Our Programmes



Sign up to our newsletter.

Follow ODI

Does decentralisation harm the poor? Lessons from forestry in Indonesia, and West and Central Africa

Time (GMT +00) 09:30 11:30


Dr Jesse Ribot - World Resources Institute

On ‘Is Decentralisation Good for the Environment? Some Reflections on Power Transfer and Institutional Choice’.

Dr John McCarthy - Van Vollenhoven Institute, Leiden University

on ‘Is decentralisation good for the environment and poverty? Lessons from the forestry sector in Indonesia’.


Dr Kate Schreckenberg - ODI

1. Is Decentralisation Good for the Environment? Some Reflections on Power Transfer and Institutional Choice’ or, ‘Decentralisation: Democratic Reform or Fig Leaf on the Reproductive Organs of the Central State’.

Unusually for an ODI lunchtime meeting, Jesse Ribot, WRI opened his presentation with his own poem on decentralisation The Business of Sustainable Development: An Africa Forest Tale. This outlined a history of forest resource appropriation in West Africa by both colonial powers and the modern state. The poem ended with a warning that central government may seek to reign back local democracy by empowering "customary" leadership.

Dr Ribot argued that decentralisation is about the transfer of powers and the actors to whom they are being transferred. It is about local democracy, i.e. discretion in the hands of representative, locally accountable authorities. And it is also about citizenship. Most decentralisation theories identify two essential ingredients:

local accountability of leaders to their people, enabling greater efficiency and greater equity;

discretionary powers in the hands of local leaders so that their powers have meaning to local people who then have an interest in accountability.

Dr Ribot added that accountability is more than elections; it is also based on procedures, culture, knowledge and identities.

However Dr Ribot asserted that it was too early to judge the effects of decentralisation in Africa on natural resources. While decentralisation may be legislated, it is often remains to be implemented or may be incomplete. And its effects may be hard to separate out from other processes such as liberalisation, globalisation, donor interventions, economic crises and natural disasters. A long-term perspective is needed for lengthy adjustment periods. But Dr Ribot argued that it was at least possible to examine the laws and changes on the ground in terms of who is attributed which powers.

The rest of Dr Ribot’s address dealt with powers being transferred, the choice of local institutions and, in particular, the ways in which decentralisations are attenuated or undermined, including through the reintegration of "customary" authorities back into local representative functions. He drew lessons from Mali, Cameroon, Uganda, Zimbabwe and South Africa, as well as a range of Asian and Latin American countries.

Dr Ribot highlighted some positive experiences. These include: the transfer of rights to reserve forests and to allocate production opportunities, concessions rights and fees on timber production; active management of forest resource by local authorities in the expectation of decentralisation (the ‘Expectation Effect’); as well as greater retention of benefits in local communities.

Overall, however, governments are putting the brakes on the transfer of powers and are attenuating the efficacy of decentralisation through their choice of local institutions. Decentralisations therefore rarely create the space for discretion that is critical for local democracy to be meaningful. Typical brakes on the transfer of powers include technocratic arguments such as destruction of natural resources as well as local conflict. Transfers of power may therefore be delayed or left to Ministerial Decrees, while specious arguments are made to keep certain decisions central. Public goods such as forests may also be privatised, taking power away from decentralised bodies.

Problems also arise over the choice of powers being transferred. Dr Ribot highlighted the transfer of burdens rather than powers, and of non-commercially valuable uses. Decentralisation may in fact amount to a form of deconcentration or "co-administration", where decentralised elected bodies may be treated as implementing agencies for central agencies. There may also be overbearing oversight, including complex management planning exercises, as well as conditionalities on capacity.

We need to ask how the empowerment of different actors shapes the prospects for democracy. In this light, Dr Ribot thought that the re-emergence of customary authorities in a range of African countries was particularly troubling, and was reproducing past patterns of indirect colonial rule. Dr Ribot argued that customary authorities, while depicted as having bottom-up legitimacy, are being deliberately resuscitated by central governments. As vote banks and strong members of patronage networks, customary authorities are easier to control than less predictable elected authorities. Particularly troubling are donor-led efforts to find and empower the "real", "traditional" natural resource managers - which takes powers that could have been attributed to democratically elected local authorities.

Dr Ribot added that goverments often argue that they give powers to customary rather than elected bodies because they are more "legitimate". But Dr Ribot argued that power itself legitimates and the reason why elected bodies are not legitimate is often because they have nothing to offer local people. The facts that

legitimacy follows power

society rallies around power

raise important sequencing issues for decentralisation. In summary, said Dr Ribot, powers and institutional choice matter deeply to whether decentralisation is democratic or not, and to whether it is effective. There are few cases where discretionary powers have been transferred and democratic institutions chosen.

In conclusion, Dr Ribot argued that decentralisation cannot be said to be bad for the environment, and there are lots of arguments to say that it can be good for the environment and for local people. He highlighted the need to respond to the range of arguments frequently made by central administrations against decentralisation.

In response to Hardenist Ecological Destruction arguments, we need to respond with good environmental analysis, and propose a set of principles of environmental subsidiarity. This includes questioning the need for complex management plans. Dr Ribot argued that minimum environmental standards may better match a decentralisation strategy than a planning approach prescribed by forestry and environment ministries.

In response to conflict arguments, the risks need to be assessed as they are often exaggerated. Are there means to transfer powers along with mediation devices so that conflicts don’t become violent?

In response to capacity arguments, we need to present multiple counter examples. Furthermore, powers cannot be transferred without any fiscal resources. Stopping transfers due to the lack of finances is often specious and in response we need to propose powers which do not require revenue.

In response to legitimacy and institutional efficiency arguments, Dr Ribot argued for principles to guide institutional choice, and the need to talk about the political nature and implications of such choices.

Dr Ribot stressed the need for constituencies and counter powers including accountability. He highlighted work in Mali in which mayors are being encouraged to address the management of urban fuelwood cutting for themselves, in the hope that this might force the forest service (DENC) to loosen control over this lucrative sector. To identify such areas of counter power, Dr Ribot stressed the need to map mechanisms by which central authorities retain control. He argued that we need to get the experiment underway so that we can ask what the effects are of democratic decentralisation on the environment, poverty and justice.

2. Is decentralisation good for the environment and poverty? Lessons from the forestry sector in Indonesia

John McCarthy, Van Vollenhoven Institute,opened his talk with an anecdote of local government corruption, involving collusion over logging rights between members of the local parliament and the district head (Bupati) of an area in North Sumatra. The logging subsequently led to severe flooding in the area. Dr McCarthy thought this story exemplary in two respects. First, it demonstrated how party politics can conspire to undermine local parliamentary democracy. Second, it showed how local people and the press can nevertheless mobilise to secure transparency and accountability.

Dr McCarthy went on to outline the structure of his talk spanning:

decentralisation within the Indonesian context;

the extent to which Indonesia has devolved management responsibilities for forestry;

the extent to which decentralisation is affecting forest management and communities;

preliminary conclusions.

(i) The Indonesian context

Dr McCarthy explained that a previous round of decentralisation in the 1970s had left local authorities accountable to central government but not to local communities. Centralised decision-making under the Soeharto regime led to rapid forest exploitation that increased the vulnerability of the poor. In fact, this level of exploitation was a driving reason for the current cycle of decentralisation initiated in 1999. However, the decision to decentralise in 1999 was taken during a moment of crisis when the then Habibi regime was in a weak position. The laws were rushed through in the hope of regaining the legitimacy to the ruling Golkar party. In the process, they empowed the districts more than they did the provinces. However, their implementation has proved difficult and complex. With 3 Presidents since 1998, no fixed policy on decentralisation has emerged.

In the meantime, deforestation has reached up to 2 million hectares per year with illegal logging at unprecedented levels (up to 55.6 million cubic metre/year). It was hoped that decentralisation could help by: improving participation of local people in district-level decision-making; ensuring that district-level decisions are based on local knowledge; and ensuring decisions reflect the common interests of local communities. Dr McCarthy added that decentralisation was in fact first introduced by the Dutch, but 5 decades on it has delivered few benefits, with no simple recipes or formulas to guarantee success.

Dr McCarthy went on to outline the legal framework for decentralisation. The framework law (Law 22 of 1999) was a compromise of differing agendas and contains numerous ambiguities. There is hence lack of clarity over the rules at every level: the law’s implementing regulations consist of complex formulations; districts have been able to create their own legal instruments at odds with higher regulation; and sectoral laws, inherited from the Soeharto era, remain at odds with the regional autonomy law.

(ii) The extent of decentralisation in the forestry sector

With decentralisation, Dr McCarthy explained that districts now share responsibilities with central government over forest management though have powers to regulate their own agencies. District Heads are also able to grant small concessions of up to 50,000ha. However these powers are circumscribed. Districts continue to lack of power over spatial planning, and most of the forestry estate remains under the control of concessionaires and/or has unclear status.

Dr McCarthy went on to examine the relationship between the district and central government. With respect to the relationship upwards, Dr McCarthy argued that the 1999 framework law had weakened the upward accountability of the districts, while central government at the same time failed to set up an effective administrative apparatus for guiding the transition. The districts are also unable to effectively influence the policy process at the centre. This disjuncture between the districts and the centre has given rise to conflict, with agencies at each level interpreting the law to support their position in struggles over jurisdiction. With respect to the relationship downwards, Dr McCarthy argued that the framework law in fact increased the accountability of district heads to local assemblies, but that the accountability of the local assemblies remained questionable. Under the existing electoral model, voters choose a party and the party committee then selects the candidates. This usually secures MPs who are predominantly loyal to the party and enables widespread money politics, where cash is exchanged for favours.

(iii) The impact of decentralisation on the forest sector and local communities

While the 1999 framework law transfers greater responsibilities to the districts, Dr McCarthy argued that they lack a tax base and are still dependent on central allocations. This leaves districts in fiscal crisis, having to improvise to raise revenue and stabilise their administrations. This problem has not been adequately registered in Jakarta.

Dr McCarthy went on to discuss the district policy process, highlighting the numerous deals that local politicians and officials may strike with the private sector, to meet both political and economic needs, and to raise district revenue and finance regional autonomy. Indeed, the survival of the District Head (who may otherwise be sacked by the District Assembly) depends on his revenue-raising capabilities. In timber rich areas, district heads have sought to raise revenue through the grant of small-scale concessions. For each concession permit granted, a local administration may raise between 50 – 100 million Rupiah each. In any one district, up to 300 such concessions may be granted without regard for social and environmental factors. These concessions may overlay existing concessions granted by central government, and logging frequently takes place well beyond concession boundaries. Dr McCarthy added that gaining access to timber now depends on local timber brokers who then negotiates horizontally with the district assembly and local government, as well as downwards with local community heads. Local timber brokers may in fact be local government officials or may sit in the assembly.

Dr McCarthy then addressed the impacts on local communities, arguing that this local timber economy fails to provide stable livelihoods. Generally, only some members of local communities benefit from logging while others watch, given the lack of money to procure permits and the lack of power to participate in decision-making processes. Many local community members only marginally benefit from the liquidation of timber resources, and conflicts frequently break out between those who have lost out and those who have gained.

(iv) Preliminary conclusions

Dr McCarthy argued that there had been a break-down in vertical accountability, especially between the districts and the provinces. Furthermore, although horizontal accountability between district and provincial assemblies and district and provincial executives is now strong, downward accountability to local communities is extremely weak. Two key problems remain. First, central government now has difficulty in ensuring national policy goals are reached. As a result there is an emerging inclination to recentralise. Second, local communities remain politically and economically marginal to the process, as central government and the regions strike deals over the division of power and resources.

In conclusion, Dr McCarthy said that decentralisation was a long-term process requiring at least another ten years. In that time, a transformation in the relationship between local government and local communities is needed towards greater accountability.

3. Summary of Discussion

A member of the audience asked Dr Ribot to expand on the lack of fiscal resources as an excuse to stop the transfer of powers and on the importance of local resources, their collection and taxation, to successful decentralisation. He also asked the speakers to expand on the link between local accountability and livelihood opportunities.

Another participant felt that Dr McCarthy’s depiction of decentralisation in Indonesia was overly negative. There are also positive experiences. These include the appearance of NGOs in some areas, who have quickly grasped the opportunities presented by decentralisation and developed programmes to educate newly empowered local assemblies in how to prepare a local budget, how to make it transparent, and how to pass local regulations. They have also worked to draft regulations on natural resource management and indigenous peoples’ land rights.

Another participant, speaking as an economist, felt that the speakers had not provided any models of decentralisation which best captured the use values of forests, including jobs, industry, export revenue and environment services. He argued that, while democratisation has an instrinsic value, it also has an instrumental value.

In response to this last remark, Dr Ribot argued that it was easy to look at theory, including public choice theory, democracy theory as well as economic theory. Theory tells us, argued Dr Ribot, that downward accountability of discretionary powers leads to better matching of services. But when you devolve power to party-chosen authorities who are really not elected (as in Indonesia), that kind of matching might not occur. It cannot be judged as decentralisation. Elections do not, therefore, tell you about downward accountability if local mayors are put in place by party political machinery. And electoral processes do not always open up universal suffrage. Although local councils in Uganda include female representatives, they still speak on behalf of their fathers and husbands. So you need to look at other mechanisms (both historical and cultural) that bind leaders to people.

Mr Ribot added that, in order to assess the material effects of decentralisation, you must first have a decentralisation. And there are a few places where decentralisation has arguably been genuine. In Kumaon in Northen India, where decentralisation took place in the 1930s, local people have demonstrated their ability to look after forests in a sustainable manner. Livelihoods benefitted from broader-based access to forest resources than the British had previously allowed. Nevertheless, the decentralisation process did not permit commercial exploitation by local people. Export value was therefore eliminated in favour of subsistence values.

Dr Ribot then addressed the issue of whether powers over local and fiscal resources are essential to successful decentralisation. In forestry, are there powers that can be devolved that do not also threaten the wellbeing of the resource, and that do not require local capacity and fiscal backing? Dr Ribot argued that there were. In Mali, natural regeneration is so robust that the transfer of powers over fuelwood management is unlikely to lead to long-term resource loss. There are many such powers that might constitute the first part of a sequencing in decentralisation.

But with respect to taxation, Dr Ribot argued that there is no such a thing as local representation (i.e. downward accountability) without local powers over taxation. Taxation provides a basis with which to hold government to account; it defines the relationship between the population and the state. He recommended Mick Moore’s work on this.

Dr McCarthy argued that he was not trying to be too negative in his depiction of decentralisation in Indonesia. But he did wish to emphasise that there is continuing ambiguity at all levels, that the transition has not been smooth and that it has given rise to a lot of disgruntlement. However, this was not surprising given the context. In any case, he could not now see how Indonesia could return to a centralised system. He added that Indonesia is so huge that it is difficult to pick up on the many responses to decentralisation, both positive and negative. His own experience had been in Central Kalimantan, where civil society is poorly developed, and in Aceh which is now a war zone. With respect to decentralisation and customary authority in Indonesia, Dr McCarthy thought this was a very interesting issue, though it is too early to draw any conclusions as to whether Jesse’s theory holds out for Indonesia. It is a can of worms yet to be opened.

One participant asked which countries had got decentralisation most right and why? Another question tackled the relationship between elected administrations and customary entities. A participant from Nepal asked how we could reconcile the need for strong community-based institutions and governance with decentralisation policy. Another participant commented that it was possible to go too far in either direction (central and local), given that these two levels of administration do not necessarily fit. He asked whether there were any examples of institutions that successfully operate across multiple levels.

Another member of the audience from Namibia commented that there had been attempts to decentralise to local institutions that had formal institutional links to the centre. However, this was not decentralisation in the true sense of the word. For that, powers would need to be transferred to a community or a committee accountable to legislation and the national constitution but not government. He also tackled the question of whether decentralisation harmed the poor. He argued that it was, in fact, a right. Given Southern Africa’s negative experiences with centralised governments in the past, decentralisation strengthened rights and is motivating democracy in an effective way. Finally, the speakers were asked why forestry, as opposed to sectors such as health, education and agriculture, offered such valuable insights into decentralisation.

Dr McCarthy thought that forests are interesting because there is no immediate constituency for protecting them. Forests are also a source of revenue (very different from health which is a burden for local government), and many people depend on them for their livelihoods. The sector is furthermore closely tied up with property rights – an issue that needs to be confronted for decentralisation to really work. Forests therefore provide an important entry point for reform.

Dr Ribot also addressed the special value of forests as a lens on decentralisation. He and David Kaimowitz had recently completed a paper comparing forests to other sectors. Even in areas such as the Sahel, it is not just livelihoods but also national wealth that are at stake. In addition, forests are so variable across space and time that their use and management can demand real particularity often best served by local knowledge. This leans in the direction of local decision-making. Furthermore, because forests are wealth generating, they are crucial in legitimising local authority. Rights to forest management give rise to a sense of citizenship. This stands it apart from cookie-cutter, spatially uniform decentralisation in sectors such as health and education.

He added that cases of successful decentralisation are limited. Nepal offers important lessons but they seem confined to lower-value forests in the middle hills. The Philippines is also worth looking at. But within every country, there are spaces of real change and it is important to focus in on these.

Finally, Dr Ribot addressed the relationship between elected and customary authority. In Africa, colonial authority empowered customary leaders to allocate land; they are hence legitimate because they are feared and have real power backed by the state. Dr Ribot emphasised the need for nested local through national powers, whereby people have recourse to higher authorities in the event of local abuses of power. Nested powers are essential for accountability. Legal pluralism can therefore put a damper on the excesses of custom.