job we provide central advice to the various country programs working in fragile and conflict-affected countries. My background is working in mainly conflict-affected countries in the Pacific for AusAid, the Australian aid agency. I've had two overseas postings during my time, the first in East Timor immediately after the withdrawal of the Indonesians in 2000 through to 2002, and more recently working in the Solomon Islands with the regional systems mission led by Australia-New Zealand, and that was for the period 2005 to 2007. And, as I said, in 2008 I took up this position with the bank. HAUSMAN: Great. Could you describe the sorts of programs that you were most involved in during that time? ELLIS: Probably for the purposes of this sort of work, my experiences in the Solomons is going to be what I'll draw from. I was the deputy program manager for the Machinery of Government Program, which is one of the three pillars of the RAMSI mission [Regional Assistance Mission to Solomon Islands], one being law and justice, one being economic governance, and the third-which is the one I was involved in-the machinery of government, which was looking at all the civil service reform requirements of the Solomon Islands. In that position I worked on a number of areas with my colleagues, looking at systems right across the various elements of the civil service. HAUSMAN: As you know, our program works with governments that face especially difficult challenges. So starting very generally, how do you think that public service management reform differs in fragile states compared to other countries or among different types of fragile states? ELLIS: I think the greatest difficulty for these countries is, firstly, they're very weak. They have very weak capacity, so that you come from a base that is extremely low. The countries I've worked in which have also suffered conflict have not only had poor capacity since decolonization, but they've also grappled with the aftermath of conflict, where there has been sometimes destruction of infrastructure, but certainly problems, psychologically, with the people that are in these countries. So they've had huge amounts of disruption, and their capacity to recover from that is actually quite low. So in many ways, the issues of shortfalls in capacity are compounded in post-conflict situations. HAUSMAN: Which key reforms among the common public administration and civil service reforms do you think require different approaches in states like these? ELLIS: I think the issue which we grapple with most is this issue of how you actually develop capacity, how you also prioritize where you're going to provide your assistance. I mean, the problems of civil service reform in these environments quite often are the same problems you see right across the development spectrum, it doesn't matter what sector you're working in. The civil service reform area is probably more problematic in that it is often highly politicized in these environments, and I think one of the problems we wrestle with is understanding the political economy of these countries, working out what are the motivations and incentives behind behavior that we're trying to assist-I don't want to say direct-but to try and improve, wrestling with the improvements that you want for building capacity in these environments. You have a lot of difficulty with those political issues-as somebody has called it, the ghost in the machine. If you don't understand the shadow state that actually flows through these environments, your ability to actually assist can be very, very minimal. In fact, you can actually do harm by not really understanding that. I think one of the problems we've got, particularly in these countries where I've been working: it is not just those political economy issues, but then these countries are also wrestling with traditional systems where you've had a colonial administrative system grafted over the top of a different system, and a clash between the traditional and customary and the more Westphalian state creates huge amounts of conflict. I think donors struggle with providing assistance that really takes account of those sorts of complexities. There's always this tendency to replicate what we know, and despite the rhetoric around "context is everything," I don't know that we really handle well reviewing the way we're going to approach things, particularly so that we truly take account of the local context. HAUSMAN: I wonder if you could give some examples of that problem from your own experience. ELLIS: Probably again going to the Solomon Islands: it's a Melanesian culture on the whole. It has a Westminster-style government grafted over top of it. You have situations such as wontok [Melanesian pidgin for "one talk": being of one language group], this notion that a person's background, their language group is their primary focus of who they are. If you ask somebody from the Solomon Islands who they are, they won't say they're from the Solomons, they will say they belong to a particular tribal group, a particular clan. So therefore the way that society works in many ways is that the allegiances that you have are to your clan, not necessarily to the wider social networks. The system worked very well in the days before globalization and monetization of the economy, and people relied very much on their social networks for support. What happens when you graft over the top of that sort of complex network of social obligation a western system of government: there is a clash then between your obligations to your tribal links with the obligations of the wider state. We see that evidenced in forms of corruption or nepotism, not necessarily always seen that way at the local level. I think our inability to navigate through the complexities of those relationships causes a lot of difficulty, particularly in civil service reform. HAUSMAN: Were there particular times that that came up in your work? ELLIS: I suppose not for me personally. As I was managing a range of programs, I wasn't involved intimately in the day-to-day dealings with the actual issues. But certainly some of the projects I was overseeing, the advisors, the consultants, the different people who were working on those were coming up against these problems. I think their capacity to navigate that process was difficult. We had a small activity that we were supporting in the Honiara City Council, which is the capital of the Solomons. We had a financial advisor there trying to ensure that decisions about remuneration for councilors on the Honiara City Council were appropriate. What they saw as appropriate would be considered probably in my country to be inappropriate. But our ability to manage to steer away from those things can be very difficult. Again it is that balance, too, of how much you're trying to manage the process and tell people what to do as distinct from taking a back seat and advising on those issues. That very difficulty of trying to recognize the sovereignty of the people that you're dealing with against trying