This report is made possible by the support of the American people through the United States Agency for International Development (USAID). 
The contents of this report are the sole responsibility of IMPAQ International, LLC and do not necessarily reflect the views of USAID or the 
United States Government.
 
MID-TERM PERFORMANCE EVALUATION
OF THE USAID’S GOVERNMENT
ACCOUNTABILITY INITIATIVE (GAI)
ACTIVITY IN SERBIA
November 2020
GAI-SUPPORTED SERVICE CENTER IN THE CITY OF SOMBOR | CREDIT: CITY OF SOMBOR
i | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
CONTENTS
TABLE OF EXHIBITS I
ACRONYMS AND ABBREVIATIONS II
EXECUTIVE SUMMARY III
1. INTRODUCTION 1
1.1 Evaluation Purpose 1
1.2 Evaluation Methodology 1
2. FINDINGS 2
2.1 Evaluation Question 1: GAI Accomplishments to Date 2
2.1.1 To strengthen the capacities and connections of the key stakeholders to increase 
government accountability at national and local levels 3
2.1.2 IR1. Increased public participation in local government decision-making and oversight 5
2.1.3 IR2: Increased follow-up to oversight institutions’ recommendations 6
2.1.4 IR3: Increased adjudication of corruption cases 8
2.2 Evaluation Question 2: Linking the Activity’s Objectives and Results 9
2.2.1 Improving Sector Policies 9
2.2.2 Effects on Participating Institutions 12
2.2.3 Management Practices 13
2.3 Evaluation Question 3: Scope of Work Expansion 14
2.3.1 Assessment of the Current SOW 14
2.3.2 Additional Areas of Intervention that Could Fit Under the SOW 18
2.3.3 Flexibility to Changing Needs and Priorities 21
2.4 Evaluation Question 4: Future Programming 22
2.4.1 Interventions with Greatest Traction 22
2.4.2 Unintended Results 24
2.4.3 Best Practices 25
2.4.4 Lessons Learned 27
2.5 Evaluation Question 5: Gender Equality 28
3. CONCLUSIONS AND RECOMMENDATIONS 29
3.1 Conclusions 29
3.2 Recommendations 31
ANNEX A. EVALUATION STATEMENT OF WORK 33
ANNEX B. ACTIVITY BACKGROUND 36
ANNEX C. ANALYSIS OF KEY PERFORMANCE INDICATORS 40
TABLE OF EXHIBITS
Exhibit 1: The WJP Open Government Index for Serbia, 2015-2020 ..................................................................4
Exhibit 2: LTI for GAI-Supported Local Self-Governments.....................................................................................4
Exhibit 3: Perceptions of GAI’s Gender Integration (n=72)..................................................................................29
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | ii
ACRONYMS AND ABBREVIATIONS
ACA Anti-Corruption Agency 
AP Action Plan
AVP Automated Case Management System for Courts
COP Chief of Party
CPI Commissioner for Information of Public Importance and Personal Data Protection
CRDA Community Revitalization through Democratic Action 
CSO Civil Society Organization
DAC Development Assistance Committee
DO Development Objective 
DPO Data Protection Officer 
EC European Commission 
EQ Evaluation Question
ERCC Electronic Register of Corruption Cases 
EU European Union 
FGD Focus Group Discussion
GAI Government Accountability Initiative
GRECO Council of Europe’s Group of States against Corruption
IOI Independent Oversight Institution 
IR Intermediate Result
KII Key Informant Interview
LAP Local Anti-Corruption Plan
LSG Local Self-Government
LTI Local Transparency Index 
MELP Monitoring, Evaluation, and Learning Plan
NA National Assembly 
NACS National Anti-Corruption Strategy 
OECD Organisation for Economic Co-operation and Development
PII Personally Identifiable Information 
PPO Public Prosecutors’ Offices 
SAI State Audit Institution 
SAPO Software Application for the Prosecutor’s Office
SOW Scope of Work 
TOC Theory of Change
USAID United States Agency for International Development 
WJP OGI World Justice Project Open Government Index
iii | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
EXECUTIVE SUMMARY
The United States Agency for International Development (USAID) in Serbia engaged IMPAQ 
International, LLC. (IMPAQ) to conduct an independent, mid-term evaluation of the USAID-funded 
Government Accountability Initiative (GAI). This $8.5 million, four-year activity started in February 2018 
and will extend through February 2022. The evaluation team independently assessed the effectiveness of 
the activities implemented to date, determined whether the Activity has achieved planned results, 
identified gaps in performance against targets, and provided recommendations on how to close these 
gaps in the Activity’s remaining period of performance. This evaluation informs USAID and GAI 
implementing partners of the success of the Activity, captures useful lessons learned, and provides 
recommendations to enable GAI to make course corrections and strengthen programming to achieve a 
greater and more sustained impact. 
METHODOLOGY
The evaluation team conducted primary and secondary research, collecting qualitative, and quantitative 
data, to address the evaluation and assessment questions in the Scope of Work (SOW), between 
September 23 and October 13, 2020. The team’s primary research consisted of 44 Key Informant 
Interviews (KIIs) and four Focus Group Discussions (FGDs), reaching a total of 66 persons. It also 
included an online survey, which received 72 responses. The team answered the following five key 
evaluation questions (EQs): 
1. To what extent is the program on track to accomplish the inputs, outputs, and likely outcomes 
expected under all three intermediate results of the activity?1
2. What links can be determined between progress towards the Activity’s objectives and the 
Activity’s results?
3. What other areas of interventions could fit under the SOW beyond what is already planned?
4. What are the interventions with greatest traction, unintended outcomes, and/or lessons learned 
from USAID assistance that the Mission can use in making decisions for future programming?
5. Did the Activity affect gender equality within the area(s) of focus?
FINDINGS AND CONCLUSIONS
EQ1: To what extent is the program on track to accomplish the inputs, outputs, and likely outcomes 
expected under all three intermediate results of the activity?
The evaluation team found that GAI is on track to successfully deliver the expected inputs and outputs, 
meeting the established targets and contributing to organizational development and capacity building of 
the institutions that are critical to increasing government accountability. Still, to date, GAI’s efforts have 
not led to any measurable improvements in government accountability, the desired likely outcomes, 
especially at the national level.
● IR1: Increase public participation in local government decision-making and oversight.
