linking local government discretion and accountability in decentralisation

8/2/2019 Linking Local Government Discretion and Accountability in Decentralisation

1/35

Development Policy Review, 2010, 28 (3): 259-293

The Authors 2010. Journal compilation 2010 Overseas Development Institute.

Published by Blackwell Publishing, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

Linking Local Government Discretion andAccountability in DecentralisationSerdar Yilmaz, Yakup Beris and Rodrigo Serrano-Berthet

Decentralisation offers significant opportunities to improve governmentaccountability by exerting stronger pressures both from below (demand)and above (supply). The literature contains many examples, however,where the potential has not been realised, partly because decentralisationreforms have often been introduced without thinking through theiraccountability implications. Even when accountability is taken intoaccount, the efforts tend to emphasise either the supply or the demand side

of the equation, but not both. Drawing on the sets of literature on fiscal,administrative and political decentralisation, this article presents amethodology for studying this.

Key words:Decentralisation, local governance, accountability

1 The local governance challenge: linking discretion andaccountabilityEstablished theories in economics and political science have articulated the efficiencyand accountability gains that could accrue from decentralisation reforms.

1They include

the internalisation of spillover effects (Oates, 1972; Mueller, 1996), the alleviation of

information asymmetry and better accountability due to the proximity of principals and

agents (Cremer et al., 1996; Raff and Wilson, 1997; Bucovetsky et al., 1998) and

competition among local governments (Tiebout, 1956). Nevertheless, the empirical

literature is full of examples of failures or mixed results in delivering these gains

(Asthana, 2003; Akin et al., 2005). Part of the reason and the starting point of this

article is that both the theory and practice of decentralisation have suffered from a

partial and fragmented approach, undermining the comprehensive and strategic

sequencing required for effective decentralisation reforms. We argue here that toincrease the developmental impact of decentralisation to receive the biggest

development bang for the buck it is essential to get the right local governance

framework, balancing discretion and accountability.

Respectively, Social Development Department, World Bank, 1818 H Street, NW, Washington, DC 20433

([email protected]); United Nations Development Program; and World Bank. The findings and

conclusions are entirely those of the authors, and do not represent the views of the World Bank, its

executive directors, or the countries they represent.

1. In this article we use the term decentralisation to refer to devolution central governments transfer of

administrative and financial decision-making authority to local governments that have clear and legally

recognised jurisdictions within which they provide public services to constituents to whom they are

accountable and not deconcentration and delegation.


2/35

260 Serdar Yilmaz, Yakup Beris and Rodrigo Serrano-Berthet


Development Policy Review 28 (3)

Although decentralisation reforms have been at the centre of public-sector reforms

around the world, they come in different types political, administrative, fiscal and

forms deconcentration, delegation, devolution in each country. Furthermore,

implementation of these reforms is always partial, based on the political context of a

country; any illustration is therefore selective and incomplete.

As decentralisation reforms become more widespread across the world,2

they often

try to increase the autonomy and discretion of local governments without thinking

through the accountability incentive structures that are crucial to obtain more responsive

and accountable governments. Even when accountability is taken into account, the

efforts tend to emphasise only internal governmental mechanisms, neglecting external

citizen vigilance and political oversight, or vice versa. In addition, the relationship

between discretion and accountability in decentralisation reforms is further complicated

when fiscal, administrative, and political aspects are separated a point often missed in

such reform efforts.

If the practice of decentralisation has been fragmented and incomplete, it is, to a

certain extent, because of the separate and parallel pathways that its political and fiscal

theories have followed. Public-finance economists have emphasised the allocative

efficiency gains produced by fiscal decentralisation, downplaying the importance of the

political dynamics surrounding the capture of those benefits by local elites. Political

scientists have highlighted the potential for democratic deepening and citizen oversight,

overlooking the fiscal underpinnings required to make local governments work. This

article is part of an ongoing attempt to bring these two strands closer together.

The first-generation theory of fiscal federalism made the case for decentralised

fiscal choice based on its superior allocative efficiency. In his work on the theory of

public finance, Musgrave (1959) argued that the policies of subnational branches ofgovernments should be permitted to differ in order to reflect the preferences of

residents. Carrying Musgraves arguments further, Oates (1972: 87) formulated the

decentralisation theorem as each public service should be provided by the jurisdiction

having control over a minimum geographic area that would internalise benefits and

costs of such provision. One of the shortcomings of first-generation theory was that it

largely assumed that local public officials seek the common good. The more empirical

literatures from political science and political economy prove this to be wrong,

highlighting in particular the problem of capture by the local elite.

An emerging second-generation theory of fiscal federalism is bridging the gap

between the economic and political approaches to decentralisation, taking as its point ofdeparture the assumption that participants in political processes (both voters and

officials) have their own objective functions that they seek to maximise in a political

2. In 1994, Dillinger reported that of the 75 developing countries with populations greater than 5 million, all

but 12 claimed to have embarked on some form of transfer of power from central to local governments.

Since then this transfer has been occurring even in inherently centralised countries, such as Jordan

(World Bank, 2002), Morocco (Vaillancourt, 1998), Egypt (UNDP, 2004), Yemen (UNCDF, 2007), the

Peoples Republic of China (Wong, 1997), Turkey (Tosun and Yilmaz, 2010), and Central and Eastern

European countries (Dunn and Wetzel, 2000; Bird et al., 1995) as well as authoritarian regimes likeAngola (Felicio and Yilmaz, 2009), Burkina Faso (World Bank, 2009), Ethiopia (Yilmaz and Venugopal,

2008) and Pakistan (World Bank, 2009).


3/35

Linking Local Government Discretion and Accountability in Decentralisation 261



setting that provides the constraints on their behaviour. Officials dont simply act on

behalf of the welfares of their constituents (Oates, 2005: 356).

Decentralisation reforms often lead to tensions among various stakeholders, not

surprisingly because decentralisation is about redistribution of power within and

between levels of government, with different actors having conflicting interests

(Hiskey, 2006). Analysing the political-economy drivers of these reforms is a crucial

element of implementing a coherent, well-thought-out decentralisation strategy. Often

decentralisation reforms are poorly designed and implemented, due to the lack of

understanding of the power and accountability relationship among various actors. The

political-economy approach highlights the influence of politics in framing and

implementing the reforms, showing that decentralisation is ultimately a highly political

process since it seeks to redistribute resources within the territorial confines of a given

nation-state (Agrawal et al., 1999: 2) and that the advantage of decentralisation is

compromised by capture by the local elites (Seabright, 1996; Bardhan and Mookherjee,

1998).

This article adds to this new more integrative body of literature by presenting a

methodology to better analyse the linkages between local discretion and accountability

in three dimensions: political, administrative, and fiscal. It argues that the relationship

between local discretion and accountability is far more complex than accountability

being an automatic outcome of increased discretion. In fact, increasing resources

allocated for public services and expanding local government discretion over the use of

these resources require a special attention to fixing accountability incentive structures.

Otherwise, decentralisation efforts will most likely not translate into more accountable

government.

