Holding governments accountable for service delivery : the local government councils scorecard initiative in Uganda Commonwealth

Uganda’s Local Government Councils Scorecard Initiative is a strategic social accountability initiative designed to enable citizens to demand excellence of their local governments and enable local governments to respond to citizen demands for effective and efficient service delivery. Begun in 2009, it is one of the signature programmes of Advocates Coalition for Development and Environment (ACODE), a public policy research and advocacy think tank in Uganda. The initiative is implemented in partnership with the Uganda Local Governments Association. The centrepiece of the initiative is the local government council scorecard, a capacity-building tool for assessing the performance of district-level elected officials in accordance with roles and responsibilities set out in the Uganda’s Local Government Act 1997 (as amended). Incorporating multiple links in the supply and demand chain of good governance and accountability – citizens, civil society organisations, local government and central government – the administration of the scorecard and dissemination of performance results is grounded in an action research methodology. Follow the positive reception of the initiative after eight years of implementation in Uganda, it is hoped that other countries will look to adapt this methodology to engage in similar types of social accountability initiatives. Muyomba-Tamale & Cunningham Local government accountability in Uganda CJLG December 2017 188 Introduction Decentralisation in Uganda and many other developing countries promised a new deal for citizens, as increased transparency, accountability and citizen engagement were to lead to more effective and efficient delivery of public services. Research over the past decade on the effectiveness of decentralisation for improving service delivery has however shown mixed results. On the positive side, decentralisation in Uganda has in general terms fostered participatory planning and heightened a sense of local ownership and improved accountability (Okidi and Guloba 2006). Across the country improvement has been documented in the areas of healthcare, water, and increased enrolment of children in primary school. In addition, local revenue collection has increased tenfold over the last 15 years. Nevertheless, in spite of these achievements there is widespread recognition that the delivery of public services is at best less than optimal or at worst has malfunctioned. Improvements in the key service areas of health, education, agriculture and roads are not commensurate with the levels of public investment in these areas. Rundown health centres and makeshift classrooms exist side by side with state-of-the-art private health centres and schools. Inadequacies in staffing and drug supplies continue to plague health centres. Moreover, local governments continue to be constrained by inadequate financial resources, difficulties with attracting and retaining qualified staff, as well as ongoing corruption, nepotism and elite capture (Bashaasha et al. 2011). Structurally, acrimony and conflict often bedevil and compromise service delivery. For instance, division of power between district chairpersons (the political head of the relevant jurisdiction) and resident district commissioners (appointed to represent central government’s interest and monitor the district’s activities) is contested and often creates conflict (Azfar et al. 2006). Following a decade of working with local governments, African think tank Advocates Coalition for Development and Environment (ACODE) understood local governments were underperforming as expected by the law. One key identified problem was that while the country had been undergoing a democratisation process, government had largely remained unaccountable to citizens for its operations. At the local level citizens remained disengaged, with limited participation and voice in the running of public affairs. This consequently led to a democratic governance deficit characterised by poor service delivery. In 2009 ACODE therefore developed the Local Government Council Scorecard Initiative (LGCSCI). The initiative is implemented in partnership with the Uganda Local Governments Association (ULGA). A preliminary assessment of the performance of local government councils (LGCs) in Uganda concluded that systemic and measurable improvement in the quality of public services is 1 ACODE (www.acode-u.org) is an independent public policy research and advocacy think tank based in Uganda, with operations in East and Southern Africa. ULGA (www.ulga.org) is an association for all local governments in Uganda, with the mission of uniting them and building their capacity for efficient and effective service delivery. Muyomba-Tamale & Cunningham Local government accountability in Uganda CJLG December 2017 189 fundamental to improving the confidence of citizens in both central and local governments. Furthermore, fully functional LGCs are the key link in the chain of public service delivery, providing an important source of power for citizens to counterbalance the power of central government (Tumushabe et al. 2010b). LGCSCI is a strategic social accountability initiative grounded in a sophisticated action research methodology. With the local government council scorecard as its centrepiece, LGCSCI increases the ability of citizens to hold their local government officials accountable for effective service delivery, and enhances the capacity of local government officials to respond to citizen demands. Context of decentralisation Beginning in the 1980s, many African countries, including Uganda, started to devolve central government functions to local jurisdictions. This shift was in response to both internal and external factors. Centralised states were being pressured from within to enact policies that would provide for more local and regional autonomy. Decentralisation was also part of the structural adjustment package that many countries had to adopt in exchange for aid and loans from the World Bank and International Monetary Fund. Decentralisation was largely conceived as a vehicle that would improve governance and public service delivery by 1) improved allocated efficiency through better matching of public services to local preferences, and 2) increased productive efficiency through increased accountability of local governments to citizens, fewer layers of bureaucracy, and better knowledge of local costs (Kahkonen and Lanyi 2001). Empowering local governments to engage in localised planning and programme implementation, and creating citizen ownership of government to enhance accountability, would lead to the achievement of two desired goals: increased efficiencies in service delivery and an engaged citizenry better equipped to participate in the democratic process (Government of Uganda