to avoid making very poor management decisions which are going to threaten the viability of the institution you're trying to help. HAUSMAN: Going back to this question about how the context of a fragile state affects civil service reforms, is there anything you would say about what initial reforms are often successful in fragile states? ELLIS: One of the lessons I've come up with through my experience is that it is very dangerous to try and assume which reform should happen first. We all know what are the core functions of government, and there's a tendency always to assume that these same views of function are also held by the people we're dealing with. But when you're talking about-and this is broader than civil service reform, but building or rebuilding a state, trying to work out what should be done first is a very difficult area, and I think one of the problems we do tend to come up against is that we think that particular reforms are important, but that may not necessarily be the priorities of the people we're dealing with. I have to say that if it is not seen as a priority for the national stakeholders, no amount of effort on our part is going to make that successful. So one of my first things is to try and work out your entry points based on the sorts of things that are considered priority by the local people. Provided they are still consistent with good governance, I think we should be getting behind their priorities and not saying, well, in this situation you may think priority one is important, but we actually think that priority two is, and therefore that's the one we wish to support. Those things will inevitably fail. I mean, for me, a very good example of that approach in my area was that one of the most successful activities I was involved in was looking at government housing, that is the housing of civil servants across the country so that they could provide services. On the face of it, in the early stages of recovery, worrying about the housing in a country may not be considered the first priority. You know, health, education, security; there were a lot of other issues that would appear to be of higher priority. Not to say that we were not trying to address those simultaneously, but this particular priority wasn't one that would have immediately leapt out to us as being important. However, we had a request from the government to look at this particular issue, and it was quite clear that there was a strong agreement across a wide range of senior civil servants and ministers that this particular issue was critical and was causing a lot of problems for them. So we put an advisor into that area. Through her work and a subsequent project, we had a huge amount of traction. Whilst the actual issue itself didn't sound like a high priority at that particular moment, as we got further into it, I found that it actually pulled out a lot of other issues that on the surface appeared irrelevant. Issues around service delivery: why was it that the Ministry of Health couldn't deliver health services where they wanted them? The fact was that the location and condition of houses in that country actually distorted decision-making around where services were being delivered. So in actual fact, housing became a fundamental concern in relation to something like basic service delivery, which was one of the issues that you would normally expect to focus on. Similarly, when analysis was done of how housing allowances were paid, or rental subsidies, or anything of that nature, it became clear that those allowances and subsidies were having a distortional impact upon how the civil service operated. It may not have been the only problem, but by analyzing the housing issue we exposed a lot of problems that were emblematic of problems right across the civil service. So that entry point became very, very useful in forming broader civil service reform. Again, we would never have expected that had we not gone down the path which was requested by the Solomon Islands' government. They hadn't articulated that depth of complexity, but it became quite apparent that that was a real issue. Of course, then we found, because it was of such a concern to them, that we had good traction and that the recommendations of the analysis, the work with the steering committee that they had formed, had an immense amount of traction and actually provided support right across the civil service-it would not have happened had we not gone down that path, if we'd decided that it wasn't considered a priority. HAUSMAN: Great, I wonder if you could go into just a little more detail on how the reform of housing took place, and then its relationship to civil service reform. ELLIS: I think one of the things about that particular area that I would recommend in the future if I were to work in another civil service project like that, was that when we were first asked to provide assistance in that area we knew very little about the issue, and it was very difficult to get much of an understanding from the people on the ground. The government was still grappling with the aftermath of the conflict, so it was very difficult to get people to tell us much about what was going on. In those circumstances I determined that rather than trying to go down the path of a normal project-based approach, what I would do first off was identify a very good single advisor to go in for six to nine months with quite loose terms of reference. Her terms of reference were really to provide some immediate policy advice to the government-and this is something they had already asked us for. So we were providing them with something they wanted. But she had a second task, which was not as explicit as the first one, and that was to really understand the environment in which this would work, to work with the ministry, the staff of that ministry, try to work out what the landscape of the reforms was. We didn't try and design up front an activity. What we tried to do was get a much deeper understanding of that environment. What I did then was, as her work progressed and as she delivered the first of the policy papers that she was required to provide to government, we then brought in a very small design team to build upon her experience, work with the stakeholders that she had built a relationship with, to develop a second phase which had a much higher profile within the Ministry of Lands where this project was located. That could not have happened without that period of analysis and relationship-building. I often say that the design period for that project was actually nine to twelve months, not two or three months, as was normally the case for the agency I worked with at the time. So we were actually providing support to the government while we were designing that particular activity, and that became a very successful model and something I would probably do again if I was faced with the same situation. HAUSMAN: Could you just go on about how the housing reform went forward? ELLIS: I suppose the housing project-the policy analysis that