Public group: Public Sector Reform & Decentralisation




Summary of the 'Grand Panel' discussion

Advanced seminar on Decentralisation
Brussels, 2-5 July 2012

During the concluding day of the seminar a ‘Grand Panel” was organised with three leading experts in the field of decentralisation: François Vaillancourt (CIRANO and Université de Montréal), Prof. Leonardo Romeo (New York University) and Jean Pierre Elong Mbassi (Secretary General of the United Cities and Local Governments of Africa, UCLGA). The panel sought to address three major questions: Why should donor agencies care more about decentralisation? How can decentralisation be “instrumental” to development? How can donor agencies like the EU provide smart support to decentralisation? This note summarises the main outcomes of the intense and thought-provoking debate , which carried important strategic and operational messages for EU development practitioners across the board.


Consider first a basic fact:  in the past decades donors have often bypassed local authorities, either by channelling ODA through line ministries or through grassroots organisations. As a result, decentralisation was all too often a “missing link” in development assistance strategies. Evidence suggests that this relative neglect of the local level has contributed to the rather mixed track record of (EU) aid.

It is time to rethink this approach. Several factors justify a qualitative change in dominant donor perspectives on decentralisation:

•    “The world is decentralised and we have to work with reality”. Looking at decentralisation is a natural thing: nowadays, the large majority of partner countries have engaged in some form of decentralisation with varying degrees of commitment, motivations and results. Each country has its own trajectory and specific mix of modalities (devolution, de-concentration, delegation and privatisation). Further, decentralisation is frequently debated in donor countries’ politics (e.g. Canada, Spain, France, Germany). This shows that decentralisation is not primarily linked to the development agenda but rather to domestic structural reforms and state building across the world. Decentralisation is therefore not a ‘donor fashion’ or an intellectual distraction. It is just a matter of sharpening our understanding of the ‘real world’.

•    Need to build legitimate institutions at local level. Local authorities are new actors with limited capacities. They have to evolve in contexts where the culture of representation is still young and where their legitimacy still needs to be built as part of a larger state building process. A key lesson of forty years of development assistance is that “we do not favour democracy by skipping local authorities (…). Why should people invest politically in their local government if they have nothing to offer?” [Romeo]. By providing contradictory signals and incentives regarding the mandate of local authorities, development assistance may undermine democratic advances and domestic accountability. Furthermore, it does not create an enabling environment for local authorities to deliver services. The resulting lack of credibility alienates citizens from their closest representatives and public authorities, with the risk of closing down spaces for development-oriented state-society relations. As a result decentralisation’s failure -so often claimed by resistant elites- becomes a self-fulfilling prophecy that may serve vested interests in keeping a status quo. And this may have other larger side effects: it is an illusion to think that the impact is circumscribed at local levels. The loss of credibility of local authorities may affect all democratic institutions across all levels of governance in already fragile states.

•    Moving from ‘trickle down’ economics to local development. The structural adjustment programmes of the 1980s in Africa helped to create a ‘territorial vacuum’. The focus was on central government and ministries of finance:  “Becoming the darling of the IMF and WB was the main concern and in the process governments tended to forget their own local people’”[Mbassi]. Advocates of such policies believed that the benefits of economic growth would smoothly ‘trickle down’ to peripheral regions and territories. Territories were expected to converge and forge the territorial cohesion as part of state unification. However, this promise has not materialised.  It is time to invest in local development in order to bring back into the equation the territories and the endogenous wealth creation processes. “Local development does not refer to ‘where’ or ‘what’ but to the ‘who and the how’ of development promotion. It is endogenous, open and incremental”.  [Romeo]. Local development is a positive sum game: it brings added value and local resources in addition to national/global resources. It is primarily about creating wealth rather than only striving to reduce poverty.

•    Decentralisation is a strategic tool to implement the EU’s Agenda for change. Closely linked to the previous point of looking at decentralisation as a means to create wealth, is the notion of “inclusive growth”, as put forward in the EU Agenda for Change. This is a highly relevant political priority for EU development cooperation but its realisation hinges on making progress in local democracy building and empowerment of local authorities. Decentralisation should not be limited to the reallocation of functions and resources. The added value of local authorities also embraces their capacity to generate economic resources that can help fostering inclusive growth at local level.