Introduction
Decentralisation in Uganda and many other developing countries promised a new deal for citizens, as increased transparency, accountability and citizen engagement were to lead to more effective and efficient delivery of public services.Research over the past decade on the effectiveness of decentralisation for improving service delivery has however shown mixed results.On the positive side, decentralisation in Uganda has in general terms fostered participatory planning and heightened a sense of local ownership and improved accountability (Okidi and Guloba 2006).Across the country improvement has been documented in the areas of healthcare, water, and increased enrolment of children in primary school.In addition, local revenue collection has increased tenfold over the last 15 years.
Nevertheless, in spite of these achievements there is widespread recognition that the delivery of public services is at best less than optimal or at worst has malfunctioned.Improvements in the key service areas of health, education, agriculture and roads are not commensurate with the levels of public investment in these areas.Rundown health centres and makeshift classrooms exist side by side with state-of-the-art private health centres and schools.Inadequacies in staffing and drug supplies continue to plague health centres.Moreover, local governments continue to be constrained by inadequate financial resources, difficulties with attracting and retaining qualified staff, as well as ongoing corruption, nepotism and elite capture (Bashaasha et al. 2011).Structurally, acrimony and conflict often bedevil and compromise service delivery.For instance, division of power between district chairpersons (the political head of the relevant jurisdiction) and resident district commissioners (appointed to represent central government's interest and monitor the district's activities) is contested and often creates conflict (Azfar et al. 2006).Following a decade of working with local governments, African think tank Advocates Coalition for Development and Environment (ACODE) understood local governments were underperforming as expected by the law.One key identified problem was that while the country had been undergoing a democratisation process, government had largely remained unaccountable to citizens for its operations.At the local level citizens remained disengaged, with limited participation and voice in the running of public affairs.This consequently led to a democratic governance deficit characterised by poor service delivery.
In 2009 ACODE therefore developed the Local Government Council Scorecard Initiative (LGCSCI).
The initiative is implemented in partnership with the Uganda Local Governments Association (ULGA). 1 A preliminary assessment of the performance of local government councils (LGCs) in Uganda concluded that systemic and measurable improvement in the quality of public services is 1 ACODE (www.acode-u.org) is an independent public policy research and advocacy think tank based in Uganda, with operations in East and Southern Africa.ULGA (www.ulga.org) is an association for all local governments in Uganda, with the mission of uniting them and building their capacity for efficient and effective service delivery.
CJLG December 2017 189 fundamental to improving the confidence of citizens in both central and local governments.

Furthermore, fully functional
LGCs are the key link in the chain of public service delivery, providing an important source of power for citizens to counterbalance the power of central government (Tumushabe et al. 2010b).
LGCSCI is a strategic social accountability initiative grounded in a sophisticated action research methodology.With the local government council scorecard as its centrepiece, LGCSCI increases the ability of citizens to hold their local government officials accountable for effective service delivery, and enhances the capacity of local government officials to respond to citizen demands.

Context of decentralisation
Beginning in the 1980s, many African countries, including Uganda, started to devolve central government functions to local jurisdictions.This shift was in response to both internal and external factors.Centralised states were being pressured from within to enact policies that would provide for more local and regional autonomy.Decentralisation was also part of the structural adjustment package that many countries had to adopt in exchange for aid and loans from the World Bank and International Monetary Fund.Decentralisation was largely conceived as a vehicle that would improve governance and public service delivery by 1) improved allocated efficiency through better matching of public services to local preferences, and 2) increased productive efficiency through increased accountability of local governments to citizens, fewer layers of bureaucracy, and better knowledge of local costs (Kahkonen and Lanyi 2001).Empowering local governments to engage in localised planning and programme implementation, and creating citizen ownership of government to enhance accountability, would lead to the achievement of two desired goals: increased efficiencies in service delivery and an engaged citizenry better equipped to participate in the democratic process (Government of Uganda

2015).
The adoption of decentralisation as a foundation for good governance and public service delivery in Uganda began in 1987.After years of conflict and war, expectations were high that decentralisation would improve local governance, democratisation and service delivery, all of which were deemed critical for durable peace and political and economic stability.After more than three decades of implementation, however the results remain a 'mixed bag' of reform, progress, stagnation, disappointment, and in some cases outright reversals.Many analysts regard decentralisation, strong macroeconomic stability, and relative peace and security as among the primary reasons behind Uganda's quick recovery from the near state collapse in the mid-1980s.Other analysts, however, observe that decentralisationthe centrepiece of a modernising democratic discourse in the developing worldhas proved problematic and challenging for several reasons: the sometimes conflicting roles of state agencies and elites, the international development community, local politicians; and lack of consistency and continuity in decision-making involving the allocative and CJLG December 2017 190 implementation functions of the central state (or provincial states) affecting local governments and service delivery (Bardhan and Mookherjee 2006).In the case of Uganda, the critical challenge to decentralisation has been that in taking that path, the country was venturing in a terrain where there were few and fragmented examples to learn from (Meyers 2014).As such, the transition period has had its ups and downs and continues to be a work in progress.A key element of that work is building the capacity of local government leaders to do the job they have been mandated to do, and building the capacity of citizens to hold them accountable.

