Posted December 1997
| Food and Agriculture Organization of the United Nations | United Nations Capital Development Fund | International Fund for Agricultural Development | German Agency for Technical Cooperation | Swiss Agency for Development and Cooperation | World Bank |
Rome
16-18 December 1997
Technical Consultation on Decentralization
This discussion is divided into three sections. The first section deals with matters in which decentralization - on currently available empirical evidence, and in the near future - has considerable promise. The second section addresses matters in which it has more limited promise, and the third section notes areas in which it has little promise.
In considering what follows, readers need to remember two things. First, there is a distinct shortage of reliable evidence about the workings of many (though not all) experiments with decentralization. Second, we are assessing institutions that are still quite young. Only a few originated before 1985, and some of them were abandoned or markedly changed after only a short time. Some of the most adventurous experiments - for example, Bolivia's - have only just gotten underway.
If these exercises are allowed to continue for sustained periods - a big if, given higher level politicians' jealousy of decentralized institutions - their performance on several of the fronts listed below will probably improve. There are signs, in many different countries, that this has begun to happen. So in time, some of the items listed in sections B and C are quite likely to move up to level A or B. The analysis below should be seen not as the final word on this subject, but as an assessment of how things look now - at what is probably an unfairly early stage in the life-spans of most of these experiments. So as we noted at the outset, even where the judgements offered below appear to be forceful assertions, they should be read as provisional assessments which may very well change over time.
Promoting greater participation and associational activity. Evidence from South Asia and West Africa clearly indicates that substantial (though often less-than-spectacular) gains can be made here through democratic decentralization (Crook and Manor, 1994). Participation grows both at elections (in terms of voting and of participating in campaigns) and between them - through increased contact or petitioning of elected representatives and (to a lesser degree) of bureaucrats, through attending official and unofficial meetings, through protests, and the like. It should be noted that poorer, low status groups tend strongly to exercise less influence and to receive fewer benefits than more prosperous groups in such cases. But participation by and associational activity within all groups tends to increase significantly.
This tends to occur even where civil society (organized interests) has previously been weak. We saw in section IV that for democratic decentralization to work reasonably well, it was not essential to have either a lively civil society or an abundance of social capital, but that both things are helpful. Since democratic decentralization fosters both of those things, sustained experiments with it can be expected, over time, to gain strength from both of these things (see, however, "Community Participation in Development," below.)
We should understand, however, that increased participation can also pose a threat. When it occurs, expectations and demands tend to rise. If decentralized bodies distribute benefits mainly among elites or malfunction in other ways, the result can be widespread anger. The same thing can happen if high-level authorities undermine decentralized bodies. Popular anger is not always a bad thing, but it is not what decentralizers intend.
Enhancing the responsiveness of government institutions. One crucial caveat is in order here. Where decentralization is democratic in character, responsiveness nearly always improves markedly. But administrative or fiscal decentralization which is unattended by democratization holds little promise here - despite naive comments to the contrary, unsupported by evidence, here and there in the literature.
Democratic decentralization tends strongly to enhance the speed, quantity and quality of responses from government institutions. Since it usually entails some empowerment and autonomy for elected bodies at intermediate or local levels, or both, those bodies usually possess the authority and resources to respond quickly to problems and pressures from below - without waiting for the approval of agencies at higher levels. The evidence from a wide array of cases indicates that this quickens responses.
Since decentralized bodies have a strong bias towards micro-level infrastructure projects, and since such projects are less costly than large-scale programs often preferred by central authorities, the quantity of outputs from government also tends to increase markedly.
Since democratic decentralization provides interests at the grass roots with influence over decisions within bodies at intermediate or local levels, the quality of such outputs is also usually enhanced - if quality is measured by the degree to which such outputs conform to the preferences of ordinary people.
Solid evidence of gains is available from an extremely broad array of countries and cultures. Opinion surveys in Colombia indicated very high levels of enthusiasm for the spending priorities of municipios. This was true even though different municipios emphasised different sectors (water here, roads there) - the explanation being that local needs and preferences differed from place to place. Similarly high levels of general satisfaction with the performance of these institutions were also expressed (World Bank, 1995, pp. 5-6).
Evidence corroborating such trends emerges from, among others, the Philippines, India, Bangladesh, Sri Lanka and Côte d'Ivoire (Panganiban, 1994; de los Reyes and Jopillo, 1995; Crook and Manor, 1994; and Slater, 1994) [31]. (These findings contradict the misinformed assertion in the World Development Report 1997 that "little comparative evidence is available with which to evaluate the relationship between decentralized government and service quality" (World Bank, 1997, p. 123).
Increasing the information flow between government and people. This is again quite commonly a major gain from democratic decentralization. Elected members and heads of local and/or intermediate authorities usually live near or within their constituencies (unlike many members of higher-level parliaments). Voters know that these people owe their position to a popular mandate and that they (usually) have the power to shape government action, so they tend to put their views to these representatives and to bureaucrats more often than before decentralization.
This results in a considerable, and often a huge increase in the amount of information flowing to persons in government - especially, but not only, where there is a lively civil society. District-level bureaucrats in the Indian state of Karnataka, for example, repeatedly stated that they experienced something like a ten-fold increase in information after decentralization there. Before it occurred, they had thought that they had an abundance of information - from their own administrative networks, state legislators, the free press, and the like - but thereafter they realised that they had been woefully underinformed. They felt empowered by the change and experienced a surge in job satisfaction (Crook and Manor, 1994).
Democratic decentralization also, although to a somewhat lesser degree, enhances the flow of information from government to citizens. This is partly the result of the (nearly always substantial, and sometimes massive) increase in the number of elected representatives who act as conduits for information from the bureaucracy and elected bodies to people at the grass roots. It also owes much to the ability of those representatives to break down popular suspicion of government by explaining official projects in terms that are intelligible to ordinary people. Major gains often occur, for example, in government programs for preventive medicine - vaccination campaigns, efforts to screen rural dwellers for serious diseases which need to be caught early, prenatal and postnatal care, and others - but not only these (Manor, 1995) [32].
Providing early warnings of potential disasters. One immensely beneficial example of improved information flows to government is the provision of early warnings of problems which, if left unattended, can burgeon into disasters - especially droughts, public sanitation problems and outbreaks of disease. Prior to decentralization, when signs of these things began to emerge - especially but not just in remote and under-represented areas - warnings often failed to reach the appropriate authorities soon enough to make action possible. After decentralization - which tends both to provide better representation for remote, underdeveloped areas and to give representatives from those areas the leverage to animate government agencies - early warnings and official responses occur far more often and reliably.
Major gains on this front were reported even from a region of India which had long had democratic representation (albeit only in state and national legislatures), a lively civil society, and a free press given to investigative doings (Crook and Manor, 1994). In places where such things have traditionally been absent, the gains can be expected to be proportionally greater.
Making development projects more sustainable. It is widely recognized in the literature that if people at the grass roots are drawn, even quite marginally, into decisions or just discussions about rural development projects, they develop the belief that they have a stake in their success. And since the quality of responses from government institutions tends to improve in the sense that they are more congruent with locally felt needs, people naturally identify more strongly with development projects.
This does a great deal to make those projects more sustainable. This applies to the management of natural resources, service delivery and much else. See for example a United Nations Development Programme - World Bank study of 121 rural water supply projects (Parker, 1995, p. 44; and Narayan, 1994).
Enhancing transparency. Democratic decentralization holds great promise here. When large numbers of decentralized bodies are thrown open to people who usually live within their constituencies, their neighbours and constituents become far better able to see and understand what goes on within government institutions. The same is true of the elected representatives themselves, who explain their doings and decisions in order to cultivate popular support (Crook and Manor, 1994; de los Reyes and Jopillo, 1995). When multiparty competition occurs within these systems, as it usually does, opposition forces constantly raise questions about the conduct of those in the majority and publicize news (or allegations) of misdeeds. These changes bring to an end the era when something as basic as the amount of funds available for development was known only to a handful of people high up in the system (a situation that facilitated large-scale, clandestine thefts of funds). Politics and the development process tend to become more untidy and contentious, but also more open and intelligible than before.
Promoting greater accountability. For all of the reasons set out just above, and because positions of power are obtained by election, democratic decentralization tends to enhance the accountability of elected representatives to citizens. This does not guarantee that elected leaders will be models of probity, responsiveness or efficiency. But if they fail badly in these respects, voters tend to oust them at the first opportunity. Once that has happened to one or two sets of leaders, citizens acquire a taste for rejecting poor representatives, and their successors in office grasp the meaning of accountability, so that it becomes likely that their conduct will improve over time.
It is more difficult to ensure the other, equally crucial type of accountability - of bureaucrats to elected representatives. That can only be achieved if high-level decentralizers empower lower-level authorities to exercise influence over bureaucrats, and then back them with helpful interventions where necessary. Those are two big ifs, but this sort of thing has happened often enough to warrant cautious optimism here.
Exercising regulatory functions. We note below that decentralized authorities sometimes have difficulty with monitoring, evaluation and planning from below. But the fragmentary evidence which is available suggests that they can be quite effective and - crucially, more responsive to local sentiments - at carrying out small-scale regulatory functions. In the Philippines, such bodies exercised great care over land use issues. They did well "in managing the establishment of cockpits and holding of cockfights, regulating the operation of tricycles for hire, inspection of food products, enforcement of the National Building Code, and legislation and enforcement of environment-related laws" (de los Reyes and Jopillo, p. 81). Cockfights may be a rather specialised interest, but several of these other roles are of considerable importance. (See also, Crook and Manor, 1994.)