Authorities from participating municipalities recognized that support from GAI has been effective in 
enhancing public participation, especially concerning local budget planning.2 Namely, GAI’s efforts to 
develop capacities and create a culture of openness, engagement, and public information sharing 
contributed to an enabling environment for public participation in GAI municipalities. Within this 
1 The evaluation changed the original wording of EQ1 included in its statement of work and the work plan to 
reflect that GAI is structured according to three IRs to achieve an overall objective.
2 KII_08, KII_14 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | iv
framework, GAI also has supported the preparation, adoption, and implementation of local anti￾corruption plans (LAPs).
Citizen participation in reporting on high-level corruption cases has increased: in 2019 courts in 
Serbia received 152 new cases based on the law on whistle-blower protection (compared to 122 in 
2018). Out of the total caseload of 220 cases, 160 cases were finalized (compared to124 in 2018). 
National partners recognized that GAI’s support for the development of capacities and mechanisms 
to protect whistleblowers at the local level contributed to the courts in Serbia receiving more cases 
based on the law on whistle-blower protection.
3 The full extent to which GAI’s support is
contributing to increased participation, however, is yet to be seen.
● IR2: Increase follow-up to oversight institutions’ recommendations. The evaluation finds 
that IR2, by prioritizing only follow-up to independent oversight institution (IOI) recommendations, 
did not include important dimensions of IOI organizational development and capacity strengthening. 
Analyzed from this broader perspective, GAI is on track to successfully enhance IOIs, most notably, 
the State Audit Institution (SAI) and the Commissioner for Information of Public Importance and 
Personal Data Protection (CPI). 
With assistance from GAI, SAI enhanced its performance audit capacities and sound risk assessment 
system to select public institutions (including LSGs) for audit, among other services. SAI’s activity 
report (including performance audit reports) was discussed in the relevant parliamentary committee
and also in the plenary session (in June 2019).
4 Consequently, the National Assembly (NA) adopted a 
conclusion recommending the government to act within its powers to ensure the implementation of 
SAI recommendations.5 Also, GAI facilitates interaction between SAI and LSGs, ensuring appropriate 
post-audit assistance and implementation of recommendations on the most challenging topics (for 
example, public procurement processes and implementation of program budgeting). 
The Anti-Corruption Agency (ACA) recorded some improvements regarding the follow up on its 
recommendations: ACA submitted in 2019 a total of 76 requests related to a failure to promptly 
submit asset declarations for misdemeanor proceedings (compared to 45 in 2018). Based on ACA’s 
work, misdemeanor courts issued 65 convictions in 2019 (compared to 49 in 2018).6
In this context, 
GAI developed the prioritization criteria for selecting declarations for annual review and proposed 
the methodology to review assets and income declaration statements, thus, prioritizing the public 
officials that submitted assets declarations.7 However, challenges were identified at the ACA 
management level that impeded activities related to assets and income declaration, due to 
“inadequate software.”
8
In addition, GAI supports ACA to prepare and adopt its mid-term strategic 
plan and its 2020 operational plan. 
3
Progress on implementation of the Action Plan for Chapter 23-The Ministry of Justice of the Republic of Serbia 
4
http://www.parlament.gov.rs/aktivnosti/kalendar.998.html
5
https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf, page 118
6 The ACA filed 25 criminal charges in 2019 (2018: 16) due to reasonable suspicion that a public official had not 
reported 27 assets or had given false information about assets intending to conceal the facts. Based on the agency’s 
previous work, 19 final judgments were issued.
7 The Law on Prevention of Conflict of Interest in Discharge of Public Office proscribes an obligation to public 
officials to report their assets to the Republic Board for resolving the conflict of interest. Under "public officials", 
the Law includes all persons elected, nominated or appointed to organs of the Republic of Serbia, autonomous 
provinces, municipalities, and towns, as well as to organs of public companies founded by the Republic, province, 
municipality or town.
8 KII_02
v | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
Concerning the implementation of CPI’s recommendations, improvements have not been recorded.
9
Most public authorities do not comply with the obligation and recommendations to provide data to 
CPI regarding citizens’ requests for information.
10 Nevertheless, GAI has been effective in supporting
CPI to increase awareness and deliver educational programs on personal data protection for local 
administrations and entities directly or indirectly funded by their public budgets. 
● IR3: Increase adjudication of corruption cases. Although corruption remains an issue of 
concern, the implementation of the Law on the Organization and Competences of State Organs in 
the Fight Against Organized Crime, Terrorism and Corruption has decreased the number of finalized 
high-level corruption cases (compared with the previous years).
11 In this context, the key informants 
recognized that "GAI played a critical role in implementing the new Law from its adoption."12
According to the national partners, "GAI provided capacity development support to specialized anti￾corruption court units (established under this law), including tools for judges, public prosecutors, and 
other legal specialists to implement the legal framework."13 Part of these efforts included GAI’s 
assistance for the design of an Electronic Register of Corruption Cases (ERCC) to enable Serbia’s 
enforcement and judicial institutions to collect and analyze corruption related data, thus creating 
basis for assessing the effectiveness of anti-corruption measures and policies. 
GAI sensitized stakeholders and contributed to the creation of an enabling environment for the
adjudication of corruption cases, GAI also implemented the Anti-Corruption Legal Clinic at the 
Faculty of Law in Belgrade, providing students with “a condensed set of knowledge on the most 
important anti-corruption topics.”14
EQ2: What links can be determined between progress toward the activity’s objectives and the 
activity’s results?
Due in part to GAI’s efficient implementation, adaptive management approach, and meaningful 
engagement with partners, preliminary results (largely at the policy and institutional levels) indicate that 
GAI is steadily progressing towards its objectives. 
● Improving Sector Policies. The policy cycle model shows that GAI’s implementation has yielded 
positive achievements to date. For example, GAI brought to the policy agenda the need to further 
strengthen, expand, and institutionalize the protection mechanism for whistleblowers. GAI also 
produced tested models that could be used to prepare specific sectoral policies, including support to 
LSGs to implement the new Law on Public Procurement. Also, the Post-Mortem Analysis of 
Corruption Cases could serve as a policy tool to improve specialized anti-corruption court units and 
Public Prosecutors Offices (PPOs). Furthermore, GAI has effectively designed tools and instruments 
to facilitate policy implementation, such as the Performance Audit Policies and Procedures Manual, 
9
EU Progress Report for the Republic of Serbia, https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
10 The CPI recognized in 2019 that a total of 4,321 (in 2018: 3,444) of citizens’ complaints – most of them 
concerning administrative silence – were well-founded and the authorities should have provided the information 
requested by citizens according to the law on access to information of public importance. In 1,786 or 41 percent 
(2018: 1,889 or 55 percent) of these well-founded cases, the authorities reported that they acted the CPI's request 
to provide the information to the complainants. Administrative enforcement of the decisions taken by the CPI has 
yet to be ensured. Ref to EU Progress Report, 2020 and the CPI Progress Reports
11 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf, page 25 
12 KII_06
13 The reference has been made to the Practicum for Investigation and Adjudication of Corruption Cases and the 
accompanying training program.