Theoretical and conceptual context
LGCSCI is a strategic social accountability initiative that enables citizens to demand excellence of their local governments and enables local governments to respond effectively and efficiently to those demands.Lee (2011) argues that accountability is a benchmark of good governance and requires transparency in the relationship between government officials and citizens, a sense of obligation among government officials to be responsive to citizens, and an empowered citizenry capable of punishing their government representatives if they fail in this (Lee 2011).Social accountability refers to building accountability through citizen engagement in which "ordinary citizens and/or civil society organisations participate directly or indirectly in exacting accountability" from public sector officials (Malena and Forster 2004, p. 3), often through the monitoring of public sector performance.Social accountability initiatives increase public sector performance by bolstering both citizen engagement and the public responsiveness of states, whereas strategic social accountability initiatives have "a theory of change that takes into account the relationship between pro-change actions and eventual goals by specifying the multiple links in the causal chain" (Fox 2014, p. 22).A tactical approach, by contrast, is limited to a specific link in the causal chain.
The scorecard initiative is strategic.Whereas some social accountability initiatives focus primarily on a single link in the chaine.g.citizens, civil society organisations (CSOs), or local governments - LGCSCI encompasses all of these.The initiative's central premise is that by monitoring the performance of LGCs and providing information about their performance to the electorate, citizens will be empowered and encouraged to demand accountability from their local elected officials.This increased demand, which it is envisaged CSOs and local governments will channel upwards to the national level, will ultimately result in a more engaged citizenry, a more responsive government, better performing local government officials, and more effective public service delivery.Activating this accountability chain requires building the capacity of the key stakeholders to demand and supply better governance and service delivery and building durable linkages through which the demand and supply can flow.LGCSCI project activities focus on both: enhancing the ability of communities, CSOs, and LGCs to demand improved service delivery, and creating the opportunities for productive engagement between these key actors through which these demands can be effectively made and responded to.
The LGCSCI assessment, and the rigorous data collection and dissemination methodology surrounding it, are central to the initiative.The scorecards are actually a set of four instruments used to assess performance of four local institutions: the LGC as a corporate body, the district chairperson, the district speaker, and individual councillors.Designed in accordance with the roles and responsibilities of LGCs as outlined in Uganda's Local Government Act 1997 (as amended), each scorecard comprises 60 to 80 indicators according to the roles and responsibilities of the four levels mentioned earlier (district council, district chairperson, district speaker and councillor).Individual indicators are assigned a maximum point score and add to a maximum 100 points.Councillors, chairpersons, speakers and councils as a whole are required to provide concrete evidence of their performance for each indicator, and the results are published and disseminated at the community, district, and national levels.(A more detailed discussion of the scorecard methodology is provided below.) The scorecard is designed to fit into what some refer to as the 'missing middle' of social accountability initiatives, with a view to turning uninformed citizens into informed citizens, unresponsive government into responsive government, and unaccountable government into accountable government (Odugbemi and Lee 2011).Its effectiveness as a tool for catalysing accountability and good governance is maximised by the presence of what Fox (2014) refers to as "voice" and "teeth".
Citizen 'voice' is a key component of strategic social accountability initiatives.Voice refers to the various ways in which citizenseither as individuals or in organised formationscan express their opinions and concerns, putting pressure on service providers, policy-makers and elected leaders to provide better services or to advocate for them (Cornwall and Leach 2010).Reviews of social accountability initiatives have shown, however, that results from initiatives that rely solely on citizen voice are generally weak (Fox 2014;Lee 2011;Joshi 2013;Gaventa and McGee 2013).Many citizen report card initiatives suffer from this problem.Researchers have found, however, that citizen voices can be strengthened with the involvement of interlocutors or intermediaries who facilitate two-way communication between governing bodies and citizens, and bridge cultural and power gaps (Fox 2014).Within the LGCSCI, both ACODE and local CSOs play this role as they interface with citizens and act as a conduit for citizen voice during the process of scorecard data collection and dissemination.
However, even with amplification of citizen voice by interlocutors, effective social accountability initiatives also still need 'teeth'i.e.capacity and willingness of government actors to respond to voice.This could be, for example, through following recommendations that emerge from citizen CJLG December 2017 192 engagement processes.It could also include governmental capacity to change practices and structures that inhibit transparency through, for example, investigating grievances and changing incentive structures to discourage wasteful, abusive or corrupt practices (Fox 2014).The publication and dissemination of scorecard results at the community, district and national levels makes visible individual councillors' performance on a broad range of good governance indicators, and provides citizens with very concrete information about their elected officials.It also promotes healthy competition between councils to achieve top performance rankings.Equally significant, however, is the enhancement of government capacity to respond to citizen voice that is built into the action research methodology underlying LGCSCI.