was done really highlighted a whole series of flaws in the structure, not only the fact that the government didn't know what houses it owned and what their condition was and where they were in the country; there had been a series over many, many years, incremental decisions about how to house and provide allowances to civil servants for their housing, and there had been no assessment done of the impact of those decisions over time. This was to the point that our analysis showed that the amount of money being spent on subsidies and rental allowances for civil servants was creeping up to about 6-8% of the entire government budget, which was the equivalent of, say, the total budget for the Ministry of Agriculture at the time-which meant it was a huge drain, and it wasn't clear that that this was sustainable, but also it wasn't clear that it was actually achieving-it wasn't achieving the service delivery aims that should have occurred from such a major investment. The other thing that we found very useful in the process was, as the reform was going ahead we had to create out of nothing a housing unit to manage the forward program. The way that was approached-again because of the very, very good relationship that the advisor had developed with the permanent secretary of that ministry and some of the other stakeholders on the committee, the steering committee for this which represented at the highest level most of the major service delivery agencies-was that we were able to identify staff that were not currently being used effectively within the agency. So without creating new positions we were able to create a housing unit out of existing staff, which is fairly rare in my experience. The first thing that tends to happen when you identify new functions is that you decide that the staff of the department needs to be larger. What we did in this case was, we actually identified where staff had the skill set required, but were not being properly utilized elsewhere in the ministry, and also we brought a member of the staff across from the public service department because there was a shift of function from one department to the other as a result of the approval of this policy by cabinet. So what happened was that there was no net increase in the size of the civil service, but we created a unit that provided a much needed service. I think that in itself is quite a lesson. The other thing was, because we had a light touch in terms of the technical assistance, we were very reliant on identifying capacity within the ministry rather than seeking to bring in capacity from outside. For me it was a salutary lesson that there often was an assumption that capacity doesn't exist. But because we had an advisor in place for so long within the ministry, she got to know a lot of the staff in the ministry. She was able to assess their capacities, and she was able to recommend to the secretary of the department where he might be able to pull staff together. That turned out to be a very useful approach, I think. HAUSMAN: Great. What did the housing unit do next? ELLIS: The housing unit first off established a register for all the government houses in the Solomon Islands. It did maintenance reports on all of them. It started to regularize the rental schemes and the subsidy arrangements, had lots of links with the Ministry of Finance to try and start assessing what the true cost of the subsidies were. I think also, when I left it was still early days, so the hard decisions about getting rid of subsidies and phasing out particular ways of working had not occurred, but there certainly were a lot of moves in that direction when I left. I think the other thing that was really important was, at the same time that that particular project was happening, we were working in a number of other areas, including the Office of the Auditor General, to recreate, to build that office back up. One of the things that again was very useful was the ability to work with different parts of the system simultaneously to provide some sort of mutual support. So what we did, because we were also working providing support to the Office of the Auditor General, we were able to support them to do an audit of the housing allowance systems. On one side the housing project helped the Office of the Auditor General to do their audit; then the auditor general was able to bring to the attention of the parliament and the executive the real problems with the system of housing management as it was. Because we had a project already operating in the housing area, they were then able to take those audit recommendations and support their implementation quickly. So there were lots of synergies between the projects. Again, that was, I think, a very useful marriage of a number of activities. It was being aware that there were many links in the chain. Where we had strategically placed advisory support, we were able to leverage off that to get a good outcome, I think, for the government. HAUSMAN: Great. And were there other ways in which this led to later civil service reforms there? ELLIS: I suppose to the extent that it identified the distortional impact of remuneration on public service behavior; I think people, agencies, and individual public servants, decisions about what they would do, where they would work, were being affected more by housing than need in terms of service delivery. I think that was a very important thing. I think too, the other thing that we've learned from some of the work we did was how to approach the building of capacity. The example in the housing project was that we were able to utilize existing capacity; identifying where good skills existed and putting those people into a position where they were able to use their skills was one approach. The other approach we had which I thought was also very useful was the one that was employed in the Office of the Auditor General, where you had an office that they completely gutted during the time of the tensions. This is the period of civil conflict, where I think something like 37 to 40 staff were in the Office of the Auditor General prior to 1998. By the time RAMSI came into the country in 2003, after the tensions, the office had two staff. So systematically, over many years, the office had been gutted of its staff. So you were left with a challenge that the Auditor General was a very important office but it had absolutely no capacity whatsoever, so it was very different from housing issues that we were faced with. In that situation, quite often what has been done by donors is to put a lot of advisors in to provide that support; you still are not creating indigenous capacity through that approach. What we did in that project was, right from the start we consciously agreed that we had to put in some external technical assistance to start with, but it was shaped around a program of graduate recruitment and training. So we just didn't put advisors in, auditors in, international auditors in to provide the support. What we did was, we did put those advisors in