Furthermore, all these relationships between discretion and accountability (brokendown into the supply and demand side

3of fiscal, administrative, and political aspects)

are far from being static. In all its aspects, decentralisation reshapes power relations

among the local residents, local governments, producers of local government services,

and higher levels of government (including central government). It sets new rules of the

3. The supply side of accountability, which is also known as public accountability, is the hallmark of and a

sine qua non for good governance (Bovens, 2005). It is the obligation of public authorities (governments,

elected representatives, corporate, and other governing bodies) to explain publicly, fully and fairly, how

they are conducting responsibilities that affect the public in important ways. Public accountability refers to

the institutionalised practice of account giving; it focuses on public-sector managers who spend publicmoney, exercise public authority, and manage a corporate body under public law. Until recently, efforts to

foster good governance focused on strengthening the supply side of accountability within state institutions.

Accordingly, donor initiatives supported institution-building in developing countries to increase thesupply

of governance processes by reforming public institutions. Recently, the development community has

concluded that domestic demandfor accountability, originating within civil-society entities and the public

at large, is at least as important for development as supply-side mechanisms. Also referred to as social

accountability, the demand side refers to an approach to building accountability that relies on civic

engagement in which ordinary citizens and/or civil-society organisations demand accountability.

Recognising the limitations of both electoral and public accountability mechanisms, demand-side/social-

accountability approaches require concerted civic education efforts during the decentralisation reform

process. This new understanding encouraged an expansion in the repertoire of instruments through which

citizens can hold the state to account, beyond voting. These instruments include traditional practices suchas public demonstrations, protests, and investigative journalism, as well as more innovative ones such as

participatory budgeting, social audits, or citizen report cards.


4/35




political game, helping new local leaders to emerge in the political competition. It thus

redefines the interactions between local leaders and their constituencies. Similarly, as a

result of new regulatory and financial powers over procurement and service delivery,

the decisions and actions of local governments have a greater impact on local

economies. Decentralisation thus leads to new interactions and contractual relationships

between local governments, between small and big private firms, and between providers

and producers of services and communities and non-governmental organisations.

The methodology presented here is not intended as a one-size-fits-all solution to

the complex public policy issue of decentralisation. We are aware of the risk of

generalising issues that are mostly specific to the local context; thus, our aim is not to

offer a universal prescription. However, in the analysis we do not distinguish among

different types of local governments, i.e. urban vs. rural, small vs. big, or among

provinces in a federal system vs. local governments in a unitary system. We argue that

factors presented in the methodology that affect discretion and accountability are valid,

regardless of the size or location of the local governments. This is not to suggest that the

size and location of the local governments do not matter in a decentralisation context.4

They are relevant during the decision-making process on what kind of discretion over

which services should be accorded to local governments. However, the issues of

optimal size and location of local governments are beyond the scope of this article. Our

methodology recommends that, once a decision on according a certain level of

discretion is made, balancing discretion and accountability is important to increase the

developmental impact of decentralisation.

In balancing discretion and accountability, the relationship between the central

government or its deconcentrated units and local governments presents challenges.

Central governments or their deconcentrated units may set the rules under which localgovernments operate, and provide a portion of the local governments financial

resources for service delivery. In these cases, local governments are accountable to

higher-level hierarchies for their conduct. This often creates a relationship of upward

accountability or dependence on central authorities, which may contradict the

arguments about local discretion and accountability. While warning against these risks,

the goal here is not to dismiss the important role of central governments in

accountability relationships in decentralisation. Provided with a meaningful level of

discretionary space, local governments need to be checked and balanced against abuse

of their discretions. In this respect, central governments have a set of accountability

instruments. However, such formal accountability rules and norms can be too vague andincomplete to deal with every situation and circumstance. They also tend to favour

upward accountability towards upper levels of government. In these situations, there

is a need for a minimum level of rule of law that will allow local governments to

challenge central government encroachment (Bahl and Martinez-Vazquez, 2006). Such

4. The size or the nature of local government (urban vs. rural) should be resolved in the discussion over the

kind of expenditure assignments to be allocated to local governments, taking into account efficiency and

effectiveness factors and the demands of local populations. Studies about the size of local governments

examine the efficiency implications of size in service delivery. According to this literature, smallergovernments that are closer to the people should be better able to provide the tax/service delivery package

that citizens want (Fox and Gurley, 2006).


5/35




minimum guarantees will allow independent bodies (such as courts) to arbitrate between

the central and local governments in case of conflict or disagreement.5

Although we accept the general notion of sequencing6

in decentralisation reforms,

we regard it as essential to conduct systematic analyses of country-specific local power

structures, interests, and socio-economic conditions to decide the best sequence for a

country. Beyond the elements of our methodology, any analysis of local governance

structures needs to pay particular attention to the way local governments (both

appointed and elected local officials) establish and maintain accountability relationships

with their surrounding local actors. Accordingly, the methodology put forward here is

simply a diagnostic tool for policy-makers to use when deciding on policy actions, not a

panacea. Following this introduction, Section 2 presents the key components of the

local political setting and discusses how to make local politics more accountable.

Section 3 introduces different dimensions of local administrative discretion and

accountability, and Section 4 discusses the linkage between discretion and

accountability in fiscal decentralisation. The final section puts forward possible

scenarios (trajectories) to achieve a high level of discretion and accountability in

decentralisation.

2 The local political setting and accountabilityAnalysing the local political setting is crucial to understanding the factors that drive

accountability (Lankina, 2008).7

Decentralisation reforms can restructure the local

political setting, reshaping local actor and voter incentives in many ways, such as

changing the size of municipalities, reformulating local electoral legislation, and

redefining formal relationships between the representative and executive bodies

(Keating, 1995). They can also change the structure of legislative bodies, the balance

5. This was the case when the Supreme Court of the Philippines ruled that the central government had to pay

the entitlements it was withholding from the local governments (Bahl and Martinez-Vazquez, 2006).

6. Sequencing in this particular context refers only to some preconditions for a successful decentralisation

reform. However, there is a broader debate on sequencing. The theoretical literature provides normative

discussions on broad steps for sequencing and implementing decentralisation programmes. But this is

limited to discretion (devolution of power) and is not linked to accountability. From our perspective,sequencing poses a big question: Can any one of three dimensions of decentralisation (political,

administrative, or fiscal) survive without the presence of the others? This is especially the case when

countries introduce only fiscal decentralisation without the other dimensions (for example, China and Viet

Nam). This article tries to partially address the sequencing issue in the last section under the discussion on

trajectories but refrains from making conclusive statements as it treats these cases as incomplete

decentralisation. The methodology does not present one or the other dimension as superior or conditional

for the other dimensions. But it presents different scenarios (trajectories) to reach the ultimate goal of a

high level of discretion and accountability in all three dimensions. The decision on how and whether to

reach this ultimate goal is a political one shaped by the local conditions and preferences in each country.

Our methodology presents a menu of options and possible trajectories. It would be too prescriptive if it

were to suggest a preference in these key and mostly political decisions.

7. The premise that the local political and institutional setting might substantially affect local performanceand accountability is extensively theorised and documented in institutionalism-inspired studies of

decentralisation (Lankina et al., 2007).


6/35




between elected local authorities and local executives and administrators,8

the way

councils are elected, the way executives are elected or appointed, and the structures for

local legislative and executive bodies to relate to citizens (Lankina, 2008).

The following sub-sections identify the key structural elements at the local level

that define the setting for political action and set out some accompanying arrangements

needed to improve the downward accountability relationships of local authorities with

the citizens.