The scorecard methodology
The LGCSCI methodology, described in ACODE's annual scorecard reports2 is grounded in an action research methodology.It combines capacity-building with evidence-based assessment of the ability of elected political leaders to fulfil their mandate (Bainomugisha et al. 2014), which is defined in Uganda's Local Government Act 1997 (as amended).This is unlike many citizen report card initiatives, which are more akin to basic consumer satisfaction surveys (Joshi 2013;Deichmann and Lall 2007).
Although some positive lessons were gleaned from the parliamentary scorecard that was implemented in Uganda between 2005 and 2008, the local government scorecard is different as it focuses more on a demand-side model of monitoring and accountability.Tumushabe et al. (2010a) argue that what was missing in the parliamentary scorecard was the realisation that citizens can be extremely powerful when they are equipped with information and empowered to demand quality service delivery from their leaders.In such a demand-side model, the roles of three major actors have to be recognised and activated through the monitoring process and deliberate outreach and capacity-building activities.The model suggests that the first and most important group of actors is citizens (Tumushabe et al. 2010a).
The LGCSCI annual assessments therefore focus on ensuring that citizens access accurate information about the performance of their leaders.Tumushabe et al. (2010b) further argued that when citizens have information about the performance of their councillors, they become empowered to demand better performance or express their disapproval by writing and submitting petitions to the speaker of council, writing letters to their councillor and engaging through meetings.The continuous existence of this demand builds upward pressure from the local councillor to the district council and upward to national elected leaders in parliament and other ministries.The model further emphasises that in order for this to happen, citizens should be fully aware not only of their own obligations but also of the roles and responsibilities of their elected leaders.Consequently, the model suggested the need to have an assessment standard with a clearly defined set of parameters and indicators where the functions become the major ingredient in building demand for improved service delivery.Essentially, unlike the methodology used in the parliamentary scorecard, the demand-side model puts citizens at the centre of the scorecard initiative.
While LGCSCI assesses political accountability, it is not a 'name and shame' undertaking.Instead, as an independent intervention, it is geared towards continuous training in order to equip political leaders to effectively fulfil their mandates.This consideration informs the design of the assessment tools and methods in such a way that researchers who are based at sub-national level can carry out capacitybuilding through and beyond the data collection process.
Importantly, the LGCSCI focuses primarily on the performance of the political arm (elected leaders) in regard to local governance.This means the initiative encompasses both executive and legislative arms working with their bureaucratic technocrats, but does not directly assess the technical or judicial arms.Unlike the central government's annual assessment of performance of minimum standards for technical staff, the scorecard initiative focuses on the political arm to deepen the demand side of democracy at local government level under Uganda's decentralised local governance system.
The first assessment, for the financial year (FY) 2008-2009, was conducted in ten districts. 3Ten new districts 4 were included during the second assessment for FY 2009-2010.In 2010, the status of Kampala District was raised to that of a city authority, which disqualified Kampala from the subsequent assessments 5 .The third assessment, for FY 2011-2012, was conducted in 26 districts. 6  The fourth assessment, for FY 2012-2013, maintained the 26 districts while the fifth assessment brought the total of assessed districts to 30. 7Funding for the assessment of the first 26 districts was provided by the Democratic Governance Facility in Uganda, while funding for the four new districts was from the Governance, Accountability, Participation and Performance programme of USAID.
Both development partners continue to support the initiative to this date.Scorecard reports for each assessment year were published and disseminated to local governments, CSOs and citizens (see, Tumushabe et al. 2009Tumushabe et al. , 2010aTumushabe et al. , 2011Tumushabe et al. , 2013;;Bainomugisha et al. 2014).

Scorecards and local government structure
The scorecards are designed to assess whether elected political leaders and their representative organs deliver on their electoral promises of promoting good governance; and, specifically, improving public 3 Amuria, Amuru, Hoima, Kampala, Kamuli, Luwero, Mbale, Moroto, Nebbi and Ntungamo. 4Bududa, Buliisa, Gulu, Jinja, Moyo Mpigi, Mukono, Nakapiripirit and Rukungiri and Soroti. 5The enactment of the KCCA Act in 2011 transformed the identify and functionality of Kampala from an ordinary local government into a government entity with direct funding and supervision from central government. 6In addition to the 20 original districts, Agago, Lira, Kabarole, Kanungu, Mbarara and Tororo were also assessed. 7By further including Apac, Arua, Masindi and Nwoya.service delivery and ensuring political accountability.In Uganda, LGCs are composed of councillors elected to represent geographically defined areas.Each council also has members elected to represent special interest groups such as women, youth, and people with disabilities.The scorecard system recognises four categories of memberschairpersons as members of the executive, district speakers, councillors, and the council as a wholeand each category has a separate scorecard.The assessment usually focuses on the previous financial year (1 July to 30 June), and lasts for a period of five In the scorecard, these themes are referred to as 'parameters'.The parameters are broken down into a set of quantitative and qualitative indicators reflecting the statutory responsibilities and functions of the elected leader or institution being assessed.

