IFPRI Home About Contact Careers Search  

Research 2020 Vision Countries Publications Datasets Library Events Media Learning  

IFPRI Publications 2020 Publications Search our Database Articles & Book Chapters Datasets Other Languages Order Form

Many developing countries are turning over functions formerly carried out by the central government to local governments. Can decentralization make government work better for poor people?

In 1994, when Jean-Paul Faguet, then working at the World Bank, heard about Bolivia's plan to empower local communities, he was skeptical. Although the plan was called Law of Popular Participation, it was designed by a small group at the central government level behind closed doors. "Popular participation" sounded good, of course, but Faguet doubted that it could simply be imposed from above. What will they legislate next, he and his colleagues at the Bank asked themselves—wealth and happiness for everyone?

In fact, says Faguet, now a researcher and lecturer at the London School of Economics (LSE), Bolivia made government more responsive to poor people in many communities and directed resources to the services poor people desired. "Decentralization put real power over public resources in the hands of ordinary citizens," Faguet says. "It changed the way the country is run."

Bolivia is among the most striking cases of decentralization of government power to the local level, an idea that has been discussed and implemented off and on throughout the developing world over the past several decades. With the emphasis on democracy and good governance in the development community since the early 1990s, decentralization has again become a popular goal, at least in principle. According to one review, 80 percent of developing countries have pursued decentralization, or attempted to, including Bangladesh, Brazil, China, Colombia, Ethiopia, Ghana, India, Nepal, the Philippines, South Africa, and Uganda.

"It's a popular stated goal," says Stephen Ndegwa, a senior public sector governance specialist at the World Bank. "Most constitutions and discussions in democracies state that they'd like to have a decentralized state."

Joachim von Braun, director general of IFPRI, sees this trend toward decentralization as driven not only by democracy, but also by economic globalization. "Globalization requires local decisionmaking power that will efficiently provide the infrastructure and services demanded by investors," he says. "This economic necessity drives 'glocalisation'—the combination of globalization with localization and decentralization."

When it works properly, decentralization can help to alleviate poverty and food insecurity by providing infrastructure and services that poor people require, like drinking water, roads, schooling, and health care. "The goal is to bring government closer to the people, with the hope of giving poor people a greater voice and making government more effective and more accountable," says IFPRI senior research fellow Regina Birner. Given that most poor people in developing countries live in rural areas, out of sight of the political elites in national capitals, "decentralization can be the single most important governance reform for rural areas," she says.

But making decentralization work effectively for poor people is a challenge, and it takes time. A 2004 study from the Organisation for Economic Co-operation and Development (OECD) examined the impact of decentralization on poverty in 18 developing countries and 3 states of India. Decentralization helped to reduce poverty in only one-third of the cases, and in some of the poorest countries with weak institutions and post-conflict situations decentralization worsened poverty.

So is decentralization the key to governance that focuses on the needs of poor people?

Decentralization can look very different depending on what kinds of functions central governments transfer to the local level. Take, for example, provision of drinking water. If drinking water provision were politically decentralized, locally elected government officials would make decisions about where and how to provide drinking water. Under administrative decentralization, the local government would have staff in charge of administering drinking water provision. And under fiscal decentralization the local government would have the revenue to finance drinking water provision.

The implications of these types of decentralization for local government effectiveness and for poor people are subject to debate. Von Braun says that political decentralization can indeed benefit poor people by involving them in planning and evaluating local public policies. Administrative decentralization alone will not help the poor, and fiscal decentralization can help reduce poverty or not, depending on how well local governments operate and what political power the poor have. To create the conditions under which decentralization can alleviate poverty, he says, "political and administrative decentralization should precede fiscal decentralization. Otherwise participation and accountability are not assured."

According to Ndegwa, on the other hand, transferring some of these functions to the local level while retaining others can undercut local effectiveness. The sticking point is often fiscal decentralization. "Central governments begin the process and even have local elections," he says, "but they often don't transfer the money to the local authorities, and without the money, you can't really do much."

It is perhaps understandable that central governments become reluctant to go all the way with decentralization. Faguet points out that decentralization is not a policy prescription with predictable results, like, say, lowering tariffs. "It's a process with very uncertain outcomes," he says. "The center has to let go of power and resources and pass them to local government. You don't know what's going to happen, and you have to live with that."