14 KII_11
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | vi
the Plea Agreement Guidelines, and the Practicum for Investigation and Adjudication of Corruption 
Offences.15,16
● Institutional Level Impact. Thus far, GAI has successfully applied tools and methods to enhance 
accountability and transparency of partner-institutions at national and local levels, ensuring a high 
degree of ownership and participation of partners. 
● Management Practices. GAI has efficiently implemented activities and demonstrated flexibility 
when encountering operational challenges to date. GAI’s achievement of planned results can, in part, 
be attributed to its management mechanisms, including implementation modalities, organizational 
structure, and the role of each partner under GAI’s components. GAI has established a clear work 
plan and a regular flow of reporting, although with limited results-oriented practice. However, the 
Monitoring, Evaluation, and Learning Plan (MELP) was not sufficient to effectively monitor GAI. 
Additionally, GAI-collected data lacks sex-disaggregation and gender sensitivity. 
EQ3: What other areas of interventions could fit under the SOW beyond what is already planned?
GAI’s existing SOW is well-defined and fully aligned with EU accession governance priorities in Serbia 
and the needs of the partner institutions. The partners recognized the importance of their organizational 
development and improvement of core services, emphasizing “that GAI assistance and its tested services 
need to continue, with focus on root causes of challenges.”17 However, there are some potential areas 
for additional GAI involvement at the sectoral level (for example, the overall anti-corruption strategy), 
institutional level (for example, strengthening local anti-corruption councils’ capacity to monitor plan 
implementation or strengthening mechanisms for private sector whistle-blowers), and individual level 
(for example, addressing the needs of youth or enhancing understanding of participation benefits for 
citizens) to complement already planned and implemented activities.
EQ4: What are the interventions with greatest traction, unintended outcomes, and/or lessons learned 
from USAID assistance that the Mission can use in decision making for future programming?
GAI has consistently responded to Serbia’s national capacity development needs and is steadily 
strengthening national and local level capacity to improve government accountability. Future progress 
and its sustainability, however, rests on the continued commitment and capacity of Serbia’s governance 
system, institutions, and citizens. 
GAI has produced unintended results in different areas of its intervention. For example, the training 
programs favored horizontal knowledge sharing among the employees from different LSGs, and among
IOIs. With support from GAI, stakeholders have been exposed to international best practices and 
models tested in the United States. This experience, together with established partnerships, generated 
additional readiness and willingness among participating institutions to propose reviews of both policy 
and legal documents to better reflect principles of transparency and accountability.
The evaluation team identified lessons learned from GAI implementation. The need to provide focused 
expertise and policy advice, together with long-term support, remain crucial to ensuring sustainable 
organizational development of IOIs. Still, without follow-up on findings and recommendations, primarily 
15 Prepared with the purpose of harmonizing ACA’s filings with the courts’ consideration of plea agreements.
16 The Practicum provided guidance on the investigation and prosecution of corruption cases.
17 The participants in the survey gave priority to capacity development (30 percent), inter-institutional cooperation 
for anti-corruption (25 percent), and the improvement of delivery of services at the local level (20 percent).
vii | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
through the involvement of NA, these efforts do not contribute to systemic change and establishment of 
good governance.
At the local level, authorities could fail in their efforts to enhance participation if they cannot ensure 
citizens’ trust through genuine program implementation and trust promotion. In this context, the need 
to ensure local ownership and commitment of LSGs and their leaders are prerequisites to sustainable 
reform efforts. Continued results-focused monitoring should be in place to ensure that targets are met 
and results reported to citizens. 
GAI also generated a number of best practices, such as the use of the LTI, engaging youth in budgeting 
processes in Sombor, supporting SAI to introduce performance audits, and strengthening cooperation 
between CPI and LSGs.18
EQ5: Did the activity affect gender equality within the area(s) of focus?
To date, GAI has not affected gender equality. This can largely be attributed to a lack of consideration of
gender issues during program design and insufficient attention to gender mainstreaming throughout 
implementation. The Activity’s gender integration efforts were primarily a formality, notably in the 
documentation of the sex of the Activity’s training participants.
RECOMMENDATIONS
Through analysis of primary and secondary data, the evaluation team identified achievements and 
challenges during the implementation of GAI and explored possible responses to these findings. The 
evaluation team formulated the following main recommendations for USAID and GAI’s consideration:
Recommendation 1. The evaluation team recommends that GAI focus on the next stage of 
strengthening government accountability at national and local levels, which should include helping SAI, 
CPI, ACA, specialized anti-corruption court units, and LSGs implement organizational strategic priorities 
and follow up on improvement of services. This could involve helping to build capacities, systems, and 
connections among the key partners and institutions. 
Recommendation 2. Development of research and analytical capacities of ACA staff remains a 
dominant priority for this organization and GAI is well-positioned to continue building on the existing 
achievements and extend its technical assistance. However, if ACA's leadership continues to resist 
implementing its oversight mandate and the new procedures and policies that GAI supports, GAI should 
reconsider supporting ACA. 
Recommendation 3. GAI should remain flexible and responsive to the needs and priorities of the 
good governance framework. Concerning responsiveness, emerging priorities include:
● Systemic Level. To engage more actively in the policy making process related to anti-corruption, 
transparency, and accountability, GAI should codify and summarize knowledge in the core areas of 
its work and share it with the government as the basis for or evidence to create a more favorable 
legal and policy environment in this area. GAI also should provide specific support to IOIs to 
respond to the corruption risks and challenges posed by COVID-19.