Data collection and verification processes
Assessing the political performance of local governments is a painstaking and meticulous process.
One has to define the indicators upon which data is to be collected for each category, and match the timing of the assessment to the processes of political governance at sub-national level, according to the country's financial and political calendars.Determining the final scores for the scorecards involves the collection and analysis of both qualitative and quantitative data.The process of data collection usually begins in June (the last month of the year under review) with assembling evidence from existing documents that contain records of council and councillor performance on most indicators.Based on this information, structured interviews are conducted with individual councillors, chairpersons and speakers during the month of July in order to verify and also understand fully the quantitative scores.In the case of the district council scorecard a technical officer, the clerk to council, is the main respondent.The respondents are given liberty to refer to documents or refer the researcher to documents to ensure accuracy and corroborate what they are saying.Information from structured interviews is then augmented and verified through key informant interviews, civic engagement meetings, and field visits, most of which take place between July and August.A brief description of each of these data collection methods follows.

Document review:
This process involves an extensive review of both published and grey literature, official government reports, and opinion articles.Key literature reviewed for LGCSCI annual assessments is derived from service delivery and infrastructure reports, budgets, planning documents, and minutes of district councils and their committees.The review provides the first source of information to the district research team, and empowers district researchers to conduct structured interviews from an informed position.Through the review process, policy gaps may be identified and shared with policy-makers for action, even before the scoring process is complete.

Structured interviews:
These are the first steps towards establishing the numerical scores for the scorecard parameters.As many councillors as possible are engaged in a face-to-face interview structured around the scorecard.This process is vital for collecting oral evidence, which is assessed later against documentary evidence.The information collected in structured interviews is also verified through field visits to specific service delivery units.There, unstructured interviews are conducted with service users at the units.Observation of service delivery units is supplemented with photography to verify assertions and claims of councillors.
Civic engagement meetings (CEMs): CEMs were introduced in 2015.Before CEMs, the LGCSCI methodology included focus group discussions (FGDs) with community members in areas where service delivery units are found.However, the FGDs were not as effective as they left out many community members and were not sufficiently reflective of the social accountability dimension of the initiative.CEMs address this civic engagement gap.As a social accountability initiative, the scorecard aims at building the capacity of citizens to demand better services using evidence on the delivery of services and conduct of their leaders.This capacity-building aim is achieved through organising CEMs at sub-county level (a level of local governance below the district).The CEMs are used as platforms to communicate elected leaders' scorecard results to their constituents.They are also used for civic education, by informing citizens about their duties as well as the roles and responsibilities of their elected councillors and chairpersons of districts and sub-counties.The mobilisation of participants at the CEMs targets organised groups like women, youth and the elderly, as well as professionals such as health workers and teachers.It is envisaged that these organised groups will share the information with other citizens with whom they come into contact on a daily basis.The outcomes of the CEMs are Civic Engagement Action Plans (CEAPs).These plans usually specify the most pressing issues on which citizens will urge their elected leaders to improve.Citizens commit to work together, with the support of their leaders, to draft petitions and letters to councils or in organising community meetings to solve existing problems.The outcomes from the CEMs and CEAPs have been empowering: citizens have been able to meaningfully engage their leaders with evidence, and leaders have become more responsive to an engaged citizenry.As citizens engage and leaders respond, the social accountability cycle is complete.

Key informant interviews:
Key informant interviews are conducted with technical officers in the district, including chief administrative officers, heads of departments, clerks to councils, senior administrative secretaries (responsible for sub-counties), and service delivery unit heads.These interviews focus on collecting summary information on the status of service delivery and verifying the CJLG December 2017 196 actions undertaken by the political actors during the financial year.The foregoing description demonstrates that the ultimate scores awarded result from a multifaceted data collection process and the triangulation of various types of data, information sources and methods.The validity and reliability of the scores is further enhanced by a multi-layered verification process during the month of September.
The process of generating the scorecard begins with the district research team responsible for collecting the data for each indicator.These researchers assign the initial scores.The second layer involves the lead researchers in each district, who supervise the fieldwork and review the scores assigned by the field researchers.The lead researchers are responsible for drafting the district-specific reports through which scorecard results are disseminated.The third layer comprises the LGCSCI core team at ACODE, which is responsible for the final validation of data.This team reviews every scorecard to ensure that the scores awarded are consistent with the evidence provided.
With an average of 30 scorecards per district, the core team verifies over 850 scorecards every year.
After verification, scores are computerised using the EpiData software package, and subsequently imported into SPSS where correlations and descriptive summaries are generated for each district.This process usually lasts a period of two months: October to November.8(See Bainomugisha et al. 2014 for a more detailed presentation of the methodology.)