These various factors generate a new rationale for investing in decentralisation for better development. They highlight the instrumental nature of decentralisation. They call upon donor agencies to see decentralisation as a process that is not ‘good’ or ‘bad’ per se. It is a tool whose performance depends on its design and implementation.


The next question is to analyse under what conditions practitioners could effectively use decentralisation as an ‘instrument’ to achieve governance results and local development/wealth creation.  The following pointers are worth considering for a strategic and development-oriented approach to decentralisation.

1)    Building ‘developmental states’ rather than promoting governance

Decentralisation is too often pushed forward as part of the good governance agenda (reflecting a ‘normative approach’). That is not likely to be successful in poor countries. Decentralisation rather needs to be understood within the long-term state building process and requires to be sequenced accordingly. A characteristic of development countries is the fragility and vulnerability of their states. In ‘limited access order societies’ , States do not enjoy the full monopoly of violence. As a result, a system of distributing rents is essential to ensure stability, avoid violence and keep the coalition in power together. These autocracies cannot be turned into democracies overnight. It is a long-term process in which decentralisation can play an instrumental role.  Hence, the nature of the state is key to defining realistic ambitions and an adequate sequencing for decentralisation. “The primary objective is to shift the behaviour of the state from predatory to developmental Democracy may then emerge under the pressure of the new political actors that development itself will have generated and when its absence will emerge as a major handicap for the developmental state project.”. [Romeo]. Among the new political actors, local authorities and their national, regional and global organizations have the potential to be an important driver in moving forward the development state agenda. Donors therefore need to enable them to play that role.

2)    See decentralisation as domestic political process rather than as a donor agenda

The three panellists strongly challenged the widespread idea that decentralisation is mainly a development and donor driven agenda. External actors may have a decentralisation agenda but it is not the only one, nor the most important or determinant one.  The focus should be on the domestic dynamics of decentralisation. Country ownership of decentralisation is not a monolith that can be captured in a single policy position. There is generally a wide diversity of interests, interpretations, motivations and rationales for national stakeholders to either promote or resist decentralisation. Power relations and alliances between these actors cut across government and society and determine the outcomes, features and the interests that decentralisation ultimately serves.

3)    Distinguish top-down and bottom-up demands for decentralisation

It is critically important to properly understand where the push for change comes from in a given societal context. Two different pathways may be followed:

•    Top-down demand. A paradox of most decentralisation reforms is that they are pushed from above. Decentralisation reforms worldwide have “rarely been conducted to foster development but have been driven first and foremost to promote political goals in the interests of political elites‘ and to legitimise the coalition in power” [Romeo]. However, these reforms engender side effects that may unintentionally open new spaces to advance governance and developmental goals.  As Douglas North puts it, “the changes (…) that occur during the transition must be explained as intentional acts consistent with the interests of the dominant coalition, but the results of those changes need not be consistent with their intentions”. Indeed, once the process is launched, its outcomes are unpredictable as they depend on a complex web of alliances and power. Therefore, opportunities for other actors to bargain arise as the process unfolds. This is where the bottom-up demand comes in and where long-term donor support can be instrumental in avoiding major drawbacks.

•    Bottom-up demand. In parallel of elite-driven decentralisation, history shows there is generally also a ‘genuine’ demand for decentralisation coming from different sectors of the population. Africa is a case in point: “People started demanding democratisation and decentralisation as a reaction to structural adjustment policies that tended to forget the fate of local population” [Mbassi]. Various movements of this type emerged around the world and were endogenously generated. This is crucial: without a consistent demand for autonomy from local authorities decentralisation cannot succeed. Their impact hinges on local leadership and active citizenship: “for decentralisation to be successful it needs to be demanded and pushed from below.  This implies that local authorities are able to raise their own money instead of only relying on transfers” [Vaillancourt]. In many places positive dynamics from the bottom-up can be observed. A new generation is coming up with a more strategic vision, new initiatives are being conducted and local authorities are increasingly getting organized to voice their demand.