● Institutional Level. GAI should: 1) continue support to local authorities – priorities could include: a) 
designing sound and realistic monitoring system for LAPs; b) enhancing the capacities of local 
independent anti-corruption bodies, in charge of LAPs monitoring; c) enhancing the capacities to 
implement recommendations to improve budgeting process and local finances; and d) enhancing 
18 Best practices reflected several criteria such as the effects on participating institutions, the durability of 
established mechanisms, and the partnerships. 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | viii
strategic communication skills of local officials in senior positions; 2) explore opportunities to 
engage with public enterprises, including public utility companies, and enhance their transparency, in 
line with the internal rules, legal provisions and requirements; 3) expand the reach of and embrace 
private sector interest, their roles and responsibilities in advancing good governance, and to fight 
corruption efforts in the country; 4) strengthen and establish whistleblower protection mechanisms 
for the private sector, enabling them to report on corruption; 5) strengthen capacities of the NA 
and its relevant committees to read, understand and comment on the oversight institution reports, 
and present findings and recommendations; 6) strengthen the technical capacity of prosecutors; and 
7) explore opportunities to support the network of DPOs.19
● Individual Level. GAI should: 1) invest in additional efforts to reach youth, offering capacity 
development programs to prevent corruption, including anti-corruption clinic models, on-line 
training programs, and the use of social networks and on-line discussion forums; 2) prepare and 
implement programs on anti-corruption framework for the journalists; and 3) increase “demand” for 
participation, by working with the public to create broader interest and willingness to take part in 
local decision-making processes.
Recommendation 4. GAI should define appropriate qualitative and quantitative indicators to enable 
GAI to adequately measure progress under components and capture GAI’s progress towards its 
objective.20 The evaluation team recommends that GAI provide a well-balanced combination of 
qualitative and quantitative indicators to capture changes and results attributable to the Activity, using 
national indicators and targets to the extent possible. The evaluation team also recommends including 
gender-sensitive indicators with a focus on “gender transformation,” identifying assumptions and risks, 
and implementing a sound risk management system. 
Recommendation 5. In addition to including gender-sensitive indicators, GAI should consider options 
to better integrate gender priorities in its activities and the overall anti-corruption framework. 
Additionally, GAI should consider activities and approaches that go beyond gender participation.
Recommendation 6. GAI should continue implementing its systemic approach to capacity 
development, following needs assessments of the main stakeholders and partners. Considering the 
current stage of the governance systems, structures, and capacities of partner institutions in Serbia, GAI 
should consider a long-term, needs-based capacity development approach. The evaluation team 
recommends that GAI design a comprehensive training evaluation approach based on the Kirkpatrick 
model, assessing four levels of learning: 1) participant satisfaction with the training; 2) immediate change 
in individual knowledge and skills; 3) change in individual performance back in the workplace; and 4) 
change in the overall performance of the institution. Finally, GAI should prepare a plan to hand-over 
training programs to the relevant national institutions and prioritize preparing them to assume these 
responsibilities.
19 GAI’s role has been to support analysis of the budgeting processes and prepare recommendations for 
improvements; however, the Action did not provide follow up support, to implement recommendations.
20 For example, indicators could not adequately measure GAI’s performance in the areas such as right to 
information, civic participation, and complaint mechanisms. GAI did not introduce appropriate indicators to 
measure development of capacities of the target groups or the effects on anti-corruption education programs on 
the selected youth. Also, GAI did not measure the effects of awareness campaigns and assess if the targeted 
audience has enhanced understanding of a specific topic. The MELP also did not include more specific indicators 
related to assistance to CPI and LSGs.
1 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
1. INTRODUCTION 
1.1 EVALUATION PURPOSE 
The United States Agency for International Development in Serbia (USAID/Serbia) recognized the 
challenges facing accountable and representative governance in Serbia, due in part to the lack of 
adequate checks on central authority and sufficient opportunities for citizens to hold their government 
and elected officials accountable, and, accordingly, funded the Government Accountability Initiative 
(GAI). GAI is a four-year activity that seeks to strengthen the capacities and connections of key Serbian 
government institutions and stakeholders, resulting in increased government accountability at the 
national and local levels. To achieve this objective, the Activity is structured according to three 
intermediary results (IRs): 1) increase public participation in local government decision-making and 
oversight; 2) increase follow-up to oversight institutions’ recommendations; and 3) increase adjudication 
of corruption cases. This $8.5 million, four-year Activity started in February 2018 and will end in 
February 2022. More background on GAI is included in Annex B.
USAID/Serbia commissioned IMPAQ International, LLC (IMPAQ) to conduct a mid-term performance 
evaluation of GAI. The evaluation provides the Mission with a fact-based and independent assessment of 
the Activity’s performance, ensuring an in-depth analysis of the Activity’s progress in achieving 
established targets. The evaluation also presents the evaluation team’s analysis of the recognized gaps 
between expected and actual performance and identifies source(s) of these gaps by considering the 
relevance of the design, project management, changes in operating environment, changes in context, 
success factors, and the Activity’s achievement of intended results and goals. The evaluation team 
conducted a mixed-methods study to assess the Activity’s impact and the likely sustainability of results. 
In parallel, the evaluation team identified relevant lessons learned with a view to strengthen the 
Activity’s remaining period of performance and inform future programming. 
In conducting the mid-term performance evaluation, the team focused on the following five evaluation 
questions (EQs) specified in the evaluation scope of work (SOW):
1. To what extent is the program on track to accomplish the inputs, outputs, and likely outcomes 
expected under all three intermediate results of the activity?21
2. What links can be determined between progress toward the Activity’s objectives and the 
Activity’s results?
3. What other areas of interventions could fit under the SOW beyond what is already planned?
4. What are the interventions with the greatest traction, unintended outcomes and/or lessons 
learned from USAID assistance that the Mission can use in making decisions for future 
programming?
5. Did the Activity affect gender equality within the area(s) of focus?
1.2 EVALUATION METHODOLOGY
To conduct the mid-term performance evaluation of GAI, the evaluation team used a theory-based and 
utilization-focused approach, incorporating complementary qualitative and quantitative methods of data 
collection and analysis utilizing online and in-person tools tailored for different stakeholder groups. The 
team’s methodology and evaluation quality standards drew on USAID Evaluation Policy and the 
Organisation for Economic Co-operation and Development’s Development Assistance Committee 
(OECD DAC).22 Before collecting data, the evaluation team conducted a thorough review of 
21 The evaluation team changed the original wording of EQ1 included in its statement of work and the work plan to 
reflect that GAI is structured according to three IRs to achieve an overall objective.
22 USAID Evaluation Policy (January 2011, updated October 2016). More details available at:
https://www.usaid.gov/sites/default/files/documents/1870/USAIDEvaluationPolicy.pdf
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 2
documentation related to the EQs to gain an in-depth understanding of GAI activities and to assess the 
availability and quality of secondary data to answer the EQs.