2.1 Factors defining the local political setting

An appropriate political setting for downward accountability requires a suitable

environment for local elected leaders to act independently (even if it conflicts with their

own parties or with the central government) and responsively (in line with the demands

of the local population). The local leadership will be influenced by at least three sets of

factors: (i) the institutional arrangements for separation of powers among the executive,

legislative, and judicial bodies; (ii) the election laws and the electoral system; and (iii)

the existence and functioning of a party system and political party laws.

Institutional separation of powers at the local level

Strongly influencing the quality of local decision-making are the checks and balances

between the executive, legislative, and judicial branches of local government and the

clear separation of powers among them. With roles and functions clearly identified,

local councillors are expected to undertake independent oversight over local executive

bodies, and local courts take the role of impartially resolving conflicts arising from localgovernments administrative actions. Accordingly, a key prerequisite for the separation

of powers, at least in theory, is a specialised local court system able to resolve such

conflicts. In many countries, this is done through administrative courts or alternative

dispute-resolution mechanisms that mediate between local governments and citizens

about administrative actions, such as land use, zoning, and business regulations.9

Another important component of the institutional separation of powers is the

relationship between the executive and the legislative. The relationship between and the

relative weight of the local executive and the local council establish how local decisions

8. We make a distinction between the elected and executive bodies within local governments because theyhave distinct roles in local decision-making. Schroeder (2004) depicts local governments (elected and

executive bodies) as the provision units rather than producers of services, meaning that they make

decisions about the type, quantity, and quality of services to be made available in the locality, how those

services are to be financed, and how produced. He ascribes the production of services to local public

employees, private contractors, or higher jurisdictions under contract with the local government. In this

sense, bureaucracy (and its supporting personnel) is the set of actors that either carry out the production of

local services or help to oversee the private (or public) contractors that serve as production units

(Schroeder 2004: 9). We use the term local appointed official interchangeably with local bureaucracy

and emphasise the important role of local elected representatives in local accountability systems,

especially in the oversight of local governments.

9. However, in less developed countries such formalised court systems usually do not extend to jurisdictions

below a certain level of government or do not always respond to certain types of small-scale conflicts.Frequently, traditional and informal structures may also address cases resulting from local government

actions, before a formal recourse to the judicial system.


7/35




are made. Different institutional mechanisms for assigning the weight to the respective

bodies are possible:

Strong mayor system. Under the mayor-council set-up, also called a strong mayor

system, a popularly elected mayor wields strong executive, often charismatic,

authority, deciding public policy and with a strong symbolic role in representing the city

(Sisk, 2001).10

This arrangement is detrimental to local accountability if local councils

are marginalised and their role reduced to rubber-stamping the preferences of the local

executive (Lankina, 2008). In Cte dIvoire and Uganda, for example, use of the strong

mayor system has led to mayoral domination of local councils (Crook and Manor, 1998;

Wunsch, 2001).

Strong council. The second type of institutional arrangement is the election of a

mayor by the council, usually from council members, giving the council more weight

than the executive. This system is also referred to as parliamentary system. The risk of

this type of local government is that debating chambers are not able to implement

sound policy decisions (Lankina, 2008).

Council-manager. The third type, frequent under many strong council systems, is

the council-manager arrangement. The council appoints and contracts with a politically

neutral administrator to run and manage the city. Such an arrangement can preclude

politically motivated patronage (Montjoy and Watson, 1995).11

Commissioner. Under a commission form of municipal government, elected

commissioners also manage separate departments. They are both legislators and

department chairs, all with equal powers though one may have the title of mayor. The

system has been criticised for violating the principle of division of powers (Montjoy and

Watson, 1995).

Existence and quality of local electoral systems

Electoral systems change the incentives of elected local leaders and voters, during and

between elections.12

An electoral system may favour big parties, undermine alternative

10. Various arrangements exist for governing the mayor-council relationship. For example, the mayor may

have charter-based veto authority over council decisions, and procedures may exist for the council to

override this veto. Or the mayor may lack veto power. Mayor-council set-ups might have an adverse effect

on policy outcomes because political and administrative roles are not sharply distinguished under thisarrangement. Empirical studies have shown how electoral turnover, particularly when a new mayor defeats

an incumbent, results in proclivities for greater public expenditures and policy changes (Wolman et al.,

1996).

11. It has been suggested that city managers are more likely to pursue policy innovations than elected mayors

because they are ostensibly guided more by actual effectiveness and efficiency, rather than short-term

electoral considerations and pressure-group demands, which would have been the case with elected

political executives (Montjoy and Watson, 1995). City managers are not subject to frequent turnover and

thus are more likely to ensure policy continuity and to have credible commitments to other actors in local

development (Clingermayer and Feiock, 1997).

12. The concept of electoral systems is defined narrowly to refer only to the rules that determine the means

by which votes are translated into seats in the process of electing politicians into office (Farrell, 2001:

15). Electoral laws is a broader concept that includes the laws regulating all facets of the electionprocess, including nomination procedures, the characteristics of the franchise, and how campaigns are

conducted (ibid.).


8/35




voices and dissent, or encourage strict hierarchies within parties.13

More important, it

can be structured in such a way as to systematically exclude certain groups. Or it may

encourage political parties to simply win the votes of particular groups over others. In

sum, the structure of local electoral systems has an impact on the quality of local

representation and its responsiveness, and therefore the performance of decentralisation

reforms (Hiskey, 2006).

Scholars of decentralisation assert that fair, competitive and regular elections

compel local politicians to exercise power in a way that allows decentralised institutions

to provide efficient and fair outcomes (Echeverri-Gent, 1993; Crook and Manor, 1998;

Blair, 2000; Faguet, 2008). But if an electoral system does not secure real competition

among local politicians, decentralisation reforms might end up strengthening the hands

of local political strongmen. Competition among local politicians increases the chances

for vulnerable groups to be included in decision-making (Lankina, 2008). However, few

scholars provide systematic insight into whether certain electoral arrangements produce

better outcomes than others (Packel, 2008). A host of factors may shape the

effectiveness of elections as an instrument of local citizens voice. Key among them:

whether elections are based on individuals or party-nominated candidates; whether

officials being elected are perceived to wield sufficient power; whether local elections

are held concomitantly with national elections; and, of course, whether elections are

manipulated (Sisk, 2001).

In this context, the choice between proportional representation and election

through single-member districts and plurality votes, also known as first-past-the-post, is

a key institutional variation. In the emerging systems of democratic decentralisation in

the developing world, there is more variation, ranging from pure PR systems, systems

mixing PR with elections from single-member districts, first-past-the-post elections insingle-member districts, to arrangements where the winning party takes all the council

seats allocated through the election (Packel, 2008).

In PR systems, voters generally vote for a political party rather than a specific

candidate. Electoral districts contain multiple representatives, and the share of votes

received by a party is translated by a fixed formula into the number of seats to be held

by that party. So PR works to ensure that a partys degree of representation

appropriately reflects its electoral support (ibid.).

The alternative to PR systems is majority or plurality voting within single-member

districts. Here, voters in a given electoral district vote for a specific candidate. With

only one representative per electoral district, voters have greater clarity ofrepresentation they have no doubt who is directly charged with accounting for their

interests. The chief drawback is that there is no guarantee that minority interests receive

any representation (ibid.). This arrangement becomes a particular concern where

minority interests are equally distributed across the polity. A party or group that has

only a slight minority could easily lose elections in each electoral district, leaving it

with no representation whatsoever (Farrell, 2001; Gallagher and Mitchell, 2005).