The scorecards
Figure 1 is a picture of the first page of a completed scorecard for a district councillor.(The content of the scorecards varies for chairpersons, speakers and councils, but the format is largely the same.)The main headings -LEGISLATIVE ROLE and CONTACT WITH ELECTORATESare the parameters.Column one contains the performance indicators, column two indicates the number of marks associated with each indicator, and researchers use the third column to describe the evidence upon which a score was awarded.On the scorecard in Figure 1, we see the indicators, scores and comments related to the councillor's performance in two of the five parameterslegislative role and contact with electorates.Councillors have four primary responsibilities within the legislative area: participation in plenary sessions, participation in committees, moving motions for approval as resolutions, and providing special skills and knowledge to the council or committee.In dialogue with a task group of key stakeholders, including ULGA, the LGCSCI team developed specific indicators to assess performance in each area.For example, good performance in the area of participation in CJLG December 2017 197 plenaries (i.e.district council meetings) is measured using three indicators: attending at least four times, debating at least four times, and debating issues related to service delivery.

Figure 1: A snapshot of a councillor scorecard
In the example included, this particular councillor attained the minimum threshold of least four meetings (attended six meetings in total), debated an issue related to service delivery (in this case, education), but did not debate at least four times in the district council meetings.Thus, he received marks for the first and third indicators, but not for the second.Since scores are "all or nothing" he received five out of the eight marks possible in this performance area.The rationale for adopting the "all or nothing" approach is guided by the need for district leaders to aspire for excellence as opposed to simply focusing on the basic minimum standards of service delivery.As explained earlier, the weight or marks attached to each indicator in all the scorecards is a product of extensive consultations with a local government expert task group and district leaders who review and agree on the scorecard CJLG December 2017 198 methodology on an annual basis.Indicators that are considered to be more important that others in deepening democratic governance and civic participation will automatically carry more marks.
The scorecard example also illustrates the verification process through the difference in pencil as well as blue and red ink.Focusing on the first performance area (Meeting with Electorates) in the CONTACT WITH ELECTORATES parameter, we see that the field researchers initially awarded the councillor marks for the second and third indicators (note circled scores with pencil), but a subsequent review of the scorecard by a member of the LGCSCI core team at the secretariat crosschecked and verified the evidence attached to the scorecard and noted that there was insufficient evidence for those marks to be given and they were therefore crossed out.The blue ink has been used in the second layer of verification while the red ink is the final layer of verification before marks are awarded.

Presentation of final scores
Once scores have been finalised, scorecard reports are prepared for each district.These reports contain the scores for district council, chairpersons, speakers and each individual councillor.The centrepiece of each district's report is the table with councillors' scores, ranked in order of highest to lowest overall score.Figure 2   The first five columns of the table contain basic information about the councillor (name, political party affiliation, sub-county represented, gender, and number of terms in office).Following that are CJLG December 2017 199 columns associated with the overall performance of each councillor.While this scorecard report was for FY 2013-2014, scores from the previous two financial years are also included so that citizens and others can see trends in their councillors' performance over time.In addition, the report includes a column showing the percentage change in the overall score from the previous year.
In the example used in Figure 2, citizens from Kakiri Town Council are able to see that their female councillor (fourth row, highlighted with red box), scored 85 points and was in joint fourth position overall.The information in the columns under the 'Performance' heading reveal that the councillor's performance has been steadily improving from 58 points in FY 2011-2012 to 76 points in FY 2012-2013 and then to 85 points during 2013-2014.Her recent score represents a 9 point or 12% increase over the previous year.This kind of presentation makes it easy for citizens to track the performance of their councillors over a period of time, an important aspect of being able to hold their councillors accountable before any election.
The scorecard report also includes scores for councillors' performance on each parameter.As explained in the discussion of Figure 1, there are multiple indicators for every parameter; the score on the report is the sum of the scores for each indicator.For example, the scores for monitoring each of the NPPAs are arrived at by adding together the councillor's scores on three separate indicators: a) conducting visits to service delivery sites; b) preparing quarterly monitoring reports on site visits; and c) taking follow-up action on issues raised in the reports.
The reports also include an accompanying narrative that highlights particular achievements and challenges and provides additional information connected to factors affecting performance.Below is an excerpt from the Wakiso District report, associated with the scorecard in Figure 2, that highlights and offers an explanation for the higher scores of councillors who sit on the District Executive Committee (DEC).
Membership in certain committees also played a role in the high level of performance of councillors.For instance, councillors that also served as members of the DEC had greater opportunities and facilitation to fulfil their roles more effectively than councillors who were not members of the DEC.A councillor that serves as a DEC member is entitled to the full-time service of council that presents him/her privileges of an office, emoluments and allowances, making execution of his/her roles easier.
Not only are the parameter-specific scores and accompanying narrative context important for enhancing citizen 'voice', as people are able to use the data to hold their local elected officials accountable in specific performance areas, but this level of detail is also helpful to those being scored as it highlights jobs well done and provides specific targets for improvement.Moreover, the scores give the LGCSCI team a sense of where more 'teeth' are needed as well, i.e. more government capacity to respond to citizen voice.For example, referring again to the NPPAs in Figure 2, monitoring of the service delivery sectors continues to be a challenge for all but the very top-scoring CJLG December 2017 200 councillors.When the team dug deeper into this, they realised that there are significant budgetary and transportation constraints that impinge on councillors' ability to discharge their responsibilities in this area.Armed with this information, the project team, ULGA and the councillors themselves have begun to demand that more resources be provided to them by central government specifically for monitoring service delivery.This is an excellent example of how the scorecard, and the LGCSCI project more broadly, triggers upward demand for accountability: not just from citizens to their local governments, but from local governments to central government as well.