Achieving political renewal. The creation of elected authorities at intermediate and local levels opens up a large number of positions of power for people - many of them young - who aspire to political influence. This eases the frustration of such people and deflects them from destructive behaviour which often results from exasperation with limited opportunities. (This can develop even in long-established democracies. In the Indian state of Karnataka, there were 224 elected posts in the legislature before decentralization in 1987, and afterward there were more than 50,000! For corroborating evidence from the Philippines, see de los Reyes and Jopillo, 1995, p. 79.)
When leaders of voluntary associations and other organized interests near the local level seize these opportunities, as they routinely do, it integrates state and society in potentially creative ways. These people develop the skills and attitudes that are needed to make the politics of bargaining and accommodation work. The ensuing gains tend to outweigh the problems (such as elite capture of decentralized authorities) which often attend such changes. This can also greatly facilitate institutional capacity building (World Bank, 1995, p. viii; Manor, 1995).
Reinforcing national-level democracy. This process is closely tied to the item just above, but it is slightly different. Jonathan Fox has ably shown that democratic decentralization to intermediate and local levels strengthens democracy at the national level. This is true in both new and old democracies, although Fox is mainly concerned with recently democratized systems in Latin America. It does so in four inter-related ways:
First, elected civilian regimes cannot be considered democratic until authoritarian enclaves are eliminated and the entire citizenry is effectively enfranchised. Second, pluralist politics must be learned, and subnational governments make a good school. Third, rising democratic leaders can most credibly challenge the corrupt old ways if they are forearmed with successful records in local government. Fourth, the widespread transition from traditionally paternalistic social policies to more efficient and targeted programs depends on balanced partnerships among national governments, and new social and civil actors (Fox, 1994, p. 106).
There is, however, bad news here too. These authorities tend to focus, again narrowly, on small-scale construction projects - road repairs, bridges, school rooms, and wells. This often (but not always) means that the delivery of important services - education, health, agricultural extension, animal husbandry, and the like - suffer at least a little. (But note that in Latin America, there appears to be more interest in service delivery [33].) So on this front, decentralization can be a mixed blessing.
Some governments have eased this problem by earmarking portions of grants to decentralized authorities for important services and sectors, while giving those authorities influence over specific decisions within those sectors (Crook and Manor, 1994, chapters two and three). But it is often difficult to know and to ensure that large number of decentralized bodies are adhering to the rules governing earmarking.
Making development programs more flexible so that they suit local conditions. Where decentralization is substantially democratic, local interests often succeed in persuading elected politicians to allocate funds to those elements of programs which best suit local needs - thus rendering them less inflexible (Ruttan, 1975; World Bank, 1974). We can also include here initiatives to ease the problem of lack of appropriate technology (World Bank, 1987). Local interests tend to press for adaptations in technology and project design to meet their particular needs. If projects cannot be adapted in this way, they often refuse to approve or implement them - which is, on balance, also good news.
Decentralized authorities also tend to excel (when, crucially, they are empowered to do so) at adapting nationally-designed laws and tax codes to local peculiarities, minimizing irrelevant matters and pernicious mandates in ways that harmonize with local conditions. This can, among others, enhance revenues from local property taxes. In the Philippines, this has been the "most remarkable strength" of local authorities (de los Reyes and Jopillo, 1995, p. 88).
There is, however, bad news here too. Elected representatives - especially those standing near the village level - often find it difficult to understand any technologically complex proposal. They therefore tend to direct resources away from all such initiatives, and to favor comparatively simple construction projects as noted just above. This often means that technologies do not get adapted to local conditions.
Changing adverse policy environments for agriculture. When this problem thwarted integrated rural development programs in the 1980s, the adverse policy environment existed at the national level (World Bank, 1987). Decentralization empowers arenas where agriculture is of greater concern, but it also brings with it a strong bias in favor of microlevel construction projects, some of which do not serve the needs of farmers. In other words, the policy environment at lower levels is often only somewhat less adverse.
Reinforcing central government commitment to rural development. This relates closely to the previous section (World Bank, 1987). Decentralization offers certain potential gains here, since rural development will count for more among rural dwellers to whom power is devolved. But three things need to be set against that.
First, as noted just above, many projects implemented by local authorities do little to assist agriculturists. Second, the small infrastructure projects which those authorities prefer often fail to conform to the central government views of what constitutes rural development, and that can create conflicts which may undermine higher-level leaders' commitments to such development. Finally, those leaders higher up are notoriously jealous of the powers and resources given to those at lower-levels. This may tend to erode central commitment to authorities lower down and, in the process, to reinforce central prejudices against rural development.
Giving greater attention to sociocultural factors. When groups at lower levels have influence, they tend strongly to ensure that programs which fly in the face of regional or local customs, preferences or social structures are either adapted or rejected - sometimes clandestinely in both cases. This, again, has mixed implications. It means that resources are not wasted on inappropriate initiatives - a major gain. Decentralization can also help to resolve collective action problems in the management of common property resources by incorporating into the formal political process local knowledge and informal local arrangements for coping with these problems - although this sometimes excludes pastoralists, to their disadvantage (Bardhan, 1997, Leach and Mearns, 1996). But it can also mean that groups which suffer discrimination as a result of local biases (hierarchical mentalities, or prejudice against minorities) are denied benefits.
Mitigating the damaging effects of "hegemonic Western modernity". This theme is often stressed both by post-modernists and by analysts who are concerned about the damage which intrusions from on high can do to the way of life of grass-roots communities. (Those two groups overlap somewhat). Many development specialists prefer to ignore or to dismiss their views. But some of these writers are acutely perceptive and deserve to be taken seriously, especially those who do not idealize local communities and exaggerate the harmony therein (Kothari, 1988 and 1989; and Nandy, 1988.).
These writers tend to favor decentralization as a means of providing people at the local level with at least some means of defending themselves from unwelcome aspects of development fuelled either by the centralized state or by potent market forces. Both of these are seen as aggregations of power that can do severe damage by treating cherished local beliefs, moral codes, informal processes for managing local affairs, and valuable local diversities - many of which are far from antidevelopment - dismissively and aggressively.
Decentralization provides only limited resources to local communities in this struggle. But insofar as it renders government institutions more responsive, and enables local-level interests to undertake small-scale projects of their own choosing - as it commonly does - it has considerable value. There is no reason why we should exclude these concerns from our definition of rural development.
This is not to say that decentralized institutions find it easy to mesh with voluntary associations at and near the local level. Their relations are problematic even in the best of circumstances (Crook and Manor, 1994). But there are modest benefits to be had which should not be ignored in this discussion.
Those in central governments who fear that their own programs and perspectives will be overwhelmed by local prejudices as a result of decentralization are worrying needlessly. The evidence clearly indicates that they retain much of their former potency, and that the real danger is that local preferences will be given too few opportunities to influence decisions. There are also ways to ensure that central programs and policies remain highly influential, while drawing local interests into creative partnerships.
Assisting ethnic or religious minorities. This item in closely tied to preceding section. When ethnic or religious minorities are concentrated in particular subregions or localities, then the devolution of power onto arenas which are roughly congruent with them can enable such groups to gain greater control of their destiny, and to defend themselves against unwanted intrusions from above (Dukesbury, 1991). This can ease their alienation from the state and the wider society, and reduce the danger of damaging conflict.
But the precise details of political geography are crucial here. Such minorities often live cheek by jowl with majority groups, and when that is true, decentralization may actually empower arenas in which prejudices against minorities are stronger than at higher levels in the political system. It may make things worse. Architects of decentralization need to consider this possibility when framing their plans.
Assisting women. The limited evidence available on the impact of decentralization on women's interests offers only modest encouragement. It appears that the empowerment of arenas at or near the local level, where prejudices against women are often stronger than at higher levels, may damage their prospects unless provisions are made to give women a meaningful voice. In some systems, some seats on councils are reserved for women nominees. This holds little promise for them, because they tend to be beholden to the male leaders who secured their nomination. Their best hope lies in the reservation of seats for which only women candidates can stand for election, but this appears to have been attempted in only a few places - notably India.
Even there, it has so far yielded only minimal benefits. During the five-year life span of one such system there, in which 25 percent of seats were so reserved, women members of intermediate-level councils combined to act in women's interests only very exceptionally - less often than ex-untouchable members did. (Some women councillors managed to break down suspicions of village women about health and child welfare programs, however, so that participation in them improved in certain areas.) In local-level councils there, no evidence whatsoever of such action was uncovered by an extensive search for it. By the end of the five years, there were indications that more formidable, educated women would come forward at the next election at the intermediate level (Crook and Manor, 1994, chapter two). But gains for women remain more a hope than a reality.
Facilitating scaling up from successful pilot projects. If projects are shaped very substantially by a desire to suit local particularities - as frequently occurs within elected authorities near the local level - this can create some difficulties in scaling up, in making successful pilot projects more widely applicable. But in most countries, variations in conditions from one locality to another are not so radical as to render successful pilot projects nonreplicable. Far more tends to be gained as a result of consultations with and support from local interests than is lost in replicability.