4)    Look beyond the façade of “frozen” decentralization processes

In many countries donors are frustrated about the lack of progress of decentralization processes. They usually ascribe the stalemate to a lack of political will. It may lead donors to stop providing support at a critical moment of the process. This type of narrow interpretations needs to be avoided. The challenge is to look behind the façade and seek to better understand the reality behind the apparent blockage. Reaching a stage of ‘frozen decentralisation’ is not surprising in a context of elite-driven decentralisation: “Once decentralisation has accomplished the initial political objectives (i.e. to legitimise the coalition in power), the parties in power loose interest on the fiscal and administrative dimensions that would be necessary for the creation of genuinely developmental local authorities ” [Romeo]. However, this does not mean at all that the reform comes to a complete halt. The bargaining process between reformers and conservatives continues at domestic level even if the reform seems frozen.  “What you see as a stalemate of decentralisation is on the contrary a big fight in the public arena to give birth to a new social contract (…).  When things are calm it means there is a fighting arena. In Africa we do not come from a democratic culture where all the issues can be discussed openly. We come from a consensus culture where you discuss sensitive things underneath. And one of the main stakes in this arena is whether democracy and especially local democracy will work or not (…)” [Mbassi]. Seen from this perspective, the ‘stalemate’ becomes an opportunity to accompany champions of the reform through critical moments all along the reform path. Pursuing support becomes even more pressing in light of the stakes at play.

5)    Recognise the critical importance of ‘autonomy’ for local authorities

A ‘frozen’ reform is the symptom of a decentralisation process without autonomy for local authorities. Increasing the levels of autonomy is the linchpin of successful decentralisation and local development. It is the keystone for local authorities to implement their dual mandate. First, their ‘specific mandate’ in supporting the implementation of national programmes. Here the principle of allocative efficiency posits that local authorities are better able to allocate resources in line with citizen’s preferences due to their proximity to the demand. On this basis, local authorities are given ‘specific functions’ to implement central plans at local level. Autonomy is a condition to materialise the promise of allocative efficiency in terms of making good use of their added value (information, proximity) and bringing additional resources to implement national programmes. Second, autonomy is also required to assume the ‘general mandate’ local authorities have (as a politically distinct entity representing local constituencies) to develop their own local policies and programs. To this aim, they have been given a ‘general mandate’. In this framework local authorities can do whatever is in the interest of their citizens as long as they do not impinge on functions that have been specifically and exclusively assigned to the other actors. Autonomy should be understood as local authorities’ ability to fulfil their ‘general mandate’, pushing forward their agendas, priorities and programs while at the same time ensuring accountability to their citizens.

6)    Use fiscal decentralisation as teaching ground for state building & domestic accountability

Panellists turned upside down the widespread idea that higher degree of autonomy leads to higher risks of corruption. There is no evidence to back up this assumption. Evidence suggests that the challenge to implement the ‘general mandate’ rather rests on limited leadership, capacity and civic engagement. The way forward is to see autonomy and accountability as two sides of the same coin: higher levels of fiscal autonomy reinforce accountability. Overreliance on transfers deprives downwards accountability of all substance since most issues are decided at central levels. Therefore, local authorities are more directly accountable to their citizens with their own resources. This, however, requires a “paradigm shift from poverty reduction towards wealth creation and mobilising own revenues” [Vaillancourt]. The power of fiscal decentralisation as an entry point for better accountability was illustrated as follows: “If the citizens of Belgium decide to not pay their taxes the country collapses. But in rent-seeking and extractive societies this would not happen because there is no state-citizen fiscal relation” [Mbassi]. Building such a bond will not happen overnight. Therefore, fiscal decentralisation is a strategic entry point to progressively build the accountability culture from below in view of state building, democratisation, larger fiscal reforms and their contribution to inclusive growth.

Yet, how to move forward the fiscal decentralisation agenda? A major issue is “where to start in the cycle ‘capacities-responsibilities-re sources”. Purist PFM theory is based on the principle ‘resources follow functions’. It sequences decentralisation reforms accordingly: first, you build capacity at local levels. Then, responsibilities can be transferred to local authorities. Only in the final stage can resources be transferred. In this theoretical context, providing direct support to local government finances through ‘Local Development Funds’ falling outside the fiscal system as such is an anathema. An alternative theory, based on a political economy approach, stresses that: “If we wait for capacities and responsibilities there would not be one single decentralisation reform, nor one single decentralised country. This is where the relevance of policy experimentation through local development funds comes in. This would allow to directly inject resources into the budget of local authorities should in order to allow them to learn by doing and gradually develop their capacity, including that of advocating for greater transfers from the state.” But this approach entails risks: “how much money are donors willing to waste until the money is being spent correctly?” [Vaillancourt]. Ultimately the panel agreed that there is no contradiction between fiscal decentralisation and local development approaches. Institutional development closes the loop for local development. In the absence of capacities at local level there will be no sustainability.