but concurrently, in parallel with that activity. We established a graduate training program. So we recruited graduates, or I should say, we helped the Auditor General's Office to recruit graduates, and we provided them with training, some of it on the job, some overseas. There was some short-term input from institutions elsewhere in the Pacific, and we built their training up to the point where when I left the Solomons, some of those graduates were taking over the middle management positions within the organization. As that was occurring, the international advisors were being withdrawn, so you were seeing a deliberative and progressive handover of that institution back to the national managers. I think the other thing that was interesting about that project which we often faced was the recruitment; the ability to recruit people is quite difficult in those environments. So what we agreed to do in that situation was, the RAMSI mission actually provided the funding to allow those graduates to be paid salaries during the period because the recruitment processes were slow and difficult in that environment. Then, as budget was made available out of the national budget, they moved across to the government payroll. The critical thing about that was that we made a decision right from the start that the terms and conditions under which they were employed with donor money mimicked the civil service conditions within the Solomon Islands. So when the time came for them to transfer across, there was no change in their conditions. Quite often, to attract recruits, there is a tendency to offer conditions and salaries higher than can be sustained by the government. Then you're left with the problem of how you exit that process. Because we started from a basis that we would mimic local terms and conditions, that problem was avoided. HAUSMAN: Now I'd like to kind of talk through a list of various kinds of civil service reform interventions, and ask you about some of your experiences in each, and some success stories that you've witnesses or been involved in in each instance, since we're talking about it now. Are there any other capacity-building efforts that you think were particularly successful that you were involved in? I know you just said a lot about- ELLIS: Yes, I think the way we approached capacity development is one of the main challenges we face. I do think that there is a tendency to think of technical assistance [TA] as the default option when faced with problems of very low capacity. I think that sometimes that's not the way to go. I think that really we should be very careful and cautious particularly about the provision of international TA into the system. We should be looking at a variety of ways of going about these things. I'm not sure that we have good enough tools yet to provide the capacity development that countries actually do need. There is obviously a role sometimes for capacity-for international TA coming into the system; however, I don't know that it should be the be-all and end-all of the approach. Certainly, as I said, the idea of ensuring that there is some way of building indigenous capacity to develop capacity is absolutely critical. One of the things that I know was being discussed was in an area that I wasn't directly involved with-was the idea of helping professional associations within the cadre of civil service servants so that they could actually take up that role. There are associations of professional accountants, for instance. So if you need public-sector accountants, rather than just running a whole lot of training courses for them, actually investing in the development of professional associations so that they take up that role can be useful. This is similar in the legal and justice sector. You're working with bar associations, with senior civil servants associations, and those sorts of things. There is a tendency to ignore that and to work directly with people. But what happens is, when you pull out, when you finish, you think you've done your job; you haven't left behind any sort of capacity that can take on that role further. So I think there is a lot of worth in looking at those sorts of things: creating, supporting twinning arrangements, the fact that if you wanted to create an association, say, for public sector accountants, the fact that there are similar bodies both in the region and internationally that you could link them with, so that some ongoing relationship is developed which they can draw upon if they need to further down the track-I just think there has to be a much more sophisticated approach to capacity development than just sitting there and plugging TA into the gaps. I just feel that ultimately that will not achieve what you're looking for in terms of sustainability. The problems that you do have in these situations, of course, is that there is always this tension about the need to get something happening quickly on the ground from this longer-term requirement to build the capacity. You've got to manage that, inevitably. But I think too often we do one and not the other, and I think these things can be done simultaneously. Some of the examples, such as the Office of the Auditor General that I described, was a case in point where we did provide TA to get things done quickly, but we also didn't lose sight of the fact that at the end of the day we had to develop a cadre of auditors that would go on after our project investments had ended. HAUSMAN: Let me ask you now a bit about professionalization and merit in civil service. Often an important goal in civil service reform is to increase the professionalism of the civil service by introducing a stricter meritocracy in recruitment and promotions. Can you describe some of the more successful cases you've seen of merit recruitment reform or other types of merit initiatives in civil service? ELLIS: I think this one is a very challenging one. This also goes back to this issue about the way in which societies operate and the political economy issues. I tend to think, first, that we often assume that everything should operate on a merit-based system, and I think we should be honest ourselves in the fact that our own societies are not purely merit-based. There are many situations where decisions are made based on personal networks and so forth. I don't think any country truly operates on a purely merit-based system. There are obviously degrees in this, and there are situations where the corruption that comes with our clientalism and neopatrimonialism really do hinder good governance and effective development. But putting that to one side, I think this is an area we found really very challenging. To be absolutely honest, I don't know how effectively we've tackled this in all situations. There have been some gains in some of the work we were doing, where we had institutionalized some levels of more regularized recruitment processes within the civil services that I was working with. But it would be dishonest of me to say that we were completely successful. One area that