DATA COLLECTION. From September 23 through October 13, 2020, the evaluation team conducted 
primary and secondary research in Serbia, collecting both qualitative and quantitative data, to inform 
each of the EQs. The primary research included Key Informant Interviews (KIIs) and Focus Group 
Discussions (FGDs), for a combined total of 66 participants (30 male and 36 female) representing 24 
organizations. KII and FGD participants were identified from the full list of partners received from GAI, 
based on their involvement, knowledge, and awareness of the Activity. The evaluation team summarizes 
its data collection methods below:
● Key Informant Interviews. The evaluation team conducted 34 online and 10 in-person KIIs with 
1) USAID and GAI implementing partners and grantees; 2) central Government of Serbia 
representatives; 3) Local Self-Government (LSG) representatives; and, 4) Civil Society Organizations 
(CSOs), business entities, and other experts.
● Focus Group Discussions. The evaluation team conducted four online FGDs with a total of 22
participants from four LSGs with which GAI-USAID signed memoranda of cooperation: the 
municipalities of Vrnjacka Banja, and cities of Sombor, Šabac, and Novi Pazar. 
● Online Survey. The evaluation team disseminated an online survey to Activity participants 
(collecting 72 responses), including LSG representatives who received technical assistance from GAI 
and representatives of Independent Oversight Institutions (IOIs) and Public Prosecutor’s Office staff 
who participated in GAI’s trainings.
23,24
DATA ANALYSIS. The evaluation team’s methodology was based on contribution analysis, enabling the 
team to evaluate credible causal claims about interventions and their results. While answering the main 
EQs, the evaluation team used the OECD DAC criteria, namely, relevance, coherence, effectiveness, 
efficiency, impact, and sustainability. Additionally, in addressing all EQs, and specifically EQ5, the 
evaluation team assessed the extent to which, if any, GAI has integrated gender mainstreaming within 
the intervention, looking across all GAI components and activities.
KNOWN LIMITATIONS. The evaluation team identified several challenges and limitations in preparing 
for and conducting the evaluation, including COVID-19, subjectivity, and semi-structured KIIs, 
inconsistent accounts, the situation in Serbia after the 2020 local elections, and ethical considerations.
Annex A presents the Evaluation Statement of Work, which includes the EQs.
2. FINDINGS
2.1 EVALUATION QUESTION 1: GAI ACCOMPLISHMENTS TO DATE
To what extent is the program on track to accomplish the inputs, outputs, and likely outcomes 
expected under all three intermediate results of the activity?
In response to EQ1, the evaluation team sought to assess the performance of GAI at its mid-point and 
determine the extent to which the program is on track to accomplish the inputs, outputs, and outcomes 
expected under each of its three IRs. Utilizing primary and secondary data, the evaluation team assessed 
the progress and performance that GAI has made toward increasing public participation in decision 
23 The response rate has been 57.2 percent as a total of 126 representatives of the stakeholder institutions have 
been invited to participate in the survey.
24 Best practices reflected several criteria such as the effects on participating institutions, the durability of 
established mechanisms, and the partnerships. 
3 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
making and oversight to boost transparency and accountability, strengthening the capacity of 
independent oversight institutions to increase the uptake of recommendations, and, increasing 
adjudication of corruption cases. The team assessed GAI’s progress in strengthening the capacities and 
connections of the key stakeholders to increase governmental accountability at both the national and 
local levels. The evaluation team discusses its findings related to the Activity’s objectives and its three 
IRs below, and provides a more detailed analysis of the related performance indicators in Annex C.
2.1.1.1 TO STRENGTHEN THE CAPACITIES AND CONNECTIONS OF THE KEY STAKEHOLDERS TO INCREASE GOVERNMENT 
ACCOUNTABILITY AT NATIONAL AND LOCAL LEVELS
Despite GAI’s results to enhance the capacities of oversight institutions, these efforts have not led to 
any measurable improvements in government accountability at the national level.25 To date, GAI has 
made progress towards its indicators and has contributed to strengthening the capacities of participating 
oversight institutions to more effectively hold the government accountable.
26 The primary data sources 
indicated results were achieved for accountable and transparent governance at the local level, ensured 
by GAI’s assistance to prepare and adopt local anti-corruption plans and establish independent 
monitoring bodies, improve budgeting processes and delivery of services, and strengthen mechanisms to 
protect whistleblowers. Still, the oversight of the “executive remains rather formalistic and, thus, the 
democratic accountability of the executive branch remains weak.”
27
The following discussion analyzes GAI’s objective and its indicators. GAI’s objective has shown some 
challenges related to its scope and focus: it is neither results-focused nor “specific, measurable, 
adequate, realistic, and time-bound.” The objective is also compound, established by merging two 
independent components related to improved capacities and connections (of institutions) and 
government accountability (at national and local levels). Despite these weaknesses, the objective has 
been integrated in GAI intervention logic and its chain of results.
28 The indicators at the objective-level,
OI-I – World Justice Project Open Government Index (WJP-OGI), OI-2 – Local Transparency Index 
(LTI), and O1-3 – Good Governance Training, have been partially relevant to capture and measure
progress towards the objective. 
● O1-1. WJP OPEN GOVERNMENT INDEX. WJP-OGI is a third-party indicator published by the 
World Justice Project to measure “government openness based on the general public’s experience 
and perceptions.”29 While the Index is focused on measuring government openness, a desired impact 
of GAI’s intervention, the evaluation team recognized the limits of WJP-OGI to adequately capture 
GAI’s direct contribution or measure GAI’s progress towards the stated objective. The scope of the 
WJP-OGI exceeds a single project, and only broad contribution links can be established. Yet, due to 
its complexity and ability to monitor desired change at the impact level, the evaluation team 
recommends using WJP-OGI for country-level assessment, as it promotes broader systematic 
thinking and collaboration with other interventions.
In analyzing the WJP-OGI over the past five years, the evaluation team observed a negative trend, 
declining from 0.51 in 2015 (Serbia was ranked 61 out of 102 countries) to 0.47 in 2020 (ranked 76 
out of 128 countries). Exhibit 1 displays the WJP Government Index for Serbia for the last five years. 
25 The project reports and other project products, supported by evidence and sources indicated in the GAI MELP, 
and triangulated with the primary data collected through interviews with stakeholders created the basis to assess 
and validate progress.