Lessons learned and ongoing adjustments
When the scorecard was initiated in 2009 there was initial resistance by political leaders, who feared that the assessment would be used by their political opponents to undermine them.This has however since changed, because the majority of politicians have come to appreciate the usefulness of the scorecard in making them effective and efficient in their work.The Memorandum of Understanding signed between ACODE and ULGA to implement the scorecard initiative speaks to the endorsement from local government leaders from all the districts in Uganda.
They also know that even if they do not achieve the status of top performer, being able to document improvement in performance is appreciated by voters.The capacity-building work that accompanies the scorecard process is multifaceted and has evolved over the course of the initiative.From the beginning, each round of assessment has begun with a scorecard inception meeting, organised to prepare the leaders for the next round of assessment.During these meetings councillors are reminded about where they are underperforming and how they can improve.Because the scorecard itself is set up in accordance with the official roles and responsibilities of local government leaders, councillors' understanding of their own roles is deepened through this process.
The past nine years have also provided ACODE with opportunities to address capacity gaps.A number of innovative ideas have been implemented with the aim of keeping the initiative relevant in changing social and economic times.The councillors' diaries are tailor-made products of the initiative, whose purpose was to address the problem of poor record-keeping by elected leaders.The councillors' diaries were introduced in 2011 and have since been a great source of information for verification of the final scorecards.Each district councillor and chairperson is given a diary where daily activities are recorded.The diaries also include information that would be useful to councillors, including a summary of elected leaders' roles and responsibilities, service delivery standards, and the budget cycle.Over the years, record-keeping has improved, which has in turn translated into improved performance.
The local government SMS platform is another innovation that was introduced in 2013 to create space within the civic infrastructure for citizens to engage their elected leaders in real time and at minimal CJLG December 2017 201 cost.The platform aimed to address the problem of high transport costs that district councillors face during monitoring of service delivery units.Since councillors are not adequately resourced to monitor government programmes on a regular basis, monitoring of distant service delivery units had tended to be ignored.With the SMS platform, however, citizens are trained to use it to send messages about social service deficiencies in their constituencies.Elected leaders are also trained to receive and respond to the messages.It is hoped that elected leaders use this information to make informed decisions about priorities for service delivery.The more informed district councillors are, the more focused council debates become.
At the beginning of 2013, peer-to-peer learning forums were also introduced to the initiative as an avenue for councillors to learn from each other.During the first two years, the learning forums involved taking councillors from a lower performing district to a higher performing one, to provide an opportunity for districts to share best practices and learn from each other.In 2015, however, the peer learning forums evolved into multi-district leadership forums (MDLFs), which involve bringing districts together at a regional level.
The MDLFs were designed to build on the success of the peer-to-peer learning sessions, as measured both by increased performance in the lower-scoring districts and the positive reviews of the participants on all sides.The MDLFs take place annually, and bring together leadership teams from five districts including a host district.The leadership teams include the district chairperson, speaker, clerk to council, the chief administrative officer and the resident district commissioner, all of whom are crucial to implementing and overseeing service delivery in local governments.The forums equip district leaders with practical skills to run efficient and effective LGCs.After two years of implementation, results from the participating districts reveal that the forums provide an opportunity to foster accountability mechanisms and enhance the responsive capacity of both political and technical local government leaders.They also create unity among local governments, enabling them to more effectively advocate for their needs with the Ministry of Local Government.
The CEAP methodology is one of the latest innovations, rolled out in 2015.The idea is that when citizens engage their leaders using the tools of civic engagement, the leaders are more likely to respond, which improves accountability and service delivery.In the early years of the scorecard project, it was assumed that the provision of data to citizens on councillor performance, coupled with a civic education session, would catalyse citizen demand.While those activities have no doubt had an impact on civic capacity, it was observed that more attention needed to be given to the 'voice' side of this strategic social accountability initiative.The result was the CEAPs, which are an attempt to build a social contract between the electorate and their local elected leaders on the electoral mandate and promises.Tied to the dissemination of scorecard results, the CEAPs engage communities in making sense of the results and using them to develop, in essence, a social accountability action plan.Through CJLG December 2017 202 this process, citizens not only deepen their understanding of the mandated roles and responsibilities of their local elected officials, but they also come to better understand their own rights and responsibilities as citizens, both of which are essential for holding their leaders accountable and engaging in the demand side of democracy.The CEAPs are implemented as part of the civic engagement meetings.At the close of the meeting, which is facilitated by a local CSO, citizens present their plan to their local government officials and agree with them on a timeline for implementation, progress reports, and monitoring.This process has built local government capacity in the parameter 'Contact with Electorates', and has helped them achieve higher scores in this performance area.Finally, the CEAPs have been integrated into the MDLFs as important documents to consider when leadership teams develop their own action plans for improving local governance, accountability and service delivery in their districts.