When local governments gain power, do they actually empower poor people? Are their decisions about delivery of infrastructure and services any different from those of central governments? Research shows that in many cases local governments are indeed more responsive to the poor.

India undertook major decentralization through two constitutional amendments in the mid-1990s. Under the 73rd Amendment, one-third of the seats on locally elected village, block (multivillage), and district councils, and one-third of the headships of these councils, are reserved for women, who were traditionally underrepresented politically. Seats and headships are also reserved for scheduled castes and scheduled tribes—two disadvantaged minorities—in proportion to their population share. A study by Raghabendra Chattopadhyay of the Indian Institute of Management and Esther Duflo of the Massachusetts Institute of Technology showed that mandated representation of these groups affected the investments made by local governments. When women were heads of village councils, the councils invested more in women's concerns, especially drinking water.

Marc Cohen, an IFPRI research fellow, agrees that where decentralization of government services is coupled with efforts to empower communities, it can contribute to food security. He studied the effects of a program known as the MERET (Managing Environmental Resources to Enable Transitions to More Sustainable Livelihoods) plan in Ethiopia. Covering 600 communities, the plan combines an increase in agricultural extension services provided by the district-level government and the creation of elected, community-based planning teams that decide on and manage soil and water conservation activities. "The program has contributed to improved soil fertility, reforestation, higher incomes, and improved food security and nutrition," says Cohen. "Past top-down approaches to soil and water conservation failed."

How well local governments serve the poor depends heavily on the level of transparency—do local people know what government officials should be doing, and can they tell when officials are shirking their responsibilities? Bolivia's Law of Popular Participation, for example, included a provision for elected local Oversight Committees to ensure that municipal government spending reflected local priorities. But oversight was also made easier in Bolivia, says LSE's Faguet, because decentralization was based on a simple plan that distributed money to municipalities on a per capita basis, with no special formulas for different types of populations. As a result, people knew how much the local government should be spending, says Faguet: "People would march to the municipal building and say, 'We know you've got so much money per person—where is it? Where are the roads and the other public services?'"

In the 1980s, China made a massive shift from appointing of village leaders to holding local elections. An IFPRI study by senior research fellow Xiaobo Zhang and others examined how democratic local government changed the provision of public goods like roads and irrigation. "Rural roads are one of the most important local public goods in China because they connect farmers to the market and allow them to switch from subsistence to commercial agriculture," says Zhang. "Our research has shown that rural roads have a big impact on poverty reduction and agricultural growth in rural China." By comparing the provision of public goods before and after elections, Zhang and colleagues found that elected local officials tended to tax farm households less and provide more public goods. The researchers attribute this change partly to the fact that budgets are now more transparent, and elections allow villagers to hold officials accountable.

Where transparency is lacking, poor people have been less satisfied with the services provided by local government, a study from India shows. IFPRI's Regina Birner and others examined local governance and poverty in two Gram Panchayats (village councils) in the state of Karnataka, India. In the case of drainage, for example, village residents expressed high levels of dissatisfaction, and one-third to one-half of them did not know who was responsible for drainage service in their community.

One of the challenges in implementing decentralization is getting central governments to loosen their grip on power and turn it over to localities. "Once politicians get into power," says the World Bank's Ndegwa, "they have second thoughts. They don't want to give up power, or they see the risk that the state will disintegrate." But decentralization can work to the central government's advantage, he points out, by reducing people's frustration with the national authorities. "Once you give people some local autonomy, they don't have the center to blame any more. But to convince a politician of this is difficult," he says.

In 1995 and 1996 the state of Kerala in India implemented a wide-ranging decentralization scheme, overcoming opposition through rapid and extensive transfer of power and resources. According to the website of Kerala's Department of Local Self-Governance, "Real and effective decentralization probably calls for a big bang approach—functions, powers, and resources are transferred at one go. If decentralization is effected in one fell blow, the suddenness would stun potential dissenters into silent acceptance; before people realize what they have lost, decentralization would have become a fait accompli." The decentralization went hand in hand with a "People's Planning Campaign" that drew public participation and created demand for decentralization. A large share—35 percent—of planning funds were given to local governments.

Bolivia's government had its own reason for deciding to decentralize, Faguet explains. National leaders were worried that business elites in one region were becoming too autonomous, so they decided to leapfrog the regional governments by decentralizing to the local level.

When power and resources flow from the central government to localities, there is a risk that local elites will seize control, monopolizing the benefits and leaving out the poorest and most vulnerable people. Dilip Mookherjee, professor of economics at Boston University, explains, however, that this does not necessarily occur. A study of the Indian state of West Bengal by Pranab Bardhan of the University of California, Berkeley, and Mookherjee found that Gram Panchayats both gave poor people a voice in local government and directed resources toward services used by the poor. West Bengal may not be a typical case, he explains. This state has been run by the Left Front, a coalition of leftist parties, since the 1970s, and it decentralized well before the constitutional mandate of 1994.

According to Mookherjee, the success of West Bengal in achieving generally pro-poor local government may be due to its relatively high level of social and economic equality. "West Bengal is far less hierarchical and unequal than other Indian states. In different states with much higher inequality, decentralization has had almost no effect," he says. "But we don't know enough about the success in West Bengal relative to other states. We need more studies on this."

Anwar Shah, program leader for public sector governance at the World Bank, points out that in communities where there is one large landlord or industrialist, that person can buy votes and install people in local government to serve his or her interests. "In these situations, you need land reform and democratic empowerment," he says.

"It's also possible that decentralizing government functions will decentralize corruption," explains IFPRI's Birner. Poor people may be no better off under a corrupt local government than under a corrupt centralized one. Corruption is often more visible, however, at the local level, she points out. People see, for example, who can suddenly afford a big house. Therefore, decentralization may increase the possibilities for fighting corruption.

Shah says that corruption may increase in the early stages after decentralization, but in the medium and long terms corruption should decrease. "Citizen awareness may take some time to materialize," he says. Fighting corruption at the local level requires not anticorruption campaigns, says Shah, but effective rule of law, including a well-paid independent judiciary, a professional police force, and political reforms that give people the right to vote. "This is not an easy agenda," he says, "but without these basic requirements, anything you do to reduce corruption will not really work."

The design of local government can also influence the level of corruption, says LSE's Faguet. Any community has a supply of possible local politicians that consists of both honest, hardworking people and dishonest, lazy people. "What conditions would attract the honest people and not the dishonest ones to work in local government?" he asks. "Well, there wouldn't be huge pots of money available, and there would be a lot of oversight."

"Political decentralization improves poor people's ability to demand services," says Birner, "but sometimes you need to improve the government's ability to meet this demand." Civil service reform may be one element in improving local capacity.

Training can also make a real difference in local government effectiveness. From 2002 to 2004 the Asian Development Bank (ADB) undertook a project to help women in local government do their jobs and serve their constituents, mainly poor women, more effectively. The project set up forums where women representatives in local government met with representatives of government line agencies, nongovernmental organizations (NGOs), and the private sector, as well as with people in the poor communities. These forums increased the transparency of local government and led to better delivery of public services to the poor.

The project also trained women how to understand budgets, keep records, set agendas, manage projects, run meetings, mediate disputes, and negotiate development programs. It offered gender sensitivity training to men in local government, after which "there was more cooperation from male representatives," says Monawar Sultana, social development specialist at ADB.

Decentralization alone is no guarantee of more effective or responsive government, says IFPRI's Cohen: "Unless there is also community participation in making decisions and managing development, decentralization may be no more responsive to community needs and desires than centralized authority."

Moreover, decentralization is not the answer to every problem. In fact, some issues require attention at a higher level of government, points out Birner—even issues that affect the lives of poor, rural people. For instance, agricultural extension and advisory services may be best managed at the local level, but agricultural research aimed at producing improved crops for poor farmers is probably better done at the level of the agroecosystem in which farmers live. Similarly, irrigation requires management at several levels. "Local people can manage the irrigation canals serving farmers," says Birner, "but someone needs to manage the whole system."

Still, Birner points out, there is much evidence that improving access to basic services is effective in reducing poverty. Consequently, as the experiences of Bolivia, China, and other countries have shown, when decentralization involves poor people in deciding on and managing the services their local government will provide, it can give them the power to transform their lives.