There are very few systematic studies evaluating the relationship between varying

institutional electoral arrangements and accountability in local governments (Packel,

13. Rules for candidacy may force party members to have closer ties with the centre rather than their local

constituency.


9/35




2008). This is true for both advanced industrial democracies and developing countries.

It is therefore very important to develop this research agenda further to better

understand how the quality of local electoral systems reflects the local electorates

preferences, the nature of local campaigning and the performance of elected leaders.

Nature of party system and structures

The rules and structures of local political representation create (positive or negative)

incentives for local elected leaders to be downwardly accountable to all citizens.

Political parties lie at the heart of this representation. They articulate and aggregate

interests, provide channels for the recruitment of leadership, adjudicate disputes

between conflicting interests, and engage in government decision-making. They provide

the linkage between the ruler and the ruled, the policy-maker and the citizen (Lawson,

1980). Given the critical importance for local governance, it is essential to understand

the features affecting the party system and structures, namely, the existence of partisanor non-partisan systems, the role of national parties in nominating local candidates,

rules governing the financing of parties and the participation of disadvantaged groups

such as women or certain minorities, and the availability of parties based on ethnicity or

religion.

Partisan or non-partisan. Although a growing literature is looking at the

relationship between electoral competition and local government performance, a smaller

literature focuses on partisan systems compared with non-partisan ones (Packel, 2008).

Advocates of non-partisanship in local elections maintain that local government pertains

to bread and butter issues, on which there can be no division along party lines (Olowu,

2003). According to this understanding, the inclusion of parties at the local level risksallowing policy-making to become contaminated by patronage and clientelism instead

of focusing on long-term benefits (Packel, 2008). The literature argues that elected

officials may be focused on securing re-election or delivering benefits to their narrow

client base, rather than delivering policies that benefit the entire community in the long

run (Lankina, 2008). In addition, officials may be more concerned with taking measures

to ensure their promotion and advancement within the internal party structure, rather

than in promoting policies that benefit the community (Ahmad et al., 2005). In Ghana,

for example, parties have been outlawed in local elections, based on an argument that

merit, not party affiliation, is the basis of representation (Crook, 1999; Francis and

James, 2003). Indias panchayats also operate on a non-partisan basis by law(Venugopal and Yilmaz, 2009).

Allowing parties to participate in local government, by contrast, acknowledges the

link between local and national government. Examples of systems allowing partisanship

in local elections are nearly all from the recently decentralising Latin American

countries, including Bolivia and Mexico. In Africa, prominent examples of party-based

systems are Cte dIvoire, Kenya, Nigeria, Senegal and South Africa (Packel, 2008).

Role of national parties in nominating local candidates. The membership and

loyalty structure of the parties, the role of national party leaders in selecting candidates

and preparing party lists for local elections, and the hierarchy structure within parties

can also shape the behaviours of political parties. These factors can have direct orindirect effects on the ability of local political leaders to exercise their powers and may


10/35




create an environment where national priorities and national party leaders dominate

local priorities and agendas.

In many countries, parties do not command stable loyalties, with elections

focusing on elites or personalities (Azfar et al., 2001). A review of relevant case studies

shows that national parties play a role in nominations (Packel, 2008), which may affect

party loyalties. Where local elections occur on a partisan basis, nomination rules that

favour national parties can serve as impediments to downward accountability. This is

the case in Senegal, where only nationally registered parties can field candidates for

local elections. In a study comparing six Latin American countries that have taken

formal measures to decentralise forest management, Larson (2003) notes with concern

that in Bolivia, Brazil, Costa Rica, Honduras and Nicaragua, local candidates are

selected by national parties. These arrangements are also likely to diminish the capacity

of local actors to demand accountability from local elected officials, because their

choices for selecting representatives are likely to be limited (Packel, 2008).

Financing of parties. Local party financing is also critical in analysing local governance

structures. Corrupt practices related to covert party funding streams, influence peddling,

and leveraging state resources for party purposes may compromise the faith of ordinary

citizens in the political processes. A recent study by the National Democratic Institute

of 22 countries highlights many areas of concern, such as the role of wealthy business

interests in funding campaigns to gain access to lucrative state contracts (Bryan and

Baer, 2005). This is still an area with a scant literature. In many developing countries,

accurate information about party funding sources and political spending practices is not

available to the public.

Participation of disadvantaged groups. Without clear rules stipulating the inclusion of

certain disadvantaged or minority segments of society, a party system may easily engender asystem of dominance by majority and more powerful social groups. Exclusion from the party

system based on race, gender, ethnicity, or religion reflects a party agenda (and inevitably, a

local council and government agenda) favouring the majority and the more powerful. In the

absence of sanctions against such exclusion, downward accountability of locally elected

leaders is damaged, leading to skewed policy-making and greater rent-seeking.

All the elements discussed in this section define the local political setting in which

local elected officials interact with the other actors within the government and with the

citizens. There is no one-size-fits-all setting that these elements would produce; many

different combinations could be determined, depending on the country context and

preferences of the local leaders. No matter what combination is chosen, however, thegoal is to provide a setting conducive to downward accountability. The following

section discusses the key aspects that need to be taken into account.

2.2 Making local politics downwardly accountable

Political accountability is a process whereby citizens hold elected officials to account

for their behaviour and performance say, through elections (Aucoin and Heintzman,

2000). Political accountability can also be improved by having elected local officials

oversee local executives, or by involving citizens directly in decision-making beyond

elections.


11/35




Strengthening the political dimension of local accountability requires some

safeguards on the supply side regarding local electoral systems and local council

oversight (public accountability). On the demand side, political accountability measures

allow for citizen-initiated legislation (petitions), referenda, or recalls of elected public

officials, including procedures for petitions to adopt, amend, or repeal an act, law, or

executive order, to demand public hearings on policy decisions and action and to appeal

to ombudsman offices in local governments (social accountability).

Public accountability approaches

Introducing safeguards into the electoral system. Despite many weaknesses, elections

remain the principal method whereby all eligible local residents can have a voice in the

outcome and can hold decision-makers accountable (Schroeder, 2004: 9). They legitimise

local authority and provide elected representatives with a mandate for action. But little

attention has been devoted to how specific electoral mechanisms fare in deliveringaccountability, or even whether voters use elections to hold officials accountable for

certain policy decisions (Rodden, 2004). Nevertheless, the general literature on

elections does show that differing electoral arrangements shape how citizens exercise

influence on policy-makers (Powell, 2000).

An established body of literature on Western local governance suggests that, even

under perfect conditions, a voice in local elections fails to ensure that elected officials

will exercise power on behalf of all segments of the community.14

These studies suggest

structural biases against marginalised and non-elite groups built into elections even in

Western democracies. Hunter (1953) shows how power in US cities is concentrated in,

and exercised on behalf of, a narrow group of elite interests. Studies on the involvementof underprivileged segments in decision-making processes demonstrate that they

rubber-stamp decisions already reached by other interests (Schattschneider, 1960;

Bachrach and Baratz, 1962). And other studies reveal a middle-class, white-collar, male

bias in the composition of elected bodies (Balme, 1989; Newton, 1976; Mabileau, 1989;

Mabileau et al., 1989; Stone, 1989).