Annual review of the scorecard
Oversight of the scorecard initiative is provided by a local government expert task group which comprises local government technical and political leaders,9 CSOs,10 the Ministry of Local Government,11 members of parliament (MPs)12 and academics. 13The task group is hosted by ACODE and ULGA and meets biannually to review the scorecard methodology and address technical gaps raised by the assessments.The task group has been instrumental in ensuring that the parameters and indicators are not only credible and fair, but also aligned to address policy gaps.The task group members therefore refine the methodology and provide guidance on any innovations.
The emphasis on refining the assessment methodology and building local government capacity initially took precedence over building channels for the demand of upward accountability from local to central government in accordance with the LGCSCI theory of change (Bainomugisha et al. 2015, p. 17).The theory of change posits that by monitoring the performance of LGCs and providing information about their performance to the electorate, citizens will demand accountability from their elected officials. .However, stakeholder reflections and a formal evaluation of the first phase of the project (2009 to 2012) pointed to the need for more work to maximise the uptake of policy recommendations that stem from the assessment process.The annual national-level synthesis report of findings across the scorecard districts includes multiple recommendations to central government for new or revised policy and practices that would give local government the resources needed to fully carry out its mandates under decentralisation.

CJLG December 2017 203
In 2017, the Local Government Parliamentary Forum was launched with the goal of strengthening the link between LGCs and the central government.The purpose of the forum is to foster dialogue and bridge the gap between local governments (frontline service providers) and legislators.Meetings of the forum are held quarterly, bringing together MPs, local government leaders and CSOs from across the country.The forum is a strategic platform to debate the recommendations emerging from the districts and synthesise scorecard reports to influence policy and ensure responsive, efficient and effective service delivery at local government level.

Discussion
While there are still improvements and adjustments to be made, the scorecard initiative has registered a number of tangible positive results.Political leaders are now more aware of their mandate, and the performance of district councils and individual councillors has improved.Discussions in many district councils focus more on service delivery, and the increase in the level of record-keeping is commendable, with more councillors using their diaries to keep track of the progress of their work.
Conflicts which had riddled a number of local governments during the first two years of the assessment have been resolved, as district leaders came to realise the impact of these conflicts on their district's performance. 14On an individual level, councillor monitoring of NPPA services has greatly improved, though work remains to be done on consistent report-writing and follow-up.
An external evaluation of LGCSCI in October 2014 found that the effectiveness of the initiative in boosting councillor performance is directly related to its methodology and tools (Democratic Governance Facility 2014).The strength of the research underpinning the initiative is also recognised at central government level, as findings from the LGCSCI were an important resource during the process of reviewing the Local Government Act in December 2015.After Uganda's local elections in 2016, ACODE partnered with ULGA to train the new district councils in the 33 project districts at a time when government did not have sufficient funding for the activity.However, perhaps the greatest testament to the value of the project so far is the fact that, while many councillors had reservations and disagreements about the project during the initial years, now other districts are requesting to be included in the LGCSCI.Ranking performance has helped to deepen democratic decentralisation by creating healthy competition among local leaders, which enhances their performance.

Conclusion
Drawing on the engagement with the scorecard process, district political leaders have expressed increasing confidence in their ability to do their work more effectively (Bainomugisha et al. 2015).
One particularly positive impact documented by ACODE was the outcome of the 2016 general elections in Uganda, which pointed to the scorecard's impact on voters' choices and decision-making.
14 Mbarara District is a good example where the conflict between the district chairperson and sections of the council was resolved after ACODE organised three conflict resolutions clinics between 2014 and 2016.
months.The main building blocks in the LGCSCI scorecard are the principles and core responsibilities of local governments.These are set out in the Constitution and the Local Governments Act.They are classified into five categories or broad themes: (i) financial management and oversight; (ii) political functions and representation; (iii) legislation and related functions; (iv) development planning and constituency servicing; and (v) monitoring service delivery of the National Priority Programme Areas (NPPAs).
contains part of one such table.