26 For example, SAI and to some extent the specialized anti-corruption court units.
27 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
28 GAI objective, which accurately describes the “intended changes in development conditions resulting from the 
joint work of GAI and key stakeholders,” OECD DAC and KII_01 
29 https://worldjusticeproject.org/our-work/research-and-data/wjp-open-government-index/wjp-open-government￾index-methodology
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 4
The WJP-OGI for Serbia and its declining value illustrates the governance challenges in Serbia, 
including the government’s lack of genuine commitment, and the lack of capacity of governance 
structures, especially at the national level, to work on challenges under the good governance 
framework. 
Exhibit 1: The WJP Open Government Index for Serbia, 2015-2020
● OI-2. LOCAL TRANSPARENCY INDEX. The LTI, a third-party indicator published by 
Transparency Serbia, provides a “direct insight into the functioning of the service centers and 
premises of the local administrations” to measure transparency levels across the country.30 In 
analyzing the 2020 LTI, the evaluation team found higher LTI scores among the LSGs that participated 
in GAI’s intervention than those that did not.31 Twelve out of the 13 participating LSGs (92 percent) 
recorded an increase on the LTI scale compared to the previous year, meeting or exceeding GAI’s 
annual targets. The evaluation team identified the Municipality of Dimitrovgrad as the sole LSG that 
did not meet its target due to organizational changes that caused delays in implementation of planned 
activities. The first group of LSGs to receive GAI’s intervention scored an average of 59 out of 100 
and the second 53. These values are considerably above the national average of 46.32 Exhibit 2 shows 
the LTI in 2019 and 2020 for GAI-supported LSGs.
EXHIBIT 2: LTI FOR GAI-SUPPORTED LOCAL SELF-GOVERNMENTS 
GAI LOCAL-SELF GOVERNMENTS
LTI 2019 LTI 2020
TARGET ACTUAL TARGET ACTUAL
First Group
Dimitrovgrad 49 38 52 51
Raška 65 44 68 47
Šabac 44 41 47 57
Sjenica 36 34 39 39
Sombor 55 52 58 80
Vranje 65 60 68 73
Vrnjačka Banja 45 62 48 63
30 https://www.transparentnost.org.rs/index.php/en/research-on-corruption/lti
31 The Local Transparency Index for 2020 was conducted in the first quarter of 2020, based on research 
conducted on 145 LSGs (cities and municipalities) and 25 city municipalities throughout Serbia.
32 The national average LTI for 2020 is 46 is an improvement in comparison to the average 2019 LTI score of 40 
for 6 points.
Source: WJP Rule of Law Index®
0.47
0.47
0.49
0.56
0.51
2020
2019
2017 - 2018
2016
2015
5 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
Second Group 
Stari Grad, City Municipality N/A 23 26 28
Kragujevac N/A 43 45 55
Niš N/A 34 37 46
Novi Pazar N/A 66 69 82
Sremska Mitrovica N/A 45 48 51
Žabalj N/A 36 39 55
Source: LTI-Transparency Serbia and GAI MELP
● OI-3. GOOD GOVERNANCE PRINCIPLES AND ACCOUNTABILITY TRAINING. GAI 
monitors the number of public-sector employees and representatives of CSOs trained in the core 
areas of good governance. The team’s analysis of available data shows that GAI has provided capacity 
development support for the planned number of trainings and participants.33 For example, 294 civil 
servants and local officials were trained on personal data protection from December 2019 to 
February 2020.34 However, the cumulative data on the number of employees and representatives 
trained on good governance have not been reported regularly in GAI quarterly reports
35 and available 
data are not disaggregated by sex, limiting the team’s ability to analyze and validate achievements 
against targets.
36
2.1.2 IR1. INCREASED PUBLIC PARTICIPATION IN LOCAL GOVERNMENT DECISION-MAKING AND OVERSIGHT
Authorities from participating municipalities recognized that support from GAI has been 
effective in enhancing public participation, especially concerning the preparation of 
municipal budgets.
37 Namely, GAI’s efforts to develop capacities and create a culture of openness, 
engagement, and public information sharing contributed to an enabling environment for public 
participation in GAI municipalities.38
Interviews with representatives of LSGs confirmed the team’s initial findings from the desk review that 
GAI has provided support in the preparation, adoption, and implementation of local anti-corruption 
plans (LAPs).39 GAI has initiated capacity development of local independent bodies to monitor LAPs, 
through a basic training program that presented their role within the local anti-corruption framework. 
However, the members of these independent bodies highlighted the need to “enhance their capacities 
33 KII_02, KII_15
34 GAI Quarterly Report, February 6- April 30, 2020
35 The Q3 2019 report highlighted that “the target value for trainings in FY 2019 was set at 690 trainees and the 
number was surpassed, as GAI trained over 1,400 persons in 57 different sessions” The evaluation team was not 
able to review the Q3 2020 report, which included similar data.
36 GAI provides this data to the overall the US Mission Performance Plan and Report for Serbia. However, the 
evaluation team did not receive these documents during the desk review and was not able to include them in its 
analysis. 
37 KII_08, KII_14 
38 1
st Group included seven local governments (from November 2018): Sombor, Vranje, Vrnjacka Banja, Šabac, 
Dimitrovgrad, Raška and Sjenica; 2nd included six local governments (November 2019): Niš, Kragujevac, Novi 
Pazar, Sremska Mitrovica, Žabalj and Belgrade City Municipality of Stari Grad. 