we found quite useful, particularly in the early days, was to try and remove the difficulty some of our national managers had in being the person that actually makes the decision. It's interesting; when I was in East Timor-this is going back a long way, but when I was there they were doing major recruitment in the civil service. This was during the UN mission, so at that time you didn't have a sovereign government, but we were transitioning to make sure that national managers were making the decisions about recruitment in the civil service. But at this particular point we were approached by a national manager who said that he had a preference for an international to actually be the delegate to make the decision about who should and shouldn't be recruited. As he put it, if he had made the decision somebody would have come around and burned down his house. If an international person made that decision, it allowed them to-strangely enough, people would accept the decision as being much more impartial, because the international did not have the social obligations and social networks that the local had. I think in some positions we've got to look at this issue about moving to merit-based recruitment as a very long-term transition. In societies where there are very strong social networks, merit-based recruitment is quite often hard for local people to comprehend. The basis of survival is based on your social networks, and people do not see, necessarily, that a purely merit-based selection process is the appropriate way to go. What I have seen in a number of areas-and this was both in the Solomons and East Timor-a decision was very sensitive, and it was very important that there was an independent decision about the selection process. Quite often, at least as a transitional mechanism, external people were brought in to assist in that process. It might mean they worked side by side, and they didn't make the decision purely themselves but provided some level of protection for the locals. So say the decision is not made by one person, one national, but that it is made by a panel, so that two or three people, including people that are external to that culture, are seen to make the decision. That seems to have some ability to minimize the tensions that arise, the clash between a merit-based system and a system that is based on other social norms. It is not ideal, but if it is seen as a long-term transition process, and as people become more familiar with the idea that merit-based promotion will provide a more effective workforce over time, it could be useful. I have to say, in all the decisions that I've been involved with, that is a very long, slow process. I think people should be very cautious about leaping to introduce pure merit-based systems on the assumption that that is the way they should go. HAUSMAN: Are there any other ways you've seen of balancing patronage and merit-based demands? ELLIS: I'd like to say I have. It is a constant source of tension in these environments. My biggest problem has been trying to come up with ways of balancing the two issues. I don't think I've got a good answer to that, beyond absolutely trying your hardest to understand the political economy of the country, as I mentioned before, trying to understand how the system does actually work, not how it looks to work. I think that, when we look at some of these civil service institutions in these fragile contexts, we often see a faŤade that looks western but behind the scenes it is not that way at all. Sometimes I think we've got to make some fairly realistic decisions about where and when we will try and push for reform or encourage reform and when we should withdraw from particular areas. Sometimes it just doesn't work. One of the hardest things I find in this area is trying to encourage the evolution of an institution, from its charismatic leadership. Many of these institutions look like formal institutions, but their success or otherwise depends on their leadership. If that leadership changes-which so often happens, as these environments are very dynamic, so you do get lots of change-you end up with a situation: what looks like a good ministry at the start because it has some change champion-you find that a change of government or some other intervening factor means that person moves away. You go from an institution that seems to be worth investing in to one that is actually quite toxic. I think that's a very difficult challenge, and one you've got to consistently be reassessing. I think sometimes we do political economy analysis as a one-off activity; that's not good enough. We need to have ways to constantly reassess and reevaluate what we're trying to do and test whether our thinking about what support would be valuable is in fact going to be the appropriate way to go. I don't think we've got a very good record on that sometimes. I tend to think we think we understand the system-it seems to make logical sense in saying we must work with this particular agency in this fashion; we tend to ignore the political economy advice that comes our way that says in actual fact, this is not working, and we tend to try and just keep working the way we are, and in some ways I think that is actually doing harm. We may be better in certain circumstances withdrawing. As I said, in the Honiara City Council, for instance, where with the change of administration half-way through the process that we were involved with, it became quite corrupt-or the advice that we were getting from the ground was that it was quite corrupt. We actually made decisions to withdraw our support. We weren't able to affect the decision-making, and we just thought it was better to withdraw until such time as we saw an improved environment in which to provide the support, continuing the way we were going was not going to end in positive results. In actual fact it may have put our advisor at some risk, as he was trying to push against some of the decisions that appeared to be not in the best interests of the organization. HAUSMAN: You mentioned before that when you were working in the Solomon Islands, you helped put in place some regularized or semi-regularized recruitment procedures. Were those the ones that you mentioned before with the panel choosing, or were there other ones, and how did that work? ELLIS: This was very early, the very early stages of development. What we were trying to do at the time: we were looking at discrete elements of the system that needed help, recruitment obviously being one of them, because it was a gutted civil service, and there needed to be increased levels of recruitment, or at least better identification of people within the system, to make sure that the benefits of the skills that were currently in the civil service were being maximized. One of the things we were trying to do, slowly, was to decentralize the decision-making process. One of the problems in many of these countries is a highly centralized system, an unwillingness or inability to delegate decision making. What we were