What are the implications of the preceding analysis for donor (EU) intervention strategies? How can the EU meaningfully engage in (frozen) decentralisation processes? What principles should guide its action? What areas merit particular attention across the board (i.e. regardless of specific country conditions)?

Three key engagement principles were strongly advocated by the experts:

•    Beware of grand theories but rather focus on reality. This means recognising that decentralisation is not a science responding to fixed laws. It is a complex, unpredictable process that can only be approached wisely from a multi-disciplinary perspective. Different lenses are needed to capture evolving realities, shed light on different elements of the process and understand the contexts in which we work.  It also implies making systematically a distinction between “what should be done” and “what can be done” –taking into account prevailing political economy conditions. The lack of ‘magic recipe’ makes the task at hand more complicated but at the same time generates new opportunities.

•    Think and act politically. Smartly supporting endogenous decentralisation processes from the outside is all about identifying and seizing windows of opportunities at country level. This, in turn, requires an on-going capacity to use political economy analysis to assess whether and how reforms could increase space for meaningful state-society interactions. The task at hand for donor agencies is to accompany the champions of the reform and make strategic use of opportunities as they arise. This requires “liaising with them, understanding their constraints, accepting the detours and the tactical retreats that they might need to take to keep reforms alive and build their constituency” [Romeo].

•    Be aware of your evolving position and leverage in the process. Partnership and dialogue relations are not static. They evolve together with stages of maturity of the reform and with the donor’s own credibility. In the first phase of the decentralisation process, there might be a clear convergence of interests between elites, demand side and donors in pushing forward the reform. Therefore, dialogue and partnership may be smooth. However, once the political objectives have been achieved, political elites tend to lose interest in the reform. They can even exert strong resistance against administrative and fiscal measures that threaten their position. Therefore, policy dialogue may become more sensitive and the donor strategy and discourse may need to be updated and adjusted to explore alternative ways to exercise influence. One of the panellists was quite specific in his advice on policy dialogue strategies: “’when discussing with the government you should abandon the normative discourse. Otherwise it won’t work because you threaten the coalition in power and drive it into an unproductive corner” [Romeo].

In addition to this, the panellists agreed on the critical importance of investing in two main areas –if decentralisation is to be optimally used as a instrument to promote local development.

First, donor agencies are advised to put local autonomy and fiscal decentralisation at the centre of their support strategies: “if I had one operational message for the people in this room I would say: promoting fiscal decentralisation and local revenue generation to counterbalance the dependency on the transfer mechanisms is key to successful decentralisation’” [Vaillancourt]. Such a strategy may help to manage the risk of reform failure (frozen decentralization) while generating multiplier and creating spillover effects. The pace of reform will depend on the maturity of fiscal and decentralisation processes and specific political economy constraints in a given country.

However, this does not mean that donors are disarmed to support autonomy in weak contexts where possibly support is most needed.  In these cases, support can be sequenced: it can start from ‘outside’ the fiscal system through local development funds or performance based grants to provide discretionary financial support to local governments. This phase can serve for policy experimentation, generate new practices and incentives and stimulate the creation of a more solid policy framework over time, including sustainability issues. As sufficient consensus is built around these local practices, donors may seek to establish a stronger link with the system and provide support to the broader fiscal decentralisation reform. In this context, ad hoc local funds can be subjected to an exit strategy and be integrated into regular financial or fiscal intergovernmental relations.  “There are successful examples of this type of sequenced approach, with positive results in terms of capacity, autonomy and accountability” [Vaillancourt].  

The second priority area relates to supporting local leadership and civic engagement. This means providing strategic capacity development (particularly to make optimal best use of local revenues and better fulfil the ‘general mandate’). It also invites donor agencies to recognise local authorities as political actors and interlocutors in policy dialogue processes. Decentralisation will not move forward without creating coalitions of local authorities so that they can better voice their demand.  “Donors should remain engaged with central governments while at the same time supporting local authorities in driving the demand for decentralisation” [Mbasssi].