Some countries allow independent candidates to run in local elections, reserve

seats for women and other vulnerable groups, allow for recalls of elected officials and

limit the length of the term they may remain in office. These mechanisms are important

in countries where money, violence, and corruption dominate elections, and politics

are based on kin, personality, or patronage (Olowu et al., 2004: 71).The impact of such measures is complex. For example, re-election procedures

could determine political accountability, although two opposing relationships are

possible (Packel, 2008). Term limits could foster accountability by preventing local

politicians from becoming entrenched in their positions and locked into relationships of

patronage. Representatives who were recently private citizens would thus be more

14. As a result of a decline in faith in representative democracy, turnouts in local elections have gradually

declined over the past few decades in many countries. Elections focused on parties have likewise alienated

voters, as parties are often perceived to have lost touch with the electorate. In some countries, like Canada,

parties are now banned from participating in local elections; it is assumed that party-free elections ensurethat elected officials speak on behalf of the community and do not espouse narrow party agendas (Sisk,

2001).


12/35




attuned to the concerns and interests of the community than career politicians (Packel,

2008). But if term limits are too restrictive, council members may not be able to master

the responsibilities of their position before their terms expire.

The theory of retrospective control of politicians asserts that, given perfect

information, voters use elections to reward or punish politicians. But without the

possibility of re-election, elections lose their power of control, and politicians may

engage in more rent-seeking (Maravall, 2007). Even with imperfect information,

eliminating the possibility of re-election may at the very least reduce opportunities for

accountability (Packel, 2008). Mexico is one country that imposes term limits of three

years on local elected officials and bars them from holding the same position again for

one term. Researchers argue that these limits impede performance (Cleary, 2007;

Grindle, 2006).

Recall gives councils or popular bodies the ability to dismiss elected leaders for

wrongdoing. Unlike elections, however, the decision is confined more to evaluations of

past actions. In Africa, recall provisions exist in Ethiopia and Nigeria (Olowu, 2003).

Indias panchayati raj law varies across different States, but some of the most

progressive legislation for recall exists in Madhya Pradesh, with detailed regulations for

gram panchayataction (Mathew and Mathew, 2003) and where thegram sabha (village

assembly open to all adult residents) has the right to dismiss the panchayat

chairperson in the event of wrongdoing (Johnson et al., 2005).

If a provision for recall exists, its design matters if it is to be an instrument of

accountability. Recall alone may not engender downward accountability if the only

actors capable of exercising this power are beholden to national political parties (Packel,

2008). In Bolivia, the widespread use of the law allowing the town council to recall

mayors in cases of misconduct, known as the voto constructivode censura (constructivecensorship vote), illustrates this risk. In 1997, one year after the first mayors entered

office following implementation of decentralising reforms, 30% of them were replaced,

indicating that the voto constructivo was being used as a political manoeuvre, not a

response to corruption (Hiskey and Seligson, 2003). Indeed, this high use demonstrated

that the recall hindered accountability, as mayors wound up being selected by council

members rather than by the electorate, and with less citizen support for the political

system than in municipalities where the recall was not exercised (ibid.).

Improving local council oversight. Local councils are the core units of

representative governments. Improving local council oversight therefore constitutes an

important part of public accountability approaches, relying as it does on the assumptionthat local elected representatives have more incentive to respond to the needs and

preferences of local populations and are more downwardly accountable than local

bureaucrats. The relationship between elected local councillors and executives also

pertains to budget planning, execution, and monitoring and evaluation. Local

councillors are supposed to oversee the executive branch during the entire process of

public financial management and provide local executives with constant feedback.

A number of factors interfere with the oversight responsibility of local councillors.

First, there is the status of local councillors (Lankina, 2008). Although executive

positions are generally considered to be full-time, in many settings councillors are low

paid and part-time, their council duties motivated by civic spirit and volunteerism(Pelissero and Krebs, 1997). Second, there is the lack of safeguards against dual


13/35




structures of accountability (Lankina, 2008). In some developing countries, councillors

often occupy dual roles as full-time executive functionaries dependent on, and

accountable to, higher-up bureaucracies and elected local officials. In some parts of

Russia, local councils are packed with senior professional employees of medical and

educational institutions (Lankina, 2004). Such councillors are less likely to aggregate

and articulate the preferences of the citizens who elected them.15

Other countries demonstrate similar conflicts of interest. In Ghana, a third of local

councillors are formally appointed by state bodies, while in Kenya as well as Ghana,

members of parliament can be ex officio members of local councils (Olowu et al., 2004).

In India, some states have power to remove elected representatives or even dissolve

panchayats (Mathew and Mathew, 2003). In many countries, councillor positions also

overlap with key patrimonial, social, or other structures in the locality, impeding

accountability.

Social accountability approaches

Social accountability mechanisms can give poor and marginalised people a more direct

voice in the policies that local governments formulate and implement than through

elections and local councils. Studies of African countries where elections are in place

show that local citizens fail to sanction poorly performing officials effectively (Azfar et

al., 2001). Patronage, clientelism, elite capture and elite biases are still widespread.

Social accountability mechanisms are often part of broader efforts to deepen democracy

and ensure a robust public sphere for citizens to give feedback and control government

action.16

The practical form of such participatory arrangements includes public meetings,citizen juries, forums for various social groups, such as the young or the elderly,

neighbourhood assemblies, multi-choice referenda accompanied by active public debate

and discussion, and activism by non-governmental organisations and other community

groups. In some settings citizens are excused from work and asked to meet to make

recommendations about local issues, thus drawing even the normally passive and

uninterested citizens into public life (Lankina, 2008).17

Logistically, some of these

deliberative forums could be complicated and time-consuming, but observers point to

their merits, namely, that consensus-based decision-making may be more legitimate

than that of elected officials which cannot be realistically scrutinised on a daily basis

(Sisk, 2001).

15. In some instances, public employees, such as school principals, appointed by higher-level bureaucracies

on a contract basis, sit in local councils. They tend to satisfy the political preferences of regional

authorities because their jobs are subject to short-term contracts, which could be revoked. Even heads of

private enterprises are subject to control and manipulation by higher authorities because their tax

privileges and licences could be withdrawn at random (Lankina, 2004).

16. These efforts represent a new approach to democratic governance that some scholars have called

Empowered Participatory Governance (Fung, 2004). It looks at how alternative political and

administrative designs can surpass conventional democratic institutional forms on the quite practical aims

of enhancing the responsiveness and effectiveness of the state while making it more fair, participatory,

deliberative, and accountable (Fung and Wright, 2001: 8).17. The size of the locality and its cohesiveness are important factors determining the effectiveness of such

forums (Sisk, 2001).


14/35




These powers to hold political leaders accountable may be expanded: (i) by

generic legislation empowering citizens to demand explanations and justifications from

local governments; (ii) by setting up specific bodies and processes for citizen oversight;

and (iii) by creating a political culture for citizen oversight.

Generic legislation. Some initiatives introduce legal mechanisms empowering

citizens to redress grievances or request explanations of municipal legislation. Public

hearings and consultations, the right to demand a public hearing, public petitions, and

the right to initiate a recall or referendum are examples.