Combating the tendency to sacrifice local needs to the administrative convenience of generalizers. To reiterate the obvious: the empowerment of elected bodies at or near the local level tends strongly to give interests at the grass roots considerable influence over the shape of development projects. They are unlikely to permit the imposition of initiatives crafted by generalizers at higher levels which fly in the face of local particularities and felt needs.
Politicians and bureaucrats at higher levels should be encouraged to recognize that this presents them not just with a problem, but also with an opportunity. If they develop dialogues with decentralized bodies on macrolevel programs, they will find that partnerships with these authorities can facilitate public understanding of, participation in, and support for those programs - most impressively in mass innoculation and other initiatives in preventive medicine, public health and sanitation, but not only these.
Such partnerships serve the interests of everyone, including high-level power holders. This offers some compensation for the erosion both of their formerly complete control over such programs and of their commandist powers more generally, although it is exceedingly difficult to persuade them of this. Tackling the problem of complexity and coordination. This is an area in which decentralization offers promise and poses potentially serious difficulties. It is promising in two rather different ways. First, as Binswanger (1994) has noted, coordination issues at lower levels in political systems are usually less complex than at higher levels, and institutions at lower levels often have the incentives, the ability and - especially - the information to achieve positive results.
Second, when authorities at lower levels possess wide-ranging powers over rural development programs, it is easier for them to bring the employees of several line ministries together for consultation and concerted action than is possible at higher levels. For example, in the late 1980s in both Bangladesh and the Indian state of Karnataka, elected leaders found it possible to draw together, frequently, officials from a broad range of government departments to discuss development projects. As a result, when a proposal to develop a small scale irrigation scheme was considered, advice was regularly available not just from irrigation specialists, but from engineers and officials dealing with agriculture, fisheries and other activities affected by and affecting the scheme. They discussed the project not just with one another but with representatives of the local communities that would be affected by it, and the resulting synergy yielded significant benefits. It would have been impossible to achieve this if the consultation had occurred in the capital cities of Bangladesh or Karnataka, where line ministries interact far less often (Crook and Manor, 1994).
On the other hand, Parker (1995, pp. 16-17) is not mistaken when he describes coordination as "the Achilles heel of rural development." The problem is that key leaders at lower levels sometimes lack the sophistication or the authority, or both, to succeed at coordination.
Even though, as Parker says, the small scale projects which arise in rural development programs are "dramatically simpler" than huge projects like major irrigation systems, the people who might coordinate such projects (elected representatives and bureaucrats) may lack the training, confidence and social skills needed to succeed.
Even if they possess these things, they may not have the power to insist that officials from various line ministries shed their time-honoured reluctance to collaborate. Or, as Parker notes, implementation of such projects may be delegated to "government bureaucracies or parastatals that were typically highly centralized...(and) out of touch with beneficiaries" (Parker, 1995). Elected office holders or presiding generalist bureaucrats may not have the authority to compel specialists in line ministries or parastatals to work cooperatively with specialists in other fields (Gesellschaft fur Technische Zusammenarbeit, 1993).
So decentralization poses both opportunities and dangers for greater coordination. It also needs to be recognized that decentralized authorities - especially those at or near the grass roots - tend to pursue a narrower range of projects than integrated rural development programs have advocated (although those programs may have been unrealistically broad).
Maintaining complex, integrated rural development programs. As noted just above, it is unrealistic to expect decentralized authorities to tackle the full array of projects which integrated rural development programs have advocated. Many such projects will be undertaken, but the bias will be towards microlevel building projects, sometimes at the expense of service delivery. In India and Bangladesh, this problem was minimized when higher level authorities ear-marked proportions of funds devolved onto local councils for various types of undertakings, with elected councils being given limited latitude for movement of funds from one budget heading to another. That did not solve the problem, but it may offer the best middle way in the drive to achieve integrated development.
Promoting cooperation between nongovernmental organizations and government. The limited evidence available indicates that democratic decentralization can make only a very limited contribution on this front, although it is possible that over time, more significant gains may become possible. International nongovernmental organizations which become frustrated with the inadequacies of central governments have sometimes found it more productive to work through decentralized authorities. But the more important issue here is the relationship between such authorities and indigenous nongovernmental organizations which are active at intermediate or local levels, or both.
We have seen that democratic decentralization can draw leaders of organized interests at the grass roots into creative roles on and relationships with decentralized bodies. But this should not be taken to imply that great progress has been made in breaking down the strong suspicions which many indigenous nongovernmental organizations harbor towards all government institutions. The limited evidence that is available even from relatively successful experiments with decentralization indicates that nongovernmental organizations seldom develop more than a tentative, cautious, arms-length relationship with these bodies, and that they are nearly as distrustful of elected leaders in such institutions as they are of power holders at higher levels. This problem my diminish over time as decentralization breaks down suspicions of government.
This writer's interviews in South Asia both with elected members of decentralized authorities and with high-level enthusiasts for decentralization have uncovered significant suspicions in both groups of indigenous nongovernmental organizations. nongovernmental organizations are often seen as unelected, insufficiently accountable and somewhat unrepresentative. Government is seen as the main and largely sufficient engine for rural development [34]. Breaking down the suspicions on both sides of this divide will take time and may never occur in many places. We lack and badly need evidence on the workings of those few systems which give indigenous nongovernmental organizations representation within state-sponsored bodies - often of a consultative type.
Reducing absenteeism among government employees. In a very small number of cases - mainly the Indian state of Karnataka between 1987 and 1991 (Crook and Manor, 1994) and to a limited extent, the Philippines [35] - the empowerment of elected councils near the local level has enabled popular representatives to put pressure on government employees in schools and local health centers to come to work in accordance with their contracts, and to work assiduously while they are on the job. The existence of a lively civil society in these places was also crucial since it meant that citizens brought poor performance to the attention of elected representatives. This ensured that services were delivered more effectively, at no additional cost to the exchequer.
In most places, however, this needs to be seen as a possibility (once democratic decentralization has had time to establish itself) rather than as a likely gain. But if decentralized institutions are given time to take root, and if organized interests grow accustomed to engaging with them, there is some hope of headway on this front.
Reducing the overall amount of corruption. It should be stressed that democratic decentralization is always attended by an increase in the number of persons who are involved in corrupt acts. This is inevitable because it increases (usually dramatically) the number of people with at least minimal access to political power. However, this need not imply that the overall amount of money diverted by corrupt means increases. Decentralization can, in some circumstances, cause it to decline.
We have seen signs of this in a very small number of cases, especially in the Indian state of Karnataka. It also appears to have occurred in the Indian state of West Bengal - although this owes much to the presence of a penetrative and reasonably disciplined Communist Party, a factor which is seldom on hand elsewhere. It may have occurred in certain pockets in the Philippines [36].
The crucial question here - to which we have no satisfactory answer - is: what caused this decline? Is it explained by factors external to the decentralized system, or by the system itself? Both clearly played a role, but which was crucial? If unusual externalities largely explain it, then it may be difficult to replicate this effect in other decentralized systems.
In Karnataka, the existence of a lively civil society, a large number of experienced if small-time political activists at low levels in the system, an established two-party system, a free press - external factors - helped to reduce corruption. These things are seldom found in abundance in other developing countries. But democratic decentralization provided the mechanisms through which they could produce this effect.
The available evidence strongly suggests that this is exceedingly difficult to achieve. We know of no examples other than the three mentioned above. But if strong, democratic, decentralized institutions are allowed to function for extended periods (a decade or more) - something on which we have next to no evidence - corruption might well be reduced. This might occur through a combination of processes: with democratic institutions stimulating (as they do) the development of civil society, while organized interests, the press, and others acquire skills and the inclination to make the system work well. This, like the previous item, is more a possibility than a likelihood, but it is important enough to bear watching. (Enhanced transparency in Karnataka led citizens to believe that corruption had increased, although the opposite was true.)
Giving greater attention to the wider context of macroeconomic policy. The impact of decentralization on popular views of the macroeconomic policy context is ambiguous, but on balance it is somewhat positive. Part of the time, it distracts rural dwellers from this context, but in many ways, it enhances popular awareness and understanding of it.
When power and resources are devolved onto bodies at lower levels, rural dwellers naturally focus their attention on those levels, since decisions that are made there affect their lives. This does not mean, however, that huge numbers of people shift their focus from central to lower levels. That sometimes happens, but most people had never focused on the national level (or any other sphere of government) in the first place. So in most cases, people who focus on lower-level authorities are being drawn anew into political awareness, which is no bad thing. Elected representatives at lower levels usually engage bureaucrats from line ministries in disputes over the direction of policies and spending, and over the limits on spending which central governments tend to impose. In these skirmishes, lower-level policy preferences tend to make at least some headway against national-level level priorities. That means that in some ways, less attention than before is given to the macroeconomic context. But since pressure from lower level authorities to permit enhanced spending usually fails, such episodes simultaneously acquaint elected members of lower-level authorities (and their constituents) with macroeconomic strategies.
Decentralization tends strongly to produce more transparent government which acquaints many people at lower levels with the amounts of money that are available for development, with policies higher up, and sometimes with the ways in which those policies manifest themselves in central authorities' earmarking of sums for education, health, agriculture, and the like.