trying to do is put that decision-making at least out to the highest levels of the line departments, but ideally lower down the management tree. I'm not sure whether this is the case, but it may also help prevent so much patronage in the system to have a more decentralized recruitment process. At least you are spreading the decision-making about who is recruited through a wider group of people. I'm not saying that that is necessarily the case. You would have to do some analysis. But you might be more efficient in your patronage in a centralized system than in a decentralized system. Alternatively, the challenge then is how to build and maintain the capacity across the wider management structures to continue to effectively do the recruitment process in a way that is transparent and accountable, so there are challenges both ways. HAUSMAN: Right, so you led directly into the next area, which is organizational structure. Do you think that reforms in organizational structure should be approached differently in fragile states versus in more developed countries, and are there particular structures that you think are more successful? ELLIS: I certainly do think they have to be approached differently. It is a bit of a conundrum in a way that fragile countries in some ways have to be more effective at public administration than developed countries. They don't have the resources to make the mistakes, if you like. So there is a real conflict there. They don't have resources, and yet quite often they make very poor judgments; their civil services are very poorly managed. This leads to something that I think these countries have to wrestle with: what are the core functions that a state must perform. They don't have the luxury of resources to do everything that a citizen might wish them to provide. I sometimes think that donors, particularly in post-conflict settings, are culpable in this. We tend to try to provide support across a huge range of government services, and in actual fact there isn't the capacity or the resources to do any of those properly. I think there is a challenge across there. I think one of the maxims should be: better do a few things well than many things poorly. Helping countries work through the prioritization that that requires is a very, very big challenge. In many of these states, fragile states, particularly the ones merging out of conflict, there is an expectation, a desire for many, many improvements, and they can't all be delivered. So that is a difficult one for everyone, but I think those priorities have to be worked through. Clearly the first one is security. If you haven't got security, you can't have development. So that has to be a fundamental requirement. There are the very, very basic services that people need: basic education, basic health, the ability to pursue a livelihood. So creating an environment whereby people can pursue their lives in a way that provides what they need to survive, the ability to raise revenues and manage those effectively, these are all the very key things. I think one of the interesting issues for me, which I've been looking at a bit while I've been in the bank-which I have to say isn't an area where I've had direct operational experience, but it's one I'm interested in exploring-is trying to get governments in some of these settings to think about how basic services are actually delivered. In my dealings in the Solomons and in other countries in the Pacific in particular-and I'm sure this is probably reflected in many other fragile states-governments feel they must directly deliver services, and sometimes that is not feasible. Again, you have a tension that if others deliver services and not the government, you weaken the bond between citizen and state, and what you are trying to do, of course, is create a legitimate state where the citizen sees the state as being useful to them and therefore having a stake in stability. But I wonder sometimes whether a state does necessarily have to deliver the services directly to create that compact. I think there are other ways in which that can be done, and the one I think would be worth exploring is more than-it's happening in some areas now, particularly in the health sector in certain areas. But it is looking at the state as an enabler and facilitator of others delivering those services, be they NGOs [non-governmental organizations], be they churches, private sector. I think sometimes we discount the effectiveness of doing it that way. The important thing is that it is not a donor working with an NGO to deliver a service. What it should be is that a donor supports the government, and the government should be the one that is actually enabling that service to be provided through a third party. I think that is a very important distinction. Quite often in post-conflict environments, donors go in, there is an urgent need for health, there is an urgent need to restore basic education; and in the desire to provide a peace dividend quickly to the community, the donor or the NGO steps in to do that work. Whilst that may be necessary in the short term, the danger is that you weaken the bond between citizen and state. So I think at least in parallel we should be working to put the state, put the government back in the driving seat in that area. A good example that I've been exploring here is some work that is being done on output-based aid, which has been used in some African states in the health sector and in water and sanitation in parts of Southeast Asia, where the government contracts private sector or NGOs to deliver services, and they're paid by output. It seems to work quite well in the health sector, from what I've seen in the research so far, and I think it is worth exploring a bit further. Similarly, some of the dialogue I was having in the Solomon Islands with some groups was looking at how people wanted their services provided. When you go out to many communities in remote parts of the Solomons, communities often prefer their services provided, say, by the church-it is a very religious community. The church quite often provides education and health services, and I've often wondered whether it would be better to try and help governments work with those institutions to provide services rather than trying to compete with them to deliver services. Similarly, looking at agricultural extension in communities, many Western countries, many developed countries have moved away from direct provision of extension services and used other intermediaries, and I think there is a lot of value in that. Again in the Solomons, I was noting that there were some very good agricultural NGOs that were working with farmers very effectively. I've often wondered whether, rather than trying to provide agricultural extension offices in remote areas-which is costly, and then they're often faced with demands from farmers that they cannot possibly handle technically-that it may be better for governments to go into partnership with those technically specialized NGOs to provide the services