Specific bodies and processes. These can comprise all citizens in the municipality

(gram sabha in India), several citizen representatives (Vigilance Committee in Bolivia),

or an elected position (citizen ombudsman in Japan). In some countries local

development programmes transfer discretionary resources to local governments on

condition that they create multi-stakeholder forums, with representation from civil

society, local councillors, and deconcentrated offices.18

In the Philippines, for example,

the Local Government Code mandates all provincial, municipal, and village

governments to establish local development councils to set the direction for economic

and social development and review local government budgets, one-quarter of the

members to come from non-governmental and community-based organisations (Estrella

and Iszatt, 2004).

Creating a political culture. By introducing mechanisms for poor and marginalised

people to participate in decision-making and for local transparency and accountability,

community-driven development programmes have promoted a culture of citizen

oversight.19

A community-based programme that empowers citizens to be informed and

have a say over local public expenditure creates the expectation that all programmes

should follow the same standards.

3 Local administrative discretion and accountabilityTo have flexibility in delivering services and the opportunity to respond to local needs,

local governments need administrative autonomy. We identify three broad areas of

authority as being crucial for local governments to be administratively autonomous:

power to make, change and enforce regulatory decisions/laws; to govern a procurement

system (based on national standards); and to make civil service/employment decisions.

18. Although these forums provide a venue for greater co-ordination and control, they need to be properly

designed to complement the role of the local council.

19. A community-driven development approach tries to improve the well-being of poor people by increasing

their control over the way investment resources are planned, executed and managed. Because communities

do not act in isolation but in a local space where they interact with local governments, sectoral service

delivery units, the private sector, and civil-society organisations, the community-driven approach hasrecently been expanded into a broader local governance approach that captures the quality of these local

interactions (Helling et al., 2005; McLean et al., 2006).


15/35




3.1 Factors affecting local administrative discretion

Ability to regulate

As part of administrative autonomy, local governments need a minimum set of powersand capacities to initiate, amend and enforce regulatory legislation on issues within their

jurisdiction, subject to national and state laws. Their powers usually extend to local

economic development, land-use planning and management, zoning, and public safety

and in certain cases, some aspects of public health, social protection, education, and

environmental protection. These enhance their discretionary authority to make decisions

and take actions concerning who can benefit from given resources or opportunities,

how, and to what extent (Agrawal and Ribot, 1999: 480). For example, research on

environmental regulatory networks for shrimp farming in Thailand shows that local

governments are the most effective regulators (Vandergeest, 2007).

They also need the power to sanction and punish non-compliance, one such toolbeing the administrative penalty, usually a monetary fine or the revocation of licences

or rights related to, say, an economic activity or land use. Although the power to

sanction requires a qualified workforce to monitor non-compliance, endowing local

governments with such power could prove more practical and cost-effective than

prosecution through litigation.20

Discretion to procure and administer services

Expanded mandates and responsibilities for new services require local governments to

be given discretion over procurement processes for goods and services,21

which in turn

requires flexibility in procurement laws and regulations and high-quality employees

well trained in public procurement, ethics, and contract management (Brennan and

Miller, n. d.). Such discretion implies that, under strictly stipulated national standards

and regulations, local governments can develop procurement strategies,22

identify

associated processes, and issue contracts for goods and services.23

Discretion over civil service and employment policies

In many countries political and fiscal devolution may have proceeded apace, but

administrative changes may only approximate deconcentration, the result being

20. Administrative penalties do not replace criminal prosecution, and the legal system is important for local

administrative autonomy. Enforcing local administrative rulings is effective and legitimate only when

there is recourse to challenging local government decisions. As discussed in Section 2, the existence of a

specialised administrative court system for local affairs would be instrumental in establishing checks and

balances between local executive and legislative bodies.

21. Private-sector participation in service delivery through service and management contracts, leasing,

concessions and joint ventures alleviates pressures on budgets. It also brings in skills and knowledge,

improves the efficiency of service delivery and insulates operations from political interventions.

22. A procurement strategy should stipulate mechanisms for assuring compliance with the competitive bidding

and against abuses, such as bid splitting, change of orders, and restrictive bid specifications.

23. A local government procurement policy is closely associated with the quality of service administration andthe way contracts are awarded to partnerships in service delivery, including for firms of various sizes,

social enterprises, minority groups businesses, and voluntary and community organisation suppliers.


16/35




weakened accountability for service delivery, and conflicts of interest (Ahmad et

al., 2005: 10). Civil servants form a crucial link in moving government closer to people;

therefore civil service management or more broadly human resource management

should be seen as an essential component in the design of decentralisation rather than a

separate, stand-alone process (Green, 2005: 129).

Local government competency and discretion over civil service and employment

policies ideally cover autonomy and transparency in pay policy (overall wage rates as

well as allowances), control of budget and establishment, autonomy in recruitment,

control of career management, and performance management (Evans, 2004: 24).24

The

degree of autonomy differs across countries; Table 1 provides a general picture of the

situation in East Asia.Each of these factors increases a local governments administrative autonomy and

influences its accountability relationship with local bureaucracies. However, many

countries suffer from misaligned structures of accountability in decentralisation

(Lankina, 2008). In Pakistan, for example, despite devolution of responsibility for

education to school districts, teachers remained provincial government employees, with

the elected district executive, nazim, having little authority over them (Ahmad et al.,

2005). Similarly in Uganda, despite devolution in health care, key decisions and drug

provision remain the prerogative of the centre, but salaries and staffing are that of the

District, undermining local incentives for efficiency and responsiveness (Azfar et al.,

2001).

Granting administrative autonomy to local governments in civil service

management results in reallocating powers and jobs, geographically and institutionally.

Issues related to status, prestige and labour mobility often impede relocation or

dislocation of civil servants across tiers of government. Bureaucracies tend to resistdecentralisation reforms because of career perspectives and institutional and political

allegiances. Such resistance can lead to field officers maintaining strong links with their

original line ministries, thereby enjoying some insulation against local control. This is

often the case in countries where central government officials were simply transferred to

local governments after decentralisation reforms were instituted. The result is that staff

burdens are transferred to local governments, without the discretionary powers over the

civil service. In India, for instance, though most State legislation has clarified the

functions to be devolved, the States have frequently been unable to transfer

administrative and technical and technical controls over locally administered

programmes. In China, on the other hand, although subnational employees make up90% of total government employment, the central government retains a significant

degree of control over numbers and wage levels, structure and career and performance

management of the subnational civil service (Green, 2005). Both in China and the

Philippines, the double subordination of local government staff to the local executive

and the relevant central agency weakens the downward accountability of local civil

servants (ibid.).

24. Some form of accountability to higher levels of authority may be unavoidable for local authorities

performing state-delegated or -funded tasks. It is, however, important that there be no confusion aboutresponsibilities or distorted incentive structures whereby the local authorities are deprived of leverage over

service providers within their jurisdictions.