Decentralization also inspires a more realistic popular understanding of what can and cannot be achieved by government - by acquainting people with, among others, spending constraints. It can also erode popular cynicism about the state, and develop a greater sense of partnership with government generally. These things tend to promote at least some tolerance for the macroeconomic policies of central governments.
Higher-level authorities need to find ways of maintaining the integrity of macrolevel policies. But devices like ear-marking usually suffice. Indeed, the main danger is usually not that those policies will be ignored and undermined, but that such devices will place excessive limits on the autonomy of lower-level authorities. These comments lend credence to Shah's arguments that decentralization poses little threat to macroeconomic restraint (Shah, 1996; Shah, 1997). And they should reassure those who write about the "dangers of decentralization" in this sphere (Prud'homme, 1995 and Tanzi, 1996).
Counteracting urban bias. Democratic decentralization tends quite strongly to counteract urban bias because it often provides institutional channels through which representatives of rural areas can exert considerable influence on higher levels and receive substantial resources (financial and administrative) from higher up. This is not invariably the case, however.
If it is to have that effect, the system must be structured in a way that provides rural areas with equitable or preferential representation in comparison to urban areas. This is obviously not a difficulty when a program of decentralization only empowers authorities serving rural parts, as often happens. Problems arise when individual authorities extend over both urban and rural areas.
Urban dwellers tend to be better organized and more aware and skilled politically than rural folk. This and their physical proximity to the offices of local authorities almost always give them unfair advantages in such situations. The problems are even more severe:
Despite these problems, however, the available evidence suggests that decentralization usually helps to counteract urban bias. But this is not to say that it impinges on every aspect of urban bias. For reasons set out below, it has little impact on biased agricultural taxation which is an important element of urban bias. Such taxation policies are usually decided high up in political systems, well beyond the influence of decentralized authorities.
Alleviating poverty which mainly arises from disparities between (as opposed to within) regions or localities. This is closely related to urban bias, but it is not the same thing. Different regions within many countries, and different localities within regions, tend to vary in terms of wealth, levels of development, and access to goods and services provided by governments and market forces.
In a study which has received considerable attention, Remy Prud'homme has argued - mistakenly - that decentralization cannot help to reduce disparities between regions or localities. There are unfortunate echoes of this in the World Development Report 1997 (World Bank, 1997). The problem with his analysis arises from his definition of decentralization. He defines it as a system in which local governments raise all of the revenues that they spend through locally imposed taxes "without the benefit of central government transfers" (Prud'homme, 1995, pp. 201-203). He is correct in saying that in such systems, disparities between localities will not be reduced, since resource-poor localities will find it difficult to raise as much revenue as richer localities. But nearly all decentralized systems entail at least some transfers from higher levels down to decentralized bodies - as they should if there is to be any hope of decentralization working well. When that happens, higher levels of government can build in elements which provide under-resourced and previously under-represented localities or regions with more resources than richer arenas, and which (crucially) provide representatives of such areas with the political leverage to ensure that they can acquire a fair share of resources. This is actually quite a common practice.
It is therefore more accurate to put this another way. Decentralization often facilitates redistributive policies to ease these disparities, but it is not invariably a positive force. A system such as Prud'homme posits will not ease such disparities. And an exercise in fiscal decentralization which undermines a pre-existing system of redistributive transfers of resources will make things worse not better. But if redistributive mechanisms for the allotment of resources from higher levels are included, and if representatives of poorer arenas are empowered to ensure their implementation, this problem will not arise.
The key here is the empowerment of these people. Much of the literature on decentralization focuses on financial formulas or targeting devices to ensure a redistribution of resources from richer to poorer areas (for example, Shah and Qureshi, 1994, pp. xvi-xvii, xxi) [38]. This is useful, but such formulas and devices are not the only or the most reliable means of achieving this.
Decentralization which is substantially democratic, and which entails some transfers of resources downward from higher levels, generates a political logic and informal practices that are usually more effective at ensuring a measure of justice for deprived localities or regions than are formal financial or administrative arrangements - provided that such empowerment occurs. If this happens, then the creation of decentralized, democratic institutions in every area of a country tends strongly to ensure that remote or impoverished areas, or both, which had previously been poorly represented at the highest levels in the political system, gain a greater voice at those levels.
Representatives of those areas articulate their needs and press for more equitable treatment from those at the apex of the system, and if they possess substantial political clout, they will almost certainly obtain helpful responses, Evidence from Colombia, Bangladesh and some Indian states (World Bank, 1995; Crook and Manor, 1994; Webster, 1992) indicates that these representatives succeed more effectively than was possible before decentralization. The empowerment of such representatives is necessary to counteract the (often unconscious) bias towards urban and prosperous arenas of politicians at higher levels and of bureaucrats at all levels. The empowerment of rural representatives and formal financial and administrative arrangements to promote redistribution should be seen not as alternatives, but as complementary.
Why? In many political systems, parochial and elite social forces tend to have more influence at the local level - and to exercise it there in a more uncompromising, exploitative manner - than at higher levels. Even where this is not true because higher-level elites have little interest in poverty alleviation, the best that can usually be hoped for is that local elites will have the same views of the poor as elites higher up. This writer has yet to discover evidence of any case where local elites were more benevolent than those at higher levels.
In other words, democratic decentralization empowers arenas which tend to be dominated by groups less, not more, amenable towards redistribution than those who dominate higher levels (Crook and Manor, 1994; Moorehead, 1991; Lund, 1993) [39]. It is therefore unrealistic to expect it to contribute much to poverty alleviation where the main problem is inequity within, not between regions or localities. This does not mean that democratic decentralization is to be avoided. It has many virtues. But when it is undertaken, efforts should be made to protect poverty alleviation programs from it - by vesting control of such programs in persons at higher levels, providing of course that such persons are more enthusiastic about redistribution.
In fairness, we need to pay attention to some writings (for example, World Bank, 1995, p. 4) which call the negative view above into question. Some countervailing analyses - anchored in empirical study and not mere hope - are worth noting. The most telling of them emerge from Latin America. As anti-poverty programs there have become more "demand-driven," and as democratic decentralization opens up channels which might be used by poorer groups to register their demands, real prospects for gains in poverty alleviation sometimes arise.
We must stress, however, that two crucial prerequisites for this which exist in many Latin American countries are found much less often in Africa, Asia or Eastern Europe. These are: considerable organizational strength among poorer groups at the local level, and a willingness of those groups to engage pragmatically with government institutions (Fox, 1994; Fisher, 1993) [40]. If more such evidence emerges, or if poorer groups elsewhere develop greater organizational strength, we may wish to move this item into the second section above.
The second of these two ifs is a realistic possibility over the medium-to-long term, thanks to new opportunities provided by decentralization. We noted that decentralization catalyzes greater participation and associational activity among all sections of society. It enables rural dwellers to develop their political awareness and to learn lobbying technology and other political skills that can advance their interests. Local-level elites seize these opportunities quite quickly, and in the short run, this tends to enhance their already substantial advantage over the poor. But if poorer groups follow suit, they may eventually make significant gains too. We cannot say with any certainty that or how soon this will happen, but it is clearly a distinct possibility.
Assisting pastoralist groups. Sometimes, the poor within a locality consist in part of pastoralists. There is a strong likelihood that decentralization will fail to benefit such people - indeed, it may well make their problems worse. The main problem is that pastoralists are mobile groups. They make their living by animal herding which requires them to move often from place to place, so that they reside in a locality only part of the time. Their prolonged absences may make it impossible for their voices to be adequately heard in decentralized democratic processes - at and between elections.
If this happens, there is a strong possibility that groups of farmers who are permanently settled on the land may take advantage of the situation to damage pastoralists' interests. They may, for example, take control of lands which the pastoralists have traditionally used for grazing - to erode grazing rights which often tend in any case to be inadequately protected by national laws (Lane and Moorehead, 1994; Evers, 1994, p. 10).
In places where pastoralists form a significant part of the rural population, the architects of decentralization need to bear this problem in mind, and seek to build safeguards into the system. But even when they do so, actual practice on the ground may damage pastoralists' interests.
Easing the problem of excessive agricultural taxation. This section should be read as a qualification of the discussion of reducing urban bias above. Trade and pricing policies in developing countries frequently impede agricultural growth and rural development by protecting industry, setting exchange rates too high, and others (Schiff and Valdés, 1992), often to the disadvantage of poorer rural groups (Bates, 1981 and Bates, 1983).
Decentralization is unlikely to make much positive impact on this front. For less prosperous farmers, it might reduce somewhat both the cost of organization and communication, and the advantage which prosperous rural groups derive from their educational and informational status (Olson, 1971). The devolution of power onto authorities serving mainly rural arenas might enhance somewhat the capacity of rural interests to articulate their views in the wider political arena (Becker, 1983 and 1985). But the recent record suggests that such bodies will have little effect on policymakers who decide these things far away at the national-level.
Reducing overall government expenditure. High-level architects of decentralization often see it as a means of reducing the overall level of government expenditure, in part because they expect it to enhance local resource mobilization. On both counts, they are usually mistaken. This misperception and the linkage between decentralization and structural adjustment initiatives (which is often made) pose a serious threat to the viability of programs of decentralization.
It needs to be understood that the inadequacies of centralized structures and commandist policies which incline many high-level leaders to undertake decentralization also produce widespread cynicism at the grass roots about government institutions and initiatives. If (as is often the case) people have also seen earlier experiments with decentralization founder, their cynicism extends more forcefully to fresh attempts to decentralize.