rather than doing it directly. I just think that is an area worthy of pursuing. So when you're looking for a civil service skill, instead of looking at civil servants with specific technical skills, what you're looking for are civil servants who have the skills to negotiate and to contract and to oversee and monitor the delivery of services by others. That's quite a different focus for civil service, but something I think is worth exploring. Not in all areas-there are some areas where the state must be the direct provider. But I don't think that we should assume that all services need to be provided directly by government. HAUSMAN: Great. So moving on to some questions about implementation of reforms. What do you think a reasonable time span is that reform leadership should consider in thinking about how long to implement various changes? ELLIS: In many of the countries I'm working in, I think, we have to look at generational change. I think it has taken developed countries a long time to get to where they are with their systems, and they're by no means perfect. Yet we do expect some of these fragile governments to implement radical changes to the way they govern in periods of five or 10 years, and I just don't think that is feasible. So I do think we've got to look at much longer timeframes. That is a big challenge for donors to stay engaged, to provide long-term and predictable funding, and also to be willing to take some risks. There will be steps forward, there will also be steps backwards in the process. Staying engaged over long periods of time is very difficult, but I think that is the way that we need to look at these particular issues. Civil service reform is a very difficult area to operate in, and I don't think we've got the answers yet, but I do know that if we're going to have any chance of succeeding we've got to work in partnership for very long periods of time. HAUSMAN: Let me ask you a hypothetical question. If you had the chance to design a unit to move reforms forward in a fragile state, what kinds of skills and staff would you want to put in place for the unit? Where would you put it in the organizational structure of the government, and what kind of relationship among ministries or officials would you recommend? ELLIS: First off, I wouldn't want to be the one making the decision about where such a reform unit should be placed. I think it is important to have some sort of leadership in any sort of reform process, definitely, but I think there is a danger in concentrating all the reform being driven out of one particular unit. The unit should be actually-it's all very contextualized. It depends on which country you're dealing with, where the change champions are, where the leadership is to put these things into place. However, if you're looking at, say, broad civil service reform, what you really want to do is establish some sort of coordinating structure that encourages ownership across a broad spectrum of the public service. There is a danger in any of these systems of putting, say, a mechanism in the Prime Minister's office or in the Ministry of Finance or the Ministry of Planning, whatever they are. Depending on their relationship with the institutions that you're seeking to reform, you may or may not get lots of resistance. So I think in the process of developing, putting in place any formal unit, you've really got to think about those political dynamics and position it-or at least have management structures around those units that maximize the potential buy-in of the institutions that you're trying to reform. I think the other danger is to try and keep the international advisor footprint fairly low, as low as is possible given what you're trying to achieve. I have a real problem with project implementation units, which tend to create parallel structures and actually can hollow out institutions which they're purportedly trying to assist. They end up being better resourced than the institution in which they reside. That causes a lot of resentment and definitely undermines ownership. So we've got to be very cautious about how much the reform process is concentrated in one unit rather than making sure that the unit is one that actually coordinates and supports the reform agendas that happen broadly across government. So a light touch with any sort of coordinating unit is my recommendation, but again I'd have to qualify it: it would depend on the particular context in which you're trying to do it. HAUSMAN: Often reform leaders look to models that have worked in other places, like in Singapore or in New Zealand. These examples might work in some places and not in others. In what context do you think, for example, the Singapore model makes sense, or the New Zealand model? Are there other examples that you think reform leaders should look at? ELLIS: I get very nervous about building on reforms in countries such as that, such as Singapore or New Zealand or Australia, in that contextually quite often they're very, very different environments. They can actually raise expectations of a system that is not sustainable in the environment that you're trying to work within. However, it is very difficult to talk reform in the abstract. People do actually need to see how something might work. You do need a goal to work to, but how you manage that is a very difficult thing. I've just recently come back from Liberia, for instance. I was interviewing the Liberian National Police, and there's a lot of unhappiness there-for good reason-about the poor state of their resources, but it is actually being fueled, I think, by the fact that many of the senior officers at various levels had been sent to Britain to have a look at policing in that country. Of course, instead of just thinking about the principles of policing and some of the high-level policing issues that are going to be valuable to them, it sort of backfired and created resentment, because they see the resource richness of a country like Britain, and it can actually be demotivating. They feel that they cannot do what they have to do because they haven't got the resources that they see in countries such as that. So I think we've got to be very careful about these models, about how we use them. Sometimes I think it's a case of trying to find models where the nature of the country parallels more closely the one you're looking at. For instance, again in the Solomon Islands, if you were looking at a budgetary system, I've often thought that rather than looking at an Australian federated system, if you were to take somebody from the Ministry of Finance in the Solomon Islands and show them the Ministry of Finance in Australia, you'd be looking at the wrong level of government in many ways. The Australian system is a federated one, and so the Ministry of Finance distributes resources to states who actually do the education, do the health and whatever. In the Solomons you don't have that. The ministry has direct dealings with the service delivery line agencies. That's a very fundamental difference. You've got to be very, very careful about that. You've really got to think closely about that. It may be more useful to show them a state within the Australian context, a Department of Finance at the state level, which is more likely to mimic the type of services and relationships that the Ministry of Finance in the Solomon Islands might have. The other thing is that we all do it, we all build from our own experience, that's human nature; and we talk about how important context is and how we should avoid the cookie-cutter approach to what we do, and yet I think sometimes that we really don't test our thinking terribly much. Inevitably I see-when I really look at it-we take what we know, and we try to apply it in the new context. I think we should be much more intellectually honest and really test and challenge our assumptions about what we're proposing to do in that environment. Sometimes I think that requires that we bring in people who really understand the environment in which we're going to operate, and ask them to critique and challenge what we're proposing to do in a way that really tests it and makes us think back to whether we should be doing a bit more development from the bottom up. I was very interested in the discussion I had-again this is a recent one-with the Carter Center in Monrovia and Liberia. This was talking about looking at providing support to the justice system in that country. When they came in, they were asked by the president to look at helping reform and reestablishing the justice sector. One thing the Carter Center said was, this is not an area they actually had worked in in the past. Having been through the process for a couple of years now, they realized maybe one of the benefits was in fact that they hadn't worked in that area in the past, because what they ended up designing for Liberia was specifically for Liberia, because they had no models upon which to base their decisions. Obviously, as I said, that model changed quickly over time; it's a very dynamic environment, and some of their assumptions in the beginning were not correct, and they had to change that. But in actual fact, in some ways that could have been an advantage. They weren't hampered by the rigidity of a system they had in their mind. This had worked in country X before; "I'm sure it will work now in this country." There's something to be said for that. How you proof yourself against that when you've already got models in your own head I'm not sure, but I do think there is a challenge for us to really think about what we're doing in a way that truly takes context into account. There is a lot of rhetoric around it, but when you really go and analyze what you've done, what you've done is you've just brought in a model that you used elsewhere. Going back to my housing project, I suppose one of the reasons why I think the housing project turned out to be pretty successful was the fact that we didn't really know a lot about the issue, and we put an advisor in who wasn't told how to do it. She was told to go in and understand the system. Out of that we would develop our approach to supporting that ministry. I think that was a very small-scale example, but I think there are lessons to be learned from that. HAUSMAN: Religious, cultural and linguistic diversity can complicate service delivery and civil service management. These are common problems. Could you talk about efforts to respond to diversity challenges as part of the reform effort? ELLIS: Let me look at that one. The countries I've worked in have been hugely diverse, 87 language groups in the Solomon Islands, 800 of them in Papua New Guinea. They are very challenging. How you manage that process-well, I don't think you do manage that process. I think what you do is try to understand it and try to work out what the implications of that diversity is for what you're doing. I think you've got to work with that diversity and not try and fight against it, which leads back to my concerns sometimes about imposition of meritocracy in these systems in the short term. Because as some of the development literature now says, you've got to go with the grain, you've got to work with the culture and not try to fight against it, because ultimately that local culture that has been there for thousands of years is going to overcome you no matter what. What I do think is worth doing, I suppose, coming from a post-conflict environment, is looking for ways of trying to get people to acknowledge and celebrate the diversity in their own country and work with the diversity across groups. I think maybe the way our interventions work should be designed to maximize interaction between different groups, in terms of gender, language groups, religious cultures, that sort of stuff. I think there is a lot of value in looking at projects designed purposefully to make that happen. I've seen community development projects that have been designed in a way that meant that they could only proceed, the resources would only flow, if the development project was one that was of benefit to more than one community or more than one cultural grouping, which in many ways forced, or at least encouraged those groups to work together for a common goal. I think, ultimately, these are very long-term processes, but if you're moving to a stable environment, what you want to do is get people relying on more than their own social group. So if it is common investment between communities, the likelihood is that you'll get a stronger cultural fabric over time. It contextualizes it in terms of culture, because I think the danger is that you might reinforce difference rather than emphasizing and acknowledging the complexity of the environment you're working in. HAUSMAN: Well, thanks so much. Are there any other issues you'd like to raise? ELLIS: No, not really, I think that has covered at least some of the views I've had from these couple of years that I've worked, in the Solomons and 15 years working in other environments. I just think the caution is to be very humble about what you're going to be able to achieve in the timeframes you're given. We do have challenges, we talk about context, we talk about leadership, we talk about being there for the long term, but the political realities of our own institutions are that that sometimes can be challenging. We talk about capacity development, but I think sometimes we don't look at our own capacities. What are the capacities of our own agencies, what are the capabilities our agencies need to do this work? I don't think we necessarily always have the skill sets ourselves to do this well. HAUSMAN: Thanks so much. ELLIS: You're welcome. Innovations for Successful Societies Series: Civil service Oral History Program Interview number: N-1 ______________________________________________________________________ 15 Use of this transcript is governed by ISS Terms of Use, available at www.princeton.edu/successfulsocieties Use of this transcript is governed by ISS Terms of Use, available at www.princeton.edu/successfulsocieties