17/35




Table 1: Staffing authority among subnationalgovernments in East AsiaEnabling mechanisms

Cambodia

China

Indonesia

Philippines

Thailand

Vietnam

Budget control

Determine wage

Dismiss surplus staff

Establishment control

Control overall staffing

numbers

in individual offices

and facilities

Recruitment

Formal employer

Have authority to hire

Have independent

merit-based recruitment

mechanism

Career management

Promotion available

Internal transfers

possible

Horizontal mobility

Performance management

Direct and supervise

Conduct evaluations

Offer financial rewards

Discipline and fire

Pay policy

Set overall wage rates

Set local incentives/

top-ups

Note: = yes; = partial; = no. Ratings refer to the subnational level and de facto practices as well as de

jure authority. Data are for most recent available year, ranging from 2000 to 2003.

Source: Green (2005).

3.2 Making local administration downwardly accountable

Accountability for the exercise of administrative powers requires avenues for

complaints and redress, accessible to officials and others. Oversight of the use of civil

service control includes measures to improve budget transparency on staff payments, on

policies and practices for new appointments, and on practices for budget and


18/35




establishment control. Similar measures are needed to ensure transparency and openness

of the procurement process to avoid misconduct and corruption. These measures call for

appropriate channels for administrative audits that can be initiated by bureaucrats or

elected leaders and by civil-society groups.

Public accountability approaches

Public accountability in the administrative sphere refers to local civil servants being

accountable to their top administrative officers and to outside officials or entities such

as public audit officers, ombudsmen, regulators, a particular administrative agency, or a

board or committee. Three major mechanisms that public sector approaches rely on to

improve this accountability are structures within bureaucratic hierarchies, specially

designed independent bodies, and administrative courts.

Accountability structures in the bureaucratic hierarchy enable higher authorities to

instigate investigations or audits of the use of administrative discretion by lowerbureaucrats, and are important as a first step to uncover information about

misadministration by local governments. Issues arising due to lack of sound control and

audit systems range from collusive practices in procurement in Indonesia (World Bank,

2003) to a lack of compliance with procurement laws in the Philippines (World Bank,

2004b).

Specially designed independent bodies authorised to conduct administrative audits

on local governments have emerged in response to increased complexities requiring

specific expertise, and have become a widespread accountability measure. Examples

include independent/external auditors who scrutinise the use of public funds,

ombudsmen who hear citizens complaints about regulations, decisions and actions,25

and theme-specific bodies (such as the anti-corruption commissions, environmental

review boards, or commissions for sustainable development) that analyse whether local

administrative decisions are in line with national sectoral strategies.26

Administrative courts with local expertise address local conflicts arising from local

governments regulatory and administrative decisions, and can ensure compliance with

national laws and regulations. Such a court system issues binding decisions and may

resolve cases that the local governments and associated independent bodies fail to

address. In France, for example, les tribunaux administratifs are the court of first

instance with full jurisdiction over disputes related to local government actions.

25. The office of the Ombudsman in the Indian State of Kerala is a 7-member body consisting of a High Court

judge as the chairperson, other judges and non-political representatives appointed after consultation with

opposition parties. In addition, there are appellate tribunals as a source of appeal against panchayat

decisions. Only impeachment could remove members of these bodies, which minimises the chance of

executive interference with their work (Mathew and Mathew, 2003).

26. Legislatures set up these agencies to make inquiries, obtain information, and issue (limited) regulations or

judgments (Zarei, 2000). In some African countries, like Ghana and Uganda, there are local judicial orconflict-resolving agencies, run by volunteers, which may be customary, providing sound redress for local

grievances (Olowu et al., 2004).


19/35




Social accountability approaches

Public accountability approaches are necessary to provide channels for uncovering basic

information on regulatory decisions and civil service practices and services. But public

approaches alone fall short of ensuring proper use of administrative discretion. Citizen-based efforts concentrate on producing the information that is most relevant for local

citizens welfare monitoring the quality of services and the way contracts and tenders

are given at the local level. Recent approaches emphasise the need for citizens to initiate

administrative audits to reveal more information on employment and pay policies as

well as relationships between local governments and service providers.

Information provision as a basis for citizen monitoring. Lack of information on

financial allocations often leads to the abuse of funds; making information available to

the public is thus the starting point for many social accountability initiatives. In India,

social audit committees, comprising individuals with impeccable reputations, scrutinise

local decision-making and publicise findings on how public money is allocated andspent (Mathew and Mathew, 2003). In the State of Rajasthan, public hearings with wide

publicity and social mobilisation forced public officials to return public money they had

misallocated (ibid.).

Monitoring procurement and implementation. The contracting and implementation

of public works and services suffer from a high risk of corruption and mismanagement.

Citizen-based initiatives complement government accountability mechanisms that

manage these risks supervising contracting and bidding, monitoring construction, and

auditing budget execution.

A typical source of corruption and collusion involves drafting tender documents in

ways that unfairly benefit one contractor over others. Citizens have organised publicconsultations in which different parties get a chance to comment on draft tender

documents, and independent outsiders conduct an in-depth analysis before the start of

bidding. In the Philippines, the Local Government Code defines ways in which

accredited non-governmental and community-based organisations can have a seat on the

prequalification, bid, and award committee for local contracts. In Argentina, the

Municipality of Morn, assisted by the local chapter of Transparency International,

introduced two mechanisms to monitor the contracting of the waste collection service,

which had been widely criticised for alleged corruption. At an extraordinary session of

the city council, attended by 500 people, the value of the contract was reduced from

about $45 million to $32 million.Citizens have also been involved in overseeing the opening and analysis of bidding

offers, as with the Social Investment Fund in Nicaragua (Grun, 2000). They can also

oversee construction while it is taking place, and are trained to see that funds are spent

as allocated and that the construction follows the standards agreed to in the contract.

Monitoring local service provision. The level and quality of service provision are

probably what citizens care about most. Participatory assessments and feedback surveys

are often accompanied by agreements on the standards of services expected.

One of the main innovations that drew attention to the potential of the social

accountability approach was the citizen report cards in Bangalore, India (Paul, 2002).

The cards are participatory surveys that solicit user feedback on the performance ofpublic services. They are used in situations where there are no demand-side data, such


20/35




as user perceptions of the quality and satisfaction with public services, and they gather

demand-side data about state-owned monopolies, many of which lack incentives to be

responsive. The report-card process relies on extensive media coverage and civil-society

advocacy to achieve greater accountability in other Indian States, Ghana, Malaysia,

Senegal and Uganda. In Uganda, a recent report-card project, aimed at reinforcing

health providers accountability to citizen-clients by enhancing communities ability to

monitor providers, increased the quality and quantity of primary health-care provision

(Bjorkman and Svensson, 2007).

Another example of social scrutiny of local governments is social funds

committees, user groups that ensure that services are delivered as intended.27

A

complement is the citizen charter, a pact between the community and service providers

that spells out expectations and roles, enabling citizens to interact more effectively with

the local authority. An example is the Citizens Charter of the Municipality of Mumbai,

India, which covers detailed public services for each municipal department. Other

strategies have relied on the creation of new multi-stakeholder institutions to promote

citizen oversight over a specific local government service. Examples include the local

school councils in Chicago in the United States (Fung, 2004) and citizen community

boards and school management committees in Pakistan (ADB/DFID/World Bank,

2004).