39 KII_010, KII_014, KII_016, KII_020 and GAI Quarterly Reports
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 6
for effective monitoring and reporting on LAP implementation,” which is beyond GAI’s current and 
planned scope.40,41
The evaluation team found that GAI’s systemic support for enhancing transparency of LSG units has 
been well-defined and effectively implemented, achieving planned results. As indicated in the Theory of 
Change (TOC), GAI identified specific issues relevant to these LSGs to increase their transparency and 
accountability (as measured by the LTI). Website improvements in Raška and Vranje, along with the 
preparation of a strategic communication strategy for the city of Niš, are some of the most notable 
results.42 In the context of enhanced transparency and integrity, the partners recognized that “GAI 
played a key role in establishing systems and mechanisms for the protection of whistleblowers.”43 The 
partners from the City of Šabac, for example, recognized that GAI assistance resulted in the creation of 
a city-level whistleblower protection system with a designated, trained person in the city’s 
administration.44
Representatives of the participating LSGs stated that GAI has provided technical support, which 
contributed to improved public engagement in the local decision-making process. Stakeholders from the 
cities of Šabac and Sombor recognized that support from GAI enabled them to develop and adopt a 
model for youth participation in budgeting processes, increasing self-confidence, and resulting in youth 
becoming more actively involved in local affairs.45 The municipalities of Vrnjacka Banja and Sremska 
Mitrovica recognized that GAI assistance enhanced their abilities for the participatory planning and the 
budgeting processes. In this context, stakeholders stated that GAI has promoted the culture of 
openness, engagement, and information sharing on public spending.46 The municipal representatives 
recognized that support from GAI “enabled them to establish oversight and control mechanisms over 
public spending, meeting their needs to establish more transparent and effective functioning of local 
authorities.”47
2.1.3 IR2: INCREASED FOLLOW-UP TO OVERSIGHT INSTITUTIONS’ RECOMMENDATIONS
Under IR2, GAI’s interventions are aimed, in part, to increase the likelihood that oversight institutions’ 
recommendations to public bodies and the NA lead to concrete, meaningful change in the Serbian 
governance environment. However, the evaluation finds that IR2 has been formulated narrowly and 
inadequately to capture the actual GAI efforts. By prioritizing only the follow-up to oversight 
institutions’ recommendations, IR2 did not include a large area of organizational development and 
strengthening the capacity of oversight institutions as the central part of GAI assistance.
48
Analyzed from a broader perspective, GAI is on track to successfully enhance oversight institutions, 
most notably, the evaluation team observed these achievements in relation to planned support to the 
40 KII_010, FGD_03
41 KII_010, FGD_03, KII_23
42 KII_24
43 KII_9, KII_24
44 The Rulebook on Internal Whistleblower Protection in the City Administration of the City of Šabac 
https://sabac.rs/wp-content/uploads/2020/06/pravilnik-o-postupku-unutrasnjeg-uzbunjivanja.doc 
45 KII_14, KII_9 and KII_19 and FGD_04
46 KII_08, KII_19, KII_14
47 FGD_01
48 Under institutional development, it is recommended to answer, “development for what,” “development for 
why,” and “development for whom.” “Capacity for development- new solutions to old problems,” Edited by: 
Sakiko Fukuda-Parr, Carlos Lopes and Khalid Malik, United Nations Development Programme, Earthscan 
Publications Ltd, London and Sterling, Virginia, 2004
7 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
State Audit Institution (SAI) and Commissioner for Information of Public Importance and Personal Data 
Protection (CPI). 
Implementation of SAI recommendations has enhanced, with support 
from GAI. The last SAI activity report, together with performance 
audit reports, was discussed in the relevant parliamentary committee 
and also in the plenary in June 2019.
49 Following these sessions, the 
National Assembly (NA) adopted a conclusion recommending the 
government to act within its powers to ensure the implementation of 
SAI recommendations.50 In this context, SAI has recognized the
assistance from GAI and enhanced its services, primarily performance 
audit capacities and sound risk assessment system to select public 
institutions (including LSGs) for audit. SAI management stated that 
support from GAI enabled them to assess whether public entities are achieving economic, efficient, and 
effective use of available resources. The number of external audit reports in 2019 reached a total of 444, 
exceeding by more than ten percent the number of reports that SAI released in 2018, which was slightly 
above 400.51,52 GAI also helped SAI establish 13 teams to conduct performance audits and followed 
sound risk assessment criteria for selecting institutions for audits (including LSGs).53,54 GAI also facilitates 
interaction between SAI and LSGs, ensuring appropriate post-audit assistance and implementation of 
recommendations on the most challenging topics (for example, public procurement process and 
implementation of program budgeting). 
Similarly, the evaluation team found that GAI facilitated communication between SAI and LSGs, including 
assistance to address the findings from the post-audit reports. SAI representatives stated that as a result 
of GAI assistance, some of the most common issues from audit reports have been clarified, and 
challenges solved, concluding that municipalities have been more effective in implementing 
recommendations from the audit reports.55
The Anti-Corruption Agency (ACA) recorded some improvements regarding the follow up on its 
recommendations. The ACA submitted in 2019 a total of 76 (compared to 45 in 2018) requests related 
to a failure to promptly submit asset declarations for misdemeanor proceedings. Based on the ACA’s 
work, misdemeanor courts issued 65 convictions in 2019,compared to 49 in 2018.56 Assistance from 
GAI has been relevant, resulting in prioritization criteria for selecting declarations for annual review and 
a proposed methodology to review assets and income declaration statements, thus prioritizing the 
49 http://www.parlament.gov.rs/aktivnosti/kalendar.998.html
50 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf, page 118
51 SAI 2019- Annual Activity Report, http://www.dri.rs/documents/annual-activity-reports.181.html
52 SAI 2018 Annual Report- available at http://www.dri.rs/php/document/download/1549/3. In 2014, it was only 135 
(SAI 2015 Annual Report- available at http://www.dri.rs/php/document/download/724/3) 
53 The risk assessments fall in three key areas: i) the calculation of salaries; ii) public procurement and iii) assumed 
obligations above approved appropriations.
54 SAI representatives stated that GAI technical assistance to introduce and implement performance audits, as well 
as the implementation of recommendations in the targeted institutions, generated financial benefit to the public 
budget.
55 KII_07. The additional reference could be SAI reports: http://www.dri.rs/revizije/izvestaji-o-reviziji.136.html and 
http://www.dri.rs/dokumenti/godisnji-izvestaji-o-radu.93.html. 
56 The ACA filed 25 criminal charges in 2019 (2018: 16) due to reasonable suspicion that a public official had not 
reported 27 assets or had given false information about assets intending to conceal the facts. Based on the agency’s 
previous work, 19 final judgments were issued.
“We have received critical 
support from GAI to develop 
capacities for performance 
audits. These capacities have 
contributed to ensure more 
than 800 million dinars of 
financial benefits through SAI 
audits.”
Interview with the 
representatives of SAI 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 8
public officials that submitted assets declarations.57 GAI also assisted in developing a prioritization 
framework scheme that ACA implemented in a limited fashion. Namely, ACA officials declared they 
would not be able to continue with the process due to “inadequate software support.”58 GAI has been 
working to improve cooperation between ACA and misdemeanor courts, focusing on harmonizing the 
review of plea agreements in misdemeanor filing. Although necessary, this effort has had a limited impact 
on ACA's service delivery capacities. In addition, GAI supported ACA to prepare and adopt its mid￾term strategic plan together with the operational plan for 2020. 