This means that decentralizers need to make efforts to break down that cynicism in order to give new decentralized authorities some hope of succeeding. To accomplish this, the new bodies usually require substantial injections of funds from higher levels in the early years of their existence. And given their reluctance to impose fresh taxes on an already cynical populace, they are likely to need significant resources from higher up over the longer term. This, together with other start-up costs - the need of decentralized authorities to build offices, possibly to pay newly hired bureaucrats, and to provide elected representatives with at least modest emoluments - implies that decentralization is likely to require at least a modest increase in government expenditure, not the reverse.
In several cases, high-level decentralizers have not recognized this. They have naively expected decentralized bodies to mobilize substantial resources from a heartily cynical electorate. The result - as for example, in Ghana - has been to cripple these bodies from the outset, and to scuttle any hope of the significant gains which decentralization can produce in better circumstances (Crook and Manor, 1994).
We should also pay attention to evidence from places like Nigeria which indicates that expenditure on decentralized system tends to creep upwards over time (Gboyega, 1993). If powerful figures at higher levels in political systems are serious about decentralization - as they should be, since it can serve their interests - they need to grasp that it is dangerous to see it as a means of cost-cutting. This is true not only early in the lifetime of decentralized bodies, but over the longer-term.
Note, however, that decentralization may eventually ease this problem somewhat. If rural dwellers develop a more sophisticated appreciation of the fiscal constraints which macroeconomic policymakers face, their representatives in decentralized authorities may become a little more willing to curtail their ambitions accordingly. But, since these people (like politicians everywhere) tend to fight their corner aggressively, it would be surprising if much headway were made on this front.
Mobilizing local resources. This is an important issue, worth discussing at length, since much of the recent interest in decentralization is based on the assumption that it can facilitate local resource mobilization. We need to distinguish between types of resource mobilization. It can mean the imposition and collection of taxes or both. Or it can mean the mobilization of resources other than taxes - investments in cash or in kind by people at the local level, in (for example) demand-driven rural investment funds.
Let us consider this second topic first. World Bank analysts working in Brazil and Morocco have found that carefully designed cofinancing mechanisms can succeed in inspiring significant investments of this sort at the grass roots [41]. Those successes need to be built upon, since they appear to offer greater promise than resource mobilization through taxes. What follows is a discussion of that bigger, more trouble-prone topic.
If successful experiments with decentralization are allowed to survive unmolested by central governments for extended periods - a big if since several have been extinguished - it is possible that local governments will become better able to mobilize local resources through taxation. But for the present, the great weight of the evidence indicates that they have had serious difficulty doing so in the brief period since decentralization occurred.
In some parts of the world - certainly in much of Latin America - it is appropriate to speak of "the enormous untapped fiscal potential of local government" when referring to local authorities in urban areas. But this study is concerned with rural development, and we must be careful about extending comments such as the following beyond the urban sector:
Effective yields on most local taxes, especially property taxation, had fallen to derisory levels over the previous decades because of the failure of cadastral surveys to keep abreast with rapid urban growth, the failure to adjust tax rates in line with high rates of inflation, and widespread administrative inefficiency and corruption facilitated by the system of self assessment for local taxation (Nickson, 1995, p. 12).Let us turn to the rural sector. Some scholars argue that it is unrealistic to expect decentralized authorities in rural areas to mobilize local resources because too few resources actually exist (Therkildsen, 1994). But counter-arguments, based on careful research in poor areas of Bangladesh - a very poor country - indicate that this is not the main problem (Blair, 1989). There are, however, seven other impediments to the early mobilization of local resources by decentralized authorities in rural parts.
It is important, however, to stress that over the medium or longer term, there is some likelihood that this will change - to some degree. We know that when decentralization works well, it improves government responsiveness, draws society into creative partnerships with the state's decentralized institutions, makes many different types of policies more sustainable, and erodes the suspicion and even cynicism which ordinary people often feel towards government.
Insofar as such things occur, mobilizing additional resources to meet the costs of public goods should become more likely. This can occur in three ways. As suspicion of government diminishes, people should become more willing to pay taxes. Conditions should also develop that encourage greater investment in profit-making enterprises which holds some promise for eventually enhancing local revenues (World Bank, 1997). And investments in local development funds - perhaps inspired by cofinancing initiatives from higher levels - should become more feasible. We already have evidence from the Philippines of revenues from taxes increasing because lower-level authorities adapted centrally-devised regulations to local particularities (de los Reyes and Jopillo, 1995, p. 88). We cannot say how much more of this may become possible, or how quickly it will occur. But it would be surprising if we saw no gains on this front, eventually.
Performing tasks off-loaded by central government. Governments often decentralize partly because they see it as a way of off-loading tasks, which they lack the resources or the inclination to perform, onto lower-level bodies. In theory, this sometimes makes sense, since (as we have seen) such bodies are often able to adapt high policy to distinctive local conditions and to break down citizens' suspicions of high-level initiatives. But there are also serious dangers here.
The main problem is the tendency of central governments to transfer tasks downward without transferring adequate - or sometimes any - resources to perform them. This can result from naiveté or cynicism - two rather different motivations which often produce similar outcomes. Given the difficulties in mobilizing local resources, such actions can either cripple decentralized authorities financially or (more often) results in tasks going unfulfilled. Citizens who see services dry up just when new decentralized institutions are created are unlikely to respond enthusiastically to the change. The lesson is clear: higher level government must devolve sufficient resources to allow lower authorities to carry out the tasks for which they are made responsible.
Another, less acute problem can also cause trouble: bureaucratic incapacity at intermediate and local levels. This can be minimized if higher level authorities are prepared to lend advice and assistance to decentralized bodies. But what is mainly needed here is a realistic assessment by decentralizers of the administrative capabilities of lower-level bodies, before off-loading takes place.
Promoting planning from below. High-level technocrats and politicians often expect decentralized authorities to engage in planning from below. Some analysts, who fix mainly on formalities and take seriously official documents that appear to show that the formal process is actually working, claim that this has been achieved. However, this writer's acquaintance with the informalities that lie behind these appearances raises grave doubts about whether this has occurred or can occur in most developing countries - for several reasons.
First, there is the problem of lack of administrative capacity. Elected bodies at or near the local level usually have extremely small staffs - sometimes no more than a single clerk with limited education and experience. Such employees can offer elected politicians little help with the complex process of developing a plan, or with the complicated paper work needed to express their intentions. They also often suffer from an inadequacy of information needed to construct a plan - no taxpayers roll, no inventory of capital resources, no assessment of the local economy, and the like.
Even where administrative resources are more plentiful - usually at intermediate levels - the process often fails to work, despite appearances to the contrary. Consider what happened in 1988 in one of the most developed districts of the Indian state of Karnataka - where the district council was served by a brilliant chief administrator with a substantial staff, where information on land holdings, and taxpayers and much else had been accurately and comprehensively computerized. On the night before the district plan was due to be submitted to higher authorities, the administrator in question telephoned a friend - a professor of economics at the local university - with a desperate plea for help in concocting a plan out of woefully incomplete and chaotically organized information that had come in from around the district. His staff were no help, because they had been trained to implement policy, and not to plan or to gather the information needed to construct a plan - a major problem in many countries. During an all-night session the two produced a plan which appeared well-considered, but the reality was very different [43]. It would be naive to expect better results in most other developing countries, where conditions are less congenial.
The problems do not end there. The elected leaders of decentralized authorities tend to understand "planning" as little more than the addition of wish lists to the annual budgeting process (see for example, de los Reyes and Jopillo, 1995, p. 88). Politicians like to keep their lists of projects open, so that items can be dropped if they prove difficult to implement or fall out of favor with constituents, and new ones can be added to please new friends. To ask them to etch their priorities in stone a year or more in advance causes them inconvenience on both fronts. Their usual response is to disregard whatever plan they have knocked together to placate higher authorities.
A further severe problem is that higher authorities who extol the virtues of planning from below often have little interest in it in practice. Decentralized authorities which submit their plans often find that they are ignored or greatly altered by higher ups. Grants from above remain heavily earmarked in ways that are inconsistent with their wishes. The great weight of evidence now available inspires grave doubts that decentralized planning can be made to work very well.
It should be stressed, however, that planning from below may improve over time. Elected representatives and interest groups may become more aware of the logic of the system, and more skilled at making it work. Civil servants who are unused to activities which support planning may become more adept at them. There is some preliminary evidence from the Philippines and Bolivia to suggest that participatory rural appraisal techniques can facilitate the planning process [44].
But even if improvements do not materialize, inadequate or nonexistent planning from below should not be seen as a matter of serious concern. Decentralized systems have clearly worked well - in for example, India (Crook and Manor, 1994) - without effective planning from below.
Promoting community participation in development. The phrase "community participation in development" has a very specific meaning. It does not refer to the many types of participation mentioned above - voting, campaigning, lobbying, contacting power holders, attending meetings, and the like. It refers to occasions when all or most of the people in a rural arena come together in a cooperative spirit and collaborate voluntarily to construct or create something. It refers to community-wide participation.
We have seen that decentralization facilitates increases in voting, lobbying, contacting, and so on. But it does not facilitate community participation in development, despite the expectations of many enthusiasts for such efforts, for several reasons.