Citizen monitoring of administrative decisions should be seen as a complement

and not a substitute for public accountability mechanisms. Citizen-based actions may in

fact fail to provide effective oversight. Local citizens often lack the knowledge to assess

adequately the quality of complex services, which results in market imperfections in

the sense that they cannot properly hold local policy-makers accountable (Lankina,

2008). What voters can assess is often wasteful, but highly visible, projects, whoseimplementation also rewards narrow clients, not the broader citizenry (Ahmad et al.,

2005). This was the case in the Czech Republic, where one badly performing

municipality decided on a very costly but also highly visible bridge across the river

running through the city (Lankina et al., 2007). Similar caution is needed to judge the

quality of participation in community activities to monitor local government service

delivery. Even if participatory mechanisms are in place, which might boost a given

localitys overall economic development, the levels of participation within the

community are likely to be lower among those less economically advantaged and the

benefits of the new participatory arrangements biased toward those who are better-off

(Lankina, 2008). A study in Uganda found that education and income weredeterminants of membership in such key local government committees as health and

school management (Azfar et al., 2001).

27. The empirical record of these mechanisms is mixed, and there is no systematic evaluation of their

effectiveness (Olowu et al., 2004).


21/35




4 Fiscal discretion and accountability28Local government performance is intrinsically linked to the scope and nature of

intergovernmental fiscal arrangements. If local governments are denied the fiscal

instruments and funding to make real use of their political and administrative autonomy,

decentralisation is doomed.29

Intergovernmental fiscal rules determine the expenditure

responsibilities and revenue resources of local governments as well as the design of the

transfers system and local governments access to capital markets (Bird, 2000).

4.1 Determinants of local fiscal discretion

Expenditure assignment

Devolving expenditure responsibilities to local governments is an important step in

increasing the participation of citizens in local decision-making.30

A genuine spirit ofdecentralisation requires assigning a meaningful level of expenditure responsibilities to

local governments with service autonomy so that they can respond to local needs. More

importantly, they need the discretion to make their own allocation decisions (with the

necessary reporting, monitoring and sanctioning).

In this devolving of responsibilities there should be no room for ambiguity. Clear

assignment of roles and responsibilities is decisive in shaping accountability

relationships. If the contest over service-delivery responsibilities is not resolved, it can

easily constrain local autonomy and reduce the local governments credibility and

responsiveness. A related issue is unfunded mandates imposed from above. These

reduce local budgetary autonomy, as was the case in East European countries as theyembarked on decentralisation in the early 1990s (Bird et al., 1995).

A clear assignment of service responsibilities requires a well-defined institutional

framework that describes the roles and responsibilities of different levels of government. Lack

of this is a common problem in East Asia. In Indonesia, for example, although the

decentralisation law of 1999 gave all authorities to local governments, since authorities are

broader than functions there is confusion about who is responsible for what, from legislation to

planning to implementation (Mountfield and Wong, 2005: 95). Clear assignment of

expenditure responsibilities becomes even more important in sectors where line ministries and

other government agencies may also deliver services at the local level often in the same

geographic area. This situation often exists for education, health, and social services in manydeveloping countries. In China and Vietnam, for example, the administrative system

28. There is a robust literature on the theoretical and implementation aspects of fiscal decentralisation the

intergovernmental framework (IGF). On the other hand, the interaction between IGF and the political and

administrative dimensions as well as fiscal discretion and accountability are inadequately studied. Thus,

this section does not provide an extensive review of IGF issues, but focuses on the linkage between IGF

and discretion and accountability.

29. Fiscal decentralisation requires public services with high local-public-good characteristics to be assigned to

local governments and the rearrangement of roles and responsibilities among different levels of government.

30. In countries where decentralisation does not end up transferring such responsibilities to local governments,central government departments and public-sector companies continue to deliver most services with high

local characteristics, such as primary education, health, and public security.


22/35




operates as nested hierarchies, and this creates ambiguity about the distribution of

functions in a multi-tiered setting (Mountfield and Wong, 2005).

Revenue assignment

Oatess decentralisation theorem (1972) states that local governments should provide

services to identifiable recipients up to the point where the value placed on the last

(marginal) amount of services for which recipients are willing to pay is equal to the

benefits they receive. The implication is that local governments must be given the

authority to exercise own-source taxation to self-finance local services at the margin,

on the premise that local governments are more accountable when relying on their own

tax bases (Faguet, 2008), and would be less accountable when the pleasure of

expenditure benefits is separated from the pain of taxation (Bahl and Schroeder, 1983).

Significant revenue autonomy and some tax-effort incentives are critical to

encourage downward accountability and increase the efficiency of local governmentoperations.

31Although there is no set of prescribed rules in revenue assignment,

subnational governments should at least have rate-setting authority over locally

assigned revenues (Bahl, 1999a). A completely local tax is one that is assessed and

collected by local governments at rates decided by and with proceeds accruing to local

governments (Bird, 2000; Bird and Vaillancourt, 1998).

In many countries, the central government interferes with local revenue autonomy,

putting direct or indirect restrictions on local governments discretionary space. In

Uganda, local governments may impose additional taxes, but only with the approval of

the Ministry of Local Government; however, there are no standards established for

approval or rejection (Azfar et al., 2001). Restrictions on revenue generation are similarin the Philippines; the Local Government Code constrains local revenue collection

through rules on rates, assessments, appeals, and administrative responsibilities

(USAID, 1999). The accountability implication of each constraint is that they might

create incentives to make inefficient investment decisions, especially if costs can be

shifted to central budgets (Campos and Hellman, 2005: 242).

Financing the fiscal gap

The design of intergovernmental transfer systems has implications for accountability

because it affects the fiscal dependence on central government and local revenue-raising

ability. Four elements determine this dependence (Yilmaz and Bindebir, 2003): the rulesdetermining the total amount of transfer the distributable pool; the way the pool is

allocated among local governments; the purpose of the transfer system an

unconditional general purpose grant versus a conditional specific transfer; and the

design and management of the system.

A high degree of dependence on transfer revenues and a poorly designed system

shift the focus of local government accountability away from citizens to central

31. In some cases, local governments do not seek revenue-raising discretion to avoid accountability. In other

cases, they do not collect taxes assigned to them. In China, for example, local governments have takenback door approaches, for example, contracts with enterprises, tax holidays, tax administration, and

trades for investment in public services (Bahl, 1999b: 79).


23/35




governments. In other words, they create an imbalance between downward and upward

accountability incentives at the local level, negatively affecting local government

discretion and the incentive to respond to citizen demands. In Nigeria, for example,

local governments over-dependence on intergovernmental transfer revenues, coupled

with uncertainty about the amount and timing, facilitates local evasion of responsibility

under the guise of fiscal powerlessness (Khemani, 2006: 22), while in Uganda, where

transfers, as earmarked conditional grants, finance more than 80% of local government

expenditures, a substantial degree of line ministry control over local expenditure

decisions is maintained.32

Financing infrastructure: local government borrowing

Local borrowing is the fourth pillar of the intergovernmental fiscal system. Access to

capital markets gives local governments an opportunity to finance local investment

needs. However, heavy reliance on borrowing has put macroeconomic stabilisation atrisk. In the late 1990s, in Argentina, for example, perversely structured

intergovernmental systems destabilised the economy.

In countries where central government cannot credibly commit to a hard budget

constraint, there are irresistible incentives for local governments to exploit borrowing

and extend their budgets beyond their means (Faguet, 2008). In Bra

linking local government discretion and accountability in decentralisation

Documents