The results of the assistance to CPI could be challenging to interpret. The results of GAI efforts could 
be associated with increased number of citizens’ complaints. In 2019, a total of 4,321 complaints had
been submitted, compared to 3,444 in 2018, most of them concerning administrative silence. Most were 
well-founded and the authorities should have provided the information requested by citizens according 
to the Law on Access to Information of Public Importance. The Activity delivered two-fold assistance. 
First, GAI supported the organizational development of CPI through exposure to international best 
practices regarding access to information and personal data protection.59 Second, in partnership with 
CPI, GAI delivered awareness and educational programs to different stakeholders. These programs 
included, for example, training for local administrations and entities (for example, utility companies or 
museums) directly or indirectly funded by public budgets to implement and follow up on the new 
requirements deriving from the Law on Personal Data Protection. 
However, concerning the implementation of CPI’s recommendations, improvements have not been 
recorded and the majority of public authorities do not comply with the obligation and recommendations 
to provide data to CPI recommendations regarding citizens’ requests for information.
60 In 2019, for 
1,786, or 41 percent (compared to 1,889, or 55 percent, in 2018), of these well-founded cases, the 
authorities reported that they acted on CPI's request to provide the information to the complainants. 
Administrative enforcement of the decisions taken by CPI has yet to be ensured.
61
2.1.4 IR3: INCREASED ADJUDICATION OF CORRUPTION CASES
The changes brought by the Law on the Organization and Competences of State Organs in the Fight 
Against Organized Crime, Terrorism, and Corruption produced some results concerning adjudication of 
corruption cases. Although corruption remains an issue of concern, the implementation of 
the Law has decreased the number of finalized high-level corruption cases, when compared 
with the previous years.
62 In this context, the key informants recognized that "GAI played a 
critical role in implementing the new Law from its adoption."63 According to the national 
57 The Law on Prevention of Conflict of Interest in Discharge of Public Office proscribes an obligation to public 
officials to report their assets to the Republic Board for resolving the conflict of interest. Under "public officials", 
the Law includes all persons elected, nominated or appointed to organs of the Republic of Serbia, autonomous 
provinces, municipalities and towns, as well as to organs of public companies founded by the Republic, province, 
municipality or town.
58 KII_02
59 The workshop has been organized around the availability of and access to information of public importance 
during the time of COVID-19 pandemic, showing parallels between Serbia, the region and Europe. KII_015
60 The CPI recognized in 2019 that a total of 4,321 (in 2018: 3,444) of citizens’ complaints – most of them 
concerning administrative silence – were well-founded and the authorities should have provided the information 
requested by citizens according to the law on access to information of public importance. In 1,786 or 41 percent 
(2018: 1,889 or 55 percent) of these well-founded cases, the authorities reported that they acted the CPI's request 
to provide the information to the complainants. Administrative enforcement of the decisions taken by the CPI has 
yet to be ensured. Ref to EU Progress Report, 2020 and the CPI Progress Reports
61 Ref to EU Progress Report, 2020 and the CPI Progress Reports
62 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf, page 25 
63 KII_06
9 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
partners, "GAI provided capacity development support, including tools for judges, public prosecutors, 
and other legal specialists to implement the legal framework." The GAI-produced Practicum for 
Investigation and Adjudication of Corruption Cases and its accompanying training program were 
referenced by the partners. 
The partners stated that support from GAI enabled them to "integrate and use financial forensic 
expertise in [the] adjudication of corruption cases, misfeasance in public procurement, and case law 
harmonization.”64 All these capacity development efforts have contributed, according to the national 
partners, to more effective adjudication of corruption cases."65
GAI responded to the need for a comprehensive approach for the analysis of corruption-related data 
through the establishment of the Electronic Register of Corruption Cases (ERCC). The ERCC should 
“enable a unified interface for specialized anti-corruption court unit and PPO data.”
66 The procurement 
process for the ERCC has been launched. Once implemented, it is expected that this tool will improve 
the analysis of information from existing case management systems and the monitoring and adjudication 
of corruption cases.67
GAI successfully tested the model to develop the capacities of the students/young 
professionals on the main anti-corruption related topics. Under IR3, GAI implemented the Anti￾Corruption Legal Clinic at the Faculty of Law in Belgrade, as a single semester program for a select 
group of 20 students completing their final year of studies. The model tested through the Anti￾Corruption Legal Clinic successfully affected youth engagement in anti-corruption activities. Although it 
only included a limited number of participants, the program was highly effective because a competent 
group of future professionals completed it.68
2.2 EVALUATION QUESTION 2: LINKING THE ACTIVITY’S OBJECTIVES AND RESULTS
What links can be determined between progress towards the Activity’s objectives and the Activity’s 
results?
EQ2 seeks to identify the links between GAI’s progress towards its objectives and its results at its mid￾way point. Drawing on the evaluation team’s desk review and interviews with key informants, the 
evaluation team found key links between GAI’s main achievements and its possible effects at the policy 
(systemic) and institutional levels. GAI has been successful in ensuring progress toward the Activity’s 
objectives, specifically regarding its interventions to improve sector policies, enhance administrative 
cultures, and strengthen management practices.
2.2.1 IMPROVING SECTOR POLICIES
GAI successfully addressed accountability issues by setting the 
basis for policies that ultimately will increase government 
accountability at national and local levels. The evaluation team 
used the policy cycle model as an analytical tool to assess and justify the 
64 KII_06
65 KII
66 KII_029, KII_06
67 KII_029
68 KII_011
The Practicum for 
Investigation and Adjudication 
of Corruption Offences 
represents an initial step in 
promoting efficient and 
innovative methods in fighting 
corruption and adopting good 
practices through 
implementation.
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 10
links between the results of the Activity and its objective at the policy level.69
● POLICY DECISIONS. The evaluation team found that GAI, by codifying its knowledge and 
experience, was able to influence “definition and agenda setting” in the policy cycle.70 GAI confirmed 
that the achievement of a more accountable governance system is a complex, slow process, with 
challenges at different levels. While GAI has helped set the policy agenda, it has not resulted in any 
new policy decisions. A brief reflection on the capacities of stakeholders, for example, showed that 
participating representatives of LSGs and public enterprises have considerably different levels of skills 
and knowledge.71 This indicated the need to prioritize systematic follow-up and continued capacity 
development. While GAI offered some tools to address these differences, the stakