Decentralization: Can It Help the Poor?

Across diverse economic and policy sectors, from health care and education to parks and wildlife management, decentralization is one of the most frequently pursued institutional reforms in developing countries today.

Decentralization is a process by which a central government transfers some of its powers or functions to a lower level of government or to a local leader or institution. In the naturalresource sector, an example of decentralization might be transferring from central to local government the responsibility for managing a tract of forest land, including the right to collect some of the income from sales of timber harvests in that forest. Or the central government might give a farmers group responsibility for managing an irrigation system, or grant a village council the right to manage wildlife and run a commercial tourism operation in a national park (WRI et al. 2003:97).

Decentralization is being driven by powerful economic, political, and technological forces. International development agencies such as the World Bank have placed decentralization in a prominent position on their agendas, and nongovernmental organizations (NGOs) and governments alike have promoted the concept, although often for different reasons. Advocates of decentralization cite the potential for greater efficiency, equity, and accountability when decision-making is brought “closer to the people” (Ribot 2004:7; WRI et al. 2003:92-97). In theory, devolving power from central government means empowering local institutions that can better discern how to manage resources and deliver services to meet the needs of local people. Modern communication options like the Internet, television, and mobile phones help make local people and organizations more aware of their rights, more able to communicate and organize, and therefore more capable of asserting their rights.

But are central governments really so eager to give up some of the powers they have traditionally wielded? In the 1980s and early 1990s, decentralization emerged as a priority in an era of economic and budget crises. Shifting responsibility for health care, education, parks, and other planning and service functions to local governments offered opportunities to reduce central government budget deficits. Central governments are all too willing to pass on to local and community institutions the responsibility for managing resources and delivering services without providing them with necessary financial or technical support. They tend to be much more reluctant, however, to give up their powers to collect and allocate user fees, fines, or other revenues (WRI et al. 2003:98).

Areas with rich natural resource endowments tend to be geographically isolated and far from centers of political power where the most momentous development decisions are made. Furthermore, central governments are often run by and for elites, and people from poor rural communities or ethnic minority groups seldom occupy senior positions in the decision-making levels of bureaucracies (Sibanda 2000:3). (See Table 3.1.)

Not All Decentralization Is Created Equal

Some decentralization advocates—governments, donors, and NGOs—view the poor as particular beneficiaries of decentralization. They envision reforms that make policies more useful to the poor, and processes that encourage the involvement of the most socially disenfranchised people in natural resource decisionmaking— those people who have the greatest stake in the outcome of management decisions (Asante and Ayee 2004:3-6, 21-22). These advocates point out that effectively implementing poverty reduction strategies often requires specific local knowledge that is best found in local institutions, and that strengthening local delivery capacity for services requires genuine devolution of authority to these institutions (Asante and Ayee 2004:5).

Some countries have responded positively to these arguments. Bolivia, for example, made decentralization across several sectors part of a package of anti-poverty reforms in the 1990s (Pacheco 2004:85, 90). Most West African countries have also declared local development a prime goal of their decentralization efforts (Ribot 2002:8).

Despite its theoretical potential, the record of decentralization has been decidedly mixed. This is true both in general and with respect to poverty reduction. In some instances, efforts to decentralize management of forests, land, water, and fisheries have shown positive outcomes: rural citizens conserving their natural resources; local councils that are increasing revenues from resource use; the poor more involved in local governance institutions and reaping more monetary benefits from local resources; and local governments providing better basic services. One of the longest-standing cases of decentralized environmental management with evident benefits to livelihoods is in Kumaon, India. Since the 1930s, elected forest councils, called van panchayats, have had the right to manage forest use, raising revenue from the sale of fodder and dead trees and enforcing regulations on forest use (Ribot 2004:22).

Similarly, some wildlife co-management schemes in Africa have yielded improved local infrastructure such as roads and schools, while community forest management in Mexico that has come about through decentralization has enabled communities to build water networks, schools, and clinics (Shyamsundar et al. 2004:9). In Ghana, devolution of power to district assemblies has improved provision of basic services and infrastructure in rural areas through construction of more feeder roads, clinics, public toilets, classrooms, and the like (Asante and Ayee 2004:8).

Yet in most decentralization efforts to date, the intended benefits for local democracy and for the poor remain largely unrealized, due to flawed implementation of the reforms. The choice of which institutions to empower with new management or decision-making responsibilities, and the ways in which those institutions are held accountable to the people, have profound implications for the effectiveness of decentralization—and whether the benefits reach the poor (Ribot 2004:25).

How Decentralization Can Harm the Poor

Governance reforms that are truly empowering for the poor, responsive to their needs, and effective in reducing poverty are rare (Crook and Sverrisson 2001:iii). In a 2001 analysis of decentralization cases from about a dozen locations in Asia, Africa, and Latin America, only Brazil, Colombia, and the Indian states of West Bengal and Karnataka showed good results in terms of increasing policy responsiveness to the poor, or reducing poverty and inequality (Crook and Sverrisson 2001:14-15).

Most reforms in the name of decentralization come up short in two areas that are critical to bringing about benefits to local populations and the poor: they don’t create accountable, representative local institutions, nor do they transfer meaningful powers to them (Ribot 2004:15). Such incomplete or partial decentralization undermines the potential benefits of governance reforms, particularly for the poor.

Decentralization without Accountability

Often, powers over natural resources are handed over to a person or body not elected by the people, and thus not wholly accountable to them, such as a traditional chief, or to a civilsociety organization such as a women’s association, or to a “user group” such as a forestry cooperative, or a pastoralists’ group. Such groups may help broaden grassroots participation in local decisions, but they speak for only a segment of the citizenry. For example, Cameroon’s community forest law devolves power to local forest-management committees. While the law requires these groups to consult “representatives” of all segments of the community, it is unclear by whom these representatives are chosen, and the results of the consultation are not binding in forest management plans (Ribot 2004:35). Similarly, in Uganda, the wildlife authority created a committee of beekeepers, but its mandate was so narrow that only interested parties participated—and these beekeepers then excluded other forest users from the committee’s deliberations (Namara and Nsabagsani 2003 in Ribot 2004:37).

Retention of Central Government Control

Another common implementation flaw is to empower a district office of the government or a local representative of the central government. Such an office or official is accountable only to central government authorities, not to the people in the town or municipality. Central governments frequently choose to transfer power to a local branch of the bureaucracy, rather than a locally elected body, as a means of maintaining central control over natural resources (Larson and Ribot 2004:6). In China, the central government devolved management of community forests in name, but in practice has shifted greater power to the provincial level, and has implemented national-level policies that override and often contradict local policies (He 2005).

Lack of Power to Generate Revenue

Even where local democratic institutions or bodies are charged with natural resource management, they are commonly entrusted with duties that are circumscribed in scope, and rarely with the power to generate revenue by setting fees or levying fines. The central government often retains the most lucrative powers—such as the right to assess wildlife hunting fees or allocate revenue from a logging or mining concessions—while granting rural communities or governments the less valuable rights to subsistence-scale harvesting, such as the collection of firewood or bamboo.

Elite Dominance of Elections, Participation, and Decisions

All too often, the fundamental differences in power between rich and poor warp the decentralization process, allowing members of elite, wealthier, more empowered groups to shape decentralization to their own ends and derive most of its benefits (Ribot 2004:41). Decentralization then becomes largely a transfer of power from national to local elites. In Indonesia, for example, many of the benefits of rural timber extraction during the Suharto era accrued to powerful business interests in Jakarta, the capital, and illegal logging was widespread. In the decentralization that followed the fall of the Suharto regime in 1998, a realignment of influence occurred, with district governments taking more control over managing timber extraction. Now the influence of local elites and business interests predominates. Rather than cracking down on illegal logging, this has tended to perpetuate the cycle, often with similar inequities and environmental damage (McCarthy 2002:879, 881-82; Djogo and Syaf 2003:9-13, 20-22).

Elites can also slant the electoral process, giving them the upper hand in local governance, and, accordingly, in the decisions made about natural resources by those institutions. Fair and competitive elections are a key means to make policies more responsive to the poor, and create a local government that is accountable to local people (Crook and Sverrisson 2001:50). But elites often have a disproportionate influence on which candidates will run for election—candidates that may then be beholden to their interests. Indeed, party politics are often dominated by local elites.

Parties, in turn, often run slates of party candidates, putting independent candidates at a disadvantage. When officials are elected from party slates rather than independently, research suggests that these officials have less accountability, in particular to the poorest citizens (Ribot 2004:27). In contrast, when independent candidates are given a fair shake, elections are more competitive, and the interests of the poor may be better served. Unfortunately, independent candidates are often barred from local elections. In a 2001 assessment of decentralization in 14 countries, only five (India, Mali, Mexico, Uganda, and Zimbabwe) permitted independent candidates in local elections (Ribot 2004:27).

Senegal shows the shortcomings, especially for poor populations, of electoral systems that do not admit independent candidates. In 1998 a new decentralized forestry law granted rural communities and their councils various rights over forests, including the right to authorize or deny commercial production of charcoal by the forest service and wealthy urban merchants— a forest use rural communities had long opposed. Yet years after the forestry law was enacted, the forest service continued charcoal production. Surprisingly, the forest service’s charcoal extraction had the approval of rural council presidents, despite the fact that almost everyone in the communities in the region opposed it. Elected from a party slate, these council presidents were beholden to the party, rather than the local popular will (Ribot 2004:27-29).

Inadequate Participation by the Poor in Decentralized Bodies

Even when decisions and policy-making are devolved to a body made up of independently elected local people, there are inherent biases against equal participation by the poor. Because of their greater confidence, literacy, or other advantages, the betteroff members of a community tend to assume positions of leadership in committees and councils. A study in West Bengal, India, showed that panchayat (village council) members from lower castes or tribes rarely spoke in meetings and, if they did, they tended to be ignored (Westergaard 1986 in Crook and Sverrisson 2001:16).

Moreover, the poorest members of the community are less able to shoulder the costs of participating in decentralized natural resource management, including membership fees, time spent in meetings or monitoring forests for poachers, and providing labor for maintenance of infrastructure such as irrigation systems (Shyamsundar et al. 2004:10). In addition, the earliest participants in projects often have more voice and opportunity to shape outcomes; the poor, joining in later stages, if at all, are less able to garner benefits (Ribot 2004:39).

Shortcomings of “User Committees”

Decentralized natural resource management often fosters the creation of user committees or user groups, which have proliferated in developing countries since the 1990s (Shyamsundar et al. 2004:5). Intended to give ordinary people a voice in local resource management, user committees do draw citizens into the policy process and give them significant influence over some programs. However, these committees aren’t usually democratically elected, and they don’t always benefit the poorest members of society. They also tend to have a short lifespan, which disadvantages poorer members of the community who need more time to develop the skills, confidence, and organizational capacities to participate on an equal footing. The only situations in which poor people are consistently able to wield influence in user committees is when the groups consist largely or entirely of poor people—for example, gatherers of non-timber forest products for subsistence use (Manor 2004:188 in Ribot and Larson 2004).
(See Box 3.2.)

Project Bias Toward Wealthier Villages and Participants

Government agencies, donors, and nonprofit groups engaged in decentralization of natural resources management often have incentives to avoid poorer constituents and invest in wealthier groups or villages with better skills or higher-quality lands needed to make projects succeed. For example, managers of a state-funded watershed development program in the Indian state of Madhya Pradesh tended to work with more prosperous farmers in the valleys, where projects were more likely to generate dramatic results, rather than engaging with poorer hill farmers (Baviskar 2004:30-31 in Ribot and Larson 2004). Similarly, selection for anti-poverty employment programs in the Indian state of Karnataka was based on information provided by village leaders—who tended to be wealthier than other participants— resulting in the inclusion of many better-off families (Sivanna 1990:200 in Crook and Sverrisson 2001:20).

Gender Inequalities in Decision-Making

Women are typically among the poorest and most disadvantaged groups in developing countries. It is no surprise that they tend to be under-represented in positions of authority in local governments, village committees, and other decentralized decision-making bodies to which powers over natural resources are increasingly being devolved. Husbands often do not like their wives to attend group activities, and traditional working patterns and government structures tend to favor men’s dominance in public decision-making. For example, in stateapproved village forest management groups in India and Nepal, women are likely to be relegated to a peripheral role (Shyamsundar et al. 2004:92-93).

In Bangladesh, an analysis of local elected governance bodies, known as Union Parishads, found that women tend to head committees related to community welfare with little influence over disbursement of resources, while men typically ran and served on committees clearly related to resource allocation, like finance, agriculture, fisheries, livestock, and infrastructure (Mukhopadhyay 2003:59). Women also have a much smaller chance of becoming elected officials in local government. A study of over 15,000 municipalities in 42 countries found that only 8 percent of all local mayors are women (UCLG 2003). (See Figure 3.2.)

When women are absent from decision- making, issues that affect them are more likely to be overlooked. The inequity of this situation is all the more glaring in light of the fact that women are often charged with responsibility for collecting and using natural resources such as water, fuelwood, and other resources for the family’s benefit.

New Demands on the Poor

Decentralization that transfers responsibility for managing services and projects to local institutions and communities without providing the financial resources needed to do so can end up creating extra burdens for the poor. For example, in Mongolia, local governments were given new responsibilities for winter preparedness and the cold-weather provisioning of livestock herds, but no new financial resources to meet this responsibility. The result was massive livestock mortality during the brutal winters of 1999-2002, and loss of one-fifth of the nation’s herd (Mearns 2004:137). In other cases, newly empowered local governments may enact new revenue-raising measures that hurt the poor. In Malawi, local governments with new decentralized responsibilities have established village-level enterprise taxes that could stifle fledgling efforts of the rural poor to build their assets and diversity their incomes by starting small businesses (Ellis et al. 2003:1507-1508).

Loss of Access to Natural Resources

Privatization—the transfer of public resources such as forests to private individuals and corporations—is often done in the name of decentralization. This transfer of management authority excludes the public from participation in decisions about the resource and often means the direct physical exclusion of people from the land or water as well, with the poor generally suffering most from such loss of access (Ribot 2004:52).

Devolving power over local resources to communities or groups within those communities can also bring problems of exclusion. For example, a community granted the power to manage a tract of public forest might decide to contract with a logging company in one area of the forest to raise revenue. In the process, it may limit local people’s collection of non-timber forest products in that section of the forest. This can impose immediate costs on poor households who depend on fuelwood and other subsistence products gleaned from the forest (Shyamsundar et al. 2004:10, 95).

Making Decentralization Work for the Poor

Decentralization can be structured in ways that make it more effective and beneficial for the poor.

Ensuring Democratic Accountability

The best way to ensure that decision-makers are accountable to local people and decision-making reflects the interests of local people is to vest powers in elected authorities who are chosen through competitive local elections (Crook and Sverrisson 2001:50). While it is often difficult to rein in the political forces that stifle open elections, the benefits can be substantial. For example, competitive local elections in West Bengal, India helped make policy more responsive to the poor, and in Colombia, competitively elected mayors— challengers to the dominant party politics—brought about better education, roads, and water supply (Crook and Sverrisson 2001:15-16, 42).

Special Measures Promoting the Interests of the Poor

A central government can increase the chances of pro-poor decentralization by making an explicit commitment to promote the interests of the poor at the local level and to ensure that marginal groups get a voice in public decisions (Ribot 2004:41). Elected local governments tend to have a poor record of serving the interests of women, the poor, and other marginal- ized populations unless required to do so by the central government (Crook and Sverrisson 2001 in Ribot and Larson 2004: 6). Special measures are needed to ensure that decentralization benefits the poorest people and most vulnerable groups— women, indigenous people, the landless, migrants, and minority castes. In 1978, for example, the government of West Bengal specifically sought to increase the power of poor and landless peasants by devolving implementation of government programs to the village councils, and mobilizing poor peasants to participate. As a result, 44 percent of those on village councils in Birbhum District are now small farm owners, sharecroppers, or agricultural laborers, and the benefits of government development programs are increasingly going to the poorer members of the community (Crook and Sverrisson 2001:15-16). Kerala State’s approach in 1996 was to give 35-40 percent of the state budget to local governance bodies for development planning, with detailed guidelines to make planning processes both participatory and equitable (Mukhopadhyay 2003:56).

Compensating the Poor for Short-Term Costs

Local institutions can find ways to compensate the poor for any rights they lose when a new management scheme restricts their access to a forest or other resource. For example, the community of San Antonio, Mexico, asked residents to forego cutting pine trees for use as roofing shingles so that they could be harvested as lumber. In return, the local logging business supplied free tin roofing materials and lumber to residents (Shyamsundar et al. 2004:96).

Community-Based Natural Resource Management

One specific approach to pro-poor decentralization of environmental resources is community-based natural resource management (CBNRM). Central governments in many parts of the developing world have begun to shift toward CBNRM in recognition of the limitations of centralized management and in response to the legitimate claims of indigenous groups and local communities to a share in the benefits of local resources. Worldwide, some 380 million hectares of forest land are now owned by or reserved for local communities—over half having been legally transferred to local control within the last 15 years (White and Martin 2002:11). This transformation in forest ownership and management began in Latin America in the late 1970s, moved through Africa in the late 1990s, and spread more recently to Asia. (See Box 3.2.)