First, many of the authorities created through decentralization are supralocal in character - they stand above (often far above) the local level. Some of those which are located at the local level cover areas that embrace numerous villages and even towns. Such bodies find it difficult to mobilize most of the population of a single village for development purposes.
Even when decentralized bodies are congruent with single villages, they face problems. People are often heartily cynical about any government initiative. Free riders are reluctant to lend their efforts. And crucially, electoral competition for places on decentralized bodies creates new divisions within villages and intensifies old ones - that is, it undermines community solidarity.
Decentralized authorities are representative institutions to which an exclusive few are elected. There is clear dissonance between the logic by which they operate and the communitarian logic which might inspire broadly inclusive participation in development efforts. Elections are usually won by village elites who, not surprisingly, frequently behave in an elitist manner thereafter.
Village societies are often inequitable, so that community spirit is limited and collective efforts tend to be seen (often rightly) as just another device for exploiting the disadvantaged. So genuine community participation in development is often an impossible, naive dream. But insofar as it is possible, it is best fostered by local voluntary associations or nongovernmental organizations, not by decentralized authorities. This problem should not be expected to wane over time.
2. Research on nongovernmental organizations in the Gambia (the country studied in the paper) indicates that their effectiveness is hindered by resource dependence and organizational weakness. They are seriously limited in their ability to complement government development programs - much less replace them. There is also a tendency for nongovernmental organizations there to duplicate efforts in certain fields and to omit others from their activities. Coordination of their efforts has been sadly lacking. (I am grateful to Ann Hudock of the Institute of Development Studies, Sussex for this information.) So there is reason to doubt whether "decentralization by default" offers a helpful alternative to conventional government efforts.
3. Bennett, 1990 and Bennett, 1994 ably analyze this.
4. There are good arguments for staying with Rondinelli's original formulations. I have chosen Parker's here mainly in the interests of consistency with other studies that parallel this one.
5. These authorities do not always have to be created anew. For example, in Latin America after May 1994, municipal mayors who had previously been appointed (and who had thus been agents of the central authorities) were "freely elected in every country except Cuba and Haiti" (Nickson, 1995, p. 2). This might be regarded as democratization rather than democratic decentralization. But since the new mayors were no longer agents of national governments, this change represented a de facto transfer of power from the national governments to the local level - it entailed both democratization and decentralization.
6. I am grateful to Suzanne Piriou-Sall for putting the point to me in these terms.
7. Such tripartite systems often develop when fiscal and democratic decentralizations occur to levels at which administrative decentralization has already taken place.
8. Indeed, we can go further and say that it also ought to entail "the elimination of exclusionary political practices, including fraud, unfair limits on voter registration,...the lack of ballot secrecy, voter intimidation, and vote buying" (Fox, 1994, p. 106). But this discussion is less about what ought to be than what is admissible under the label of democratic decentralization. Very limited infringements of the standards set out in the quotation above - for example, in vote buying, a widespread practice - should not cause us to exclude a system from consideration under this label.
9. This offers a contrast to the narrower definition adopted in some other studies - for example, Blair (1995).
10. See for example, the communes of Cote d'Ivoire assessed in Crook and Manor, 1994.
11. This failure is very common among people who study China - not least because the Chinese themselves make a habit of using the word local to refer to provincial and other intermediate levels - but it also arises in studies of some other countries. See for example, Fabian and Straussman (1994) on Hungary.
12. Indeed, there is a discernible shift of emphasis by some development agencies such as United States Agency for International Development towards decentralization in urban areas (Blair, 1995).
13. See, on Latin America, Nickson (1995, p. 3). I am grateful to Jean-Francois Bayart for calling my attention to similar tendencies in the literature on Francophone Africa.
14. Since we are discussing democratic decentralization, popular participation - which might also be seen as crucial - is assumed to be present. Participation alone does not guarantee the success of experiments with decentralization (Crook and Manor, 1994).
15. I am grateful to Jorge Dominguez for this insight and form of words.
16. Interview with B. Rachiah, Bangalore, January 11, 1988.
17. Times of India, March 15, 1996.
18. See for example World Bank (1995, p. ix) and much of the literature on decentralization in Bangladesh which is dominated by public administration specialists.
19. I am grateful to T. Mathew of the Indian Administrative Service for information on this.
20. See the studies of Estonia, Poland, Hungary and the Czech Republic and the conclusion of Gibson and Hanson (1996) and European Bank for Reconstruction and Development (EBRD) (1994).
21. Discussion with Harry Blair, London, April 17, 1997.
22. I am grateful to V.K. Natraj for calling this to my attention. There are suggestions of this in de los Reyes and Jopillo, 1995.
23. Interview with a civil servant who was present on that occasion, Brighton, March 17, 1997.
24. John Gaventa has evidence of such trends even within the United States.
25. The words here are not Hirschman's but are from a summary of his studies in Haggard and Kaufman, 1992.
26.See for example, Robinson (1988).
27. This was true in Bangladesh under H.M. Ershad's military regime between 1985 and 1991.
28. The utility of this for democracy, civil society and policy makers is vividly apparent from Jenkins, (1997) on the Indian state of Maharashtra.
29. I am grateful to Bangladeshi colleagues at a workshop on decentralization at Rajendrapur, March 10-11, 1997, for calling this to my attention.
30. For example, John Gaventa found clear evidence of this in the Philippines in 1997.
31. For similar evidence from a wider array of countries, see Shah, 1997, especially pages 15-17.)
32. The comments in this section received strong corroboration from a set of as yet unpublished U.S. Agency for International Development investigations during 1996-97 in a diversity of countries in Asia, Africa and Latin America where democratic decentralization has been undertaken. Discussion with Harry Blair, London, April 17, 1997.
33. This emerges from U.S. Agency for International Development studies of democratic decentralization in Asia, Africa and Latin America - discussion with Harry Blair, London, April 17, 1997.
34. This emerged, for example, from a talk with an enlightened champion of democratic decentralization, Dr. Kamal Siddiqui - then Secretary to the Prime Minister of Bangladesh - in Dhaka, February 2, 1993.
35. I base this on discussions with Elena Panganiban.
36. I am grateful to Neil Webster and Elena Panganiban for information on these two cases.
37.This is true, for example, in parts of Latin America and of Francophone Africa. See the case of Côte d'Ivoire where many of the communes empowered in the mid-1980s were predominantly or substantially urban.
38. See for example, Shah and Qureshi (1994, pp. xvi-xvii).
39. This is corroborated by the recent research of Anand Inbanathan in India, and by the findings of several U.S. Agency for International Development teams working in various countries. I am grateful to Inbanathan and Harry Blair for this information.
40. I am also grateful to Emanuel de Kadt of Institute of Development Studies, Sussex, who has worked with the Chilean government on social policy, for information on this point.
41. Communication from Suzanne Piriou-Sall, December 5, 1996.
42. For example, the British Broadcasting Corporation World Service reported on February 8, 1996 that township residents in South Africa are reluctant to pay taxes even to newly elected local bodies because of the cynicism instilled in them by long years of the apartheid system.
43. This is based on interviews with the two people in question in Karnataka, in April 1993.
44. I am grateful to John Gaventa and James Blackburn for information on this.
Arnstein, S.R. 1969. "A Ladder of Citizen Participation." AIP Journal (July).
Austin, G. 1966. The Indian Constitution. Oxford, UK: Oxford University Press.
Bardhan, P. 1997. "Decentralised Development." The V.K. Ramaswamy Memorial Lecture, Delhi School of Economics, (manuscript).
Barrier, C. 1990. "Developpement Rural en Afrique de l"Ouest Soudano-Sahelienne: Premier Bilan sur l'Approche Gestio des Terroirs Villageois." Les Cahiers de Recherche Developpement no. 25.
---------. 1991. Leçons des Projets de Gestion des Resources Naturelles en Milieu Rural. CCCE, Paris.
Bates, Robert. 1981. Markets and States in Tropical Africa: The Political Bases of Agricultural Policies. Berkeley: University of California Press.
---------. 1983. "Patterns of Market Intervention in Agrarian Africa," Food Policy, 8 (November): 297-304.
Bates, Robert and Da-Hsiang Donald Lien. 1985. "A Note on Taxation, Development, and Representative Government," Politics and Society (1): 53-71.
Becker, Gary. 1983. "A Theory of Competition among Pressure Groups for Political Influence." Quarterly Journal of Economics 98 (3): 371-400.
---------. 1985. "Public Policies, Pressure Groups and Deadweight Costs." Journal of Public Economics 28 (3): 329-47.
Bennett, R.J. (ed.). 1990. Decentralisation, Local Governments and Markets: Towards a Post-Welfare Agenda. Oxford, UK: Clarendon Press.
---------. (ed.). 1994. Local Government and Market Decentralisation: Experiences in Industrial, Developing and Former East Bloc Countries. Tokyo: United Nations University Press.
Binswanger, Hans P. 1994. Agriculture and Rural Development: Painful Lessons. Agriculture and Natural Resources Management. World Bank, Washington, DC.
Binswanger, Hans P., and Klaus Deininger. 1997. "Explaining Agricultural and Agrarian Policies in Developing Countries" Journal of Economic Literature.
Blair, Harry. 1989. Can Rural Development Be Financed from Below? Dhaka.
---------. 1995. Assessing Democratic Decentralization: A Center for Development Information and Evaluation (CDIE) Concept Paper . United States Agency for International Development, Washington, DC.
Brillantes, A.B. 1994. "Redemocratization and Decentralization in the Philippines: The Increasing Leadership Roles of nongovernmental organizations." International Review of Administrative Sciences (December): 575-86.
Brinkley, D. 1996. Washington Goes to War. New York: Ballantine Books.
Carvajal, E.P. 1995. "Decentralization and Democracy: The New Latin American Municipality." La Comisión Económica para América Latina (CEPAL) Review, volume, number] (April): 39-53.
Chai-Anan, S. 1991. "The Three-Dimensional State." In James Manor (ed.) Rethinking Third World Politics. London: Longmans.
Cheema, G.S. and D.A. Rondinelli (eds.). 1983. Decentralization and Development: Policy Implementation in Developing Countries. Beverly Hills, CA: Sage.
Crook, Richard. 1991. "State, Society and Political Institutions in Cote d"Ivoire and Ghana." In James Manor (ed.) Rethinking Third World Politics. London: Longmans.
Crook, Richard and James Manor. 1994. Enhancing Participation and Institutional Performance: Democratic Decentralization in South Asia and West Africa. Overseas Development Administration, London.
Davis, Daniel, David Hulme and Philip Woodhouse. 1994. "Decentralization by Default: Local Governance and the View from the Village in The Gambia." Public Administration and Development 14: 253-69.
Dearlove, J. 1995. "Village Politics." In R. Benewick and P. Wingrove (eds.) China in the 1990s. London: Macmillan.
de Janvry, Alain. 1981. The Agrarian Question and Reformism in Latin America. Baltimore: Johns Hopkins University Press.
Dillinger, William. 1994. Decentralization and its Implications for Urban Service Delivery. Urban Management Progam Discussion Paper 16, World Bank, Washington, DC.
Domenach, J.-M. 1997. Regarder la France - Essai sur le Malaise Francaise. Paris: Perrin.
Dominguez, J. I. 1986. Cuba: Internal and International Affairs. Beverly Hills, CA: Sage.
Drennan, W. 1995. "The Impact of Korean Local Elections." Strategic Forum October.
Dukesbury, J. M. 1991. Decentralization and Democratic Pluralism: The Role of Foreign Aid. United States Agency for International Development/CDIE/POA, Washington, DC.
European Bank for Reconstruction and Development. 1994. Transition Report October. London.
Edwards, Sebastian. 1994. "The Political Economy of Inflation and Stabilization in Developing Countries." Economic Development and Cultural Change, pp. 235-66.
Edwards, Sebastian and Guido Tabellini. 1991. "Explaining Fiscal Policies and Inflation in Developing Countries." Journal of International Money and Finance, 10 (Supplement): S16-48.
Evers, Y. 1994. Local Institutions and Natural Resource Management in the West African Sahel: Policy and Practice of "Gestion de Terroir" in the Republic of Mali. Rural Resources, Rural Livelihoods No. 5, Institute for Development Policy and Management, Manchester, UK.
Fabian, Katalin and Jeffrey Straussman. 1994. "Post-Communist Transition of Local Government in Hungary: Managing Emergency Social Aid." Public Administration and Development, 14: 271-80.
Faure, Y.A. 1989. "Cote d"Ivoire: Analysing the Crisis." In D.B. Cruise O'Brien, J. Dunn and R. Rathbone (eds.), Contemporary West African States. Cambridge, UK: Cambridge University Press.
Fisher, J. 1993. The Road from Rio: Sustainable Development and the Non-Governmental Movement in the Third World. Westport, CT: Praeger.
Fox, J. 1994. "Latin America's Emerging Local Politics." Journal of Democracy (April): 105-15.
Fuhr, Harald. n.d. Decentralized Policies: Expenditure and Revenue Assignment Options - Is there any Best Practice? What can we Recommend? World Bank discussion paper.
Gaventa, J. 1997. Participation, Poverty and Social Exclusion in North and South. Working Paper, Institute of Development Studies, University of Sussex, Brighton.
Gazaryan, A. and M. Jeleniewski. 1996. "Political and Economic Issues in the Re-creation of Lithuanian Local Government." In J. Gibson and P. Hasson (eds.) Transformation from Below: Local Power and the Political Economy of Post-Communist Transitions. Cheltenham, UK: Edward Elgar.
Gboyega, A. 1993. "Local Government Reform in Nigeria." In Mawhood (ed.), Local Government in the Third World: Experience of Decentralization in Tropical Africa, Second edition. Africa Institute of South Africa.
Gibson, J. and P. Hanson. 1996. "Decentralization and Change in Post-Communist Systems." In J. Gibson and P. Hanson (eds.), Transformation from Below: Local Power and the Political Economy of Post-Communist Transitions. Cheltenham: Edward Elgar.
Grindle, M. 1989. The New Political Economy: Positive Economics and Negative Politics. Harvard Institute for International Development Development Discussion Paper 311, Cambridge, MA.
Gesellschaft fur Technische Zusammenarbeit (GTZ). 1993. Regional Rural Development-Regional Rural Development Update: Elements of a Strategy for Implementing the Rural Regional Development Concept in a Changed Operational Context. Eschborn, Germany.
Haggard Stephen and Robert Kaufman (eds.). 1992. The Politics of Economic
Adjustment: International Constraints, Distributive Conflicts, and the State.
Princeton: Princeton Universty Press.
Hargreaves, J.D. 1979. The End of Colonial Rule in Africa. London: Macmillan.
Hendrickson, P. 1996. The Living and the Dead. New York: Alfred A. Knopf.
Hessling, G., and B.M. Ba. 1994. "Land Tenure and Resource Management in the Sahel - Regional Synthesis and Summary." Paper presented at the CILSS, OECD, Club du Sahel Regional Conference on Land Tenure and Decentralisation in the Sahel, January, Praia, Cape Verde.
Hobsbawm, E. 1994. Age of Extremes: The Short Twentieth Century, 1914-1991. London: Abacus.
Hoffman, S. 1997. "Look Back in Anger." New York Review of Books (July 17): 45-50.
Huntington, S.P. 1992. The Third Wave: Democratization in the Late Twentieth Century. Norman and London: University of Oklahoma Press.
Jenkins, R. 1995. "Liberal Democracy and the Political Management of Structural Adjustment in India: Conceptual Tensions in the Good Government Agenda." Institute of Development Studies Bulletin (April): 37-48.
Jenkins, Robert. 1997. "Democratic Adjustment: Explaining the Political Sustainability of Economic Reform in India." University of Sussex, Ph.D. thesis.
Kasfir, N. 1993. "Designs and Dilemmas of African Decentralization." In Mawhood (ed.), Local Government in the Third World: Experience of Decentralization in Tropical Africa, Second edition. Africa Institute of South Africa..
Kochanek, S. 1976. "Mrs. Gandhi's Pyramid: The New Congress." In H.C. Hart (ed.) Indira Gandhi's India: A Poltical System Reappraised. Boulder, CO: Westview.
Kothari, R. 1988. Transformation and Survival: In Search of a Humane World Order. Delhi: Ajanta.
---------. 1989. Rethinking Development: In Search of Humane Alternatives. Delhi: Ajanta.
Lane, C., and R. Moorehead. 1994. Who Should Own the Range? New Thinking on Pastoral Resource Tenure in Drylands Africa. Pastoral Land Tenure Series 3, Drylands Programme, International Institute for Environment and Development, London.
Leach, M. and R. Mearns. 1996. The Lie of the Land: Challenging Received Wisdom on the African Environment. Oxford and Protsmouth, NH: James Currey and Heinemann.
Levi, M. 1981. "The Predatory Theory of Rule." Politics and Society, pp. 431-65.
Lofchie, M. 1994. "The New Political Economy of Africa." In D.E. Apter and C.G. Rosberg (eds.), Political Development and the New Reaism in Sub-Saharan Africa. Charlottesville and London: University Press of Virginia.
Lund, C. 1993. Waiting for the Rural Code: Perspectives on a Land Tenure Reform in Niger. Drylands Programme Issues Paper 44, International Institute for Environment and Development, London.
Manor, James. 1979. "The Failure of Political Integration in Sri Lanka." Journal of Commonwealth and Comparative Politics 17 (1): 21-46.
---------. 1981. "Party Decay and Political Crisis in India." The Washington Quarterly (Summer): 25-40.
---------. 1986. "Region and Nation in China and India." Fairbank Center, Harvard University, Cambridge, MA.
---------. 1989. The Expedient Utopian: Bandaranaike and Ceylon. Cambridge, UK: Cambridge University Press.
---------. 1991 (ed.). Rethinking Third World Politics. London: Longmans.
---------. 1995. Capacity Development and the Ability to Change. Swedish International Development Agency, Stockholm.
---------. 1997. "Explaining the Popularity of Decentralization." World Bank, Rural Development Department, World Bank, Washington, D.C.
Mawhood, P. (ed.). 1993. Local Government in the Third World: Experience of Decentralization in Tropical Africa, Second edition. Africa Institute of South Africa.
Medhi, K. 1995. "The Political Economy of Decentralization in Thailand." In Southeast Asian Affairs. Institute of Southeast Studies, Singapore.
Mehta, A. 1978. Report of the Committee on Panchayati Raj Institutions. Government of India, Ministry of Agriculture and Irrigation, New Delhi.
Moore, M.P. 1997. "Death without Taxes: Aid Dependence, Democracy and State Capacity in the Fourth World." Draft manuscript, Institute of Development Studies, University of Sussex.
Moorehead, R. 1991. "Structural Chaos: Community and State Management of Common Property in Mali." University of Sussex, Ph.D. thesis.
Moris, J. 1972. "Administrative Authority and the Pattern of Effective Agricultural Administration in East Africa." The African Review (June).
Morris, A. and Lowder, S. (eds.). 1992. Decentralization in Latin America. New York: Praeger.
Mulasa, T. 1970. "Central Government and Local Authorities." In G. Hyden and others, Development Administration in Kenya. Nairobi: Oxford University Press.
Muthien, Y.G. and Khosa, M.M. 1995. "The Kingdom, the Volkstaat and the New South Africa: Drawing South Africa's New Regional Boundaries." Journal of Southern African Studies (June): 303-22.
Nandy, A. (ed.). 1988. Science, Hegemony and Violence. Delhi: Oxford University Press.
Narayan, Deepa. 1994. The Contribution of People's Participation: Evidence from 121 Rural Water Supplies. United Nations Development Program-World Bank Water and Sanitation Program, Washington, DC.
Nayar, B.R. 1997. "The State and Economic Performance: Globalisation and Marginalisation in India's Shipping." Journal of Commonwealth and Comparative Politics (March): 20-50.
Nehru, J. 1960. A Bunch of Old Letters. Bombay: Asia Publishing House.
Nellis, J.R. 1983. "Decentralization in North Africa: Problems of Policy Implementation." In Cheema and Rondinelli (eds.), Decentralization and Development: Policy Implementation in Developing Countries. Beverly Hills, CA: Sage.
Ng'ethe, N. 1993. "The Politics of Democratization through Decentralization in Kenya." Conference paper, Institute of Commonwealth Studies, University of London.
Nickson, R.A. 1995. Local Government in Latin America. Boulder and London: Lynne Rienner.
North, Douglass. 1981. Structure and Change in Economic History. New York: W.W. Norton and Company.
---------. 1990. Institutions, Institutional Change and Economic Performance. Cambridge, UK: Cambridge University Press.
O'Donnell, G. 1973. Modernization and Bureaucratic-Authoritarianism: Studies in South American Politics. Berkeley: University of California Press.
Oi, Jean. 1992. "Fiscal Reform and the Economic Foundations of Local State Corporations in China." World Politics 45 (October): 99-126.
Olson, M., 1971. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press.
Oyugi, W.O. 1993. "Local Government in Kenya: A Case of Institutional Decline." In Mawhood (ed.), Local Government in the Third World: Experience of Decentralization in Tropical Africa, Second edition. Africa Institute of South Africa.
Painter, T. 1991. Approaches to Improving the Use of Natural Resources for Agriculture in Sahelian West Africa. CARE International, New York.
---------. 1993. Getting It Right: Linking Concepts and Action for Improving the Use of Natural Resources in Sahelian West Africa. Drylands Programme Issues Paper 40, International Institute for Environment and Development, London.
Panganiban, E. 1994. "Democratic Decentralization in Contemporary Times: The New Local Government Code of the Philippines." Conference paper, Institute of Commonwealth Studies, University of London.
Parker, Andrew. 1995. Decentralization: The Way Forward for Rural Development? Policy and Research Working Paper 1475, Washington, DC.
Piore, M., and C. F. Sabel. 1984. The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books.
Prud'homme, Remy. 1995. "The Dangers of Decentralization." The World Bank Research Observer 10 (August): 210-26.
Putnam, R., and others. 1993. Making Democracy Work. Princeton: Princeton University Press.
de los Reyes, R.P. and S.M.G. Jopillo. 1995. Responding to LGC Demands and Challenges: Experiences of Two Iloilo Towns. Institute of Philippine Culture, Ateneo de Manila University, Quezon City.
Riggs, F.W. 1963. "Bureaucrats and Political Development: A Paradoxical View." In J. LaPalombara (ed.), Bureaucracy and Political Development. Princeton: Princeton University Press.
Robinson, M. 1988. Local Politics: The Law of the Fishes. Delhi: Oxford University Press.
Rodriguez, V. 1992. "Mexico's Decentralization in the 1980s: Promisses, Promises, Promises..." In Morris and Lowder (eds.), Decentralization in Latin America. New York: Praeger.
Rondinelli, Dennis. 1981. "Government Decentralization in Comparative Perspective: Theory and Practice in Developing Countries." International Review of Administrative Science.
Ruttan, Vernon. 1984. "Integrated Rural Development Programmes: A Historical Perspective." World Development 12 (Aril): 393-401.
Sabel, C. F. 1997. "Constitutional Ordering in Historical Context." In R. Boyer and J. Rogers (eds.), Contemporary Capitalism: The Embeddedness of Institutions. Cambridge, UK and New York: Cambridge University Press.
Saich, A.J. 1993. Discos and Dictatorship: Party-State and Society Relations in the People's Republic of China. Leiden: Rijks Universiteit Leiden.
Schiff, Maurice and Alberto Valdés. 1992. The Plundering of Agriculture in Developing Countries. World Bank, Washington, DC.
Shah, Anwar. 1991. The New Fiscal Federalism in Brazil. World Bank Discussion Paper 124, Washington, DC.
---------. 1996. "Quality of Governance and Fiscal Decentralization: Fine in Theory But What is the Pratice?." World Bank, Washington, DC.
---------. 1997. Fostering Responsive and Accountable Governance: Lessons from Decentralization Experience. World Bank Washington, DC.
Shah, Anwar and Qureshi Z. 1994. Intergovernmental Fiscal Relations in Indonesia: Issues and Reform Options. World Bank Discussion Paper 239, Washington, DC.
Sidaway, J.D. and Simon, D. 1993. "Geopolitical Transition and State Formation: The Changing Political Geographies of Angola, Mozambique and Namibia." Journal of Southern African Studies (March): 6-28.
Sidel, J. 1997. "Out of Africa, Off the Hacienda: Local Bosses, Mafias and Warlords in Southeast Asia." Draft paper for the Pacific Asia Project, Institute of Commonwealth Studies, University of London.
Simon, David. 1993. "Namibia's New Geopolitics." Indicator SA (Spring): 73-76.
Slater, R. 1994. "Strengthening Local Government in Sri Lanka." Conference paper, Institute of Commonwealth Studies, University of London.
Smith, B. 1985. Decentralization: The Territorial Dimension of the State. (London: George Allen and Unwin.
Souza, C. 1994. "Political and Financial Decentralisation in Democratic Brazil." Local Government Studies (Winter): 588-609.
de Swaan, A. 1988. In Care of the State: Health Care, Education and Welfare in Europe and the USA in the Modern Era. Cambridge, UK: Polity Press.
Tanzi, V. 1996. "Fiscal Federalism and Decentralization: A Review of Some Efficiency and Macroeconomic Aspects." Proceedings of the Annual World Bank Conference on Development Economics, 1995, Washington, DC.
Tendler, J. 1997. Good Government in the Tropics. Baltimore and London: Johns Hopkins University Press.
Therkildsen, O. 1994. "Economic Decline and Democratic Decentralization in Rural Sub-Saharan Africa." Conference paper, Institute of Commonwealth Studies, University of London.
Tiebout, C.M. 1956. "A Pure Theory of Local Expenditure." Journal of Political Economy (October).
Tilly, C. 1992. Coercion, Capital and European States, AD 990-1992. Cambridge, MA and Oxford: Blackwell (revised edition).
Tordoff, William. 1994. "Decentralisation: Comparative Experience in Commonwealth Africa." Journal of Modern African Studies, 32 (December): 555-80.
Toulmin, C. 1994. Gestion de Terroir: Concept and Development. UNSO, United Nations, New York.
van Zyl, Johan, Tulio Barbosa, Andrew Parker, Loretta Sonn. 1995. Decentralized Rural Development and Enhanced Community Participation. A Case Study from Northeast Brazil. Policy and Research Working Paper 1498. World Bank, Washington, D.C.
Webster, N. 1992. "Panchayati Raj in West Bengal: Popular Participation for the People or the Party?" Development and Change, pp. 129-63.
World Bank. 1974. Rural Development and Bank Policies: A Progress Report Agriculture and Rural Development Department Report 588, Washington, DC.
---------. 1987. World Bank Experience with Rural Development. Operations Evaluation Department, Washington, DC.
---------. 1993. New Lessons from Old Projects: The Workings of Rural Development in Northeast Brazil. Operations Evaluation Department, Washington, DC.
---------. 1995. Colombia Local Government Capacity: Beyond Technical Assistance. World Bank Report 14085-C, Washington, DC.
---------. 1997. World Development Report 1997: The State in a Changing World. New York: Oxford University Press.
Wraith, R. 1972. Local Administration in West Africa. New York: Africana Publishing Corporation.
Young, C. 1994. The African Colonial State in Comparative Perspective. New Haven and London: Yale University Press.
Zhang, Y. 1995. "China: Democratization or Recentralization?"
The Pacific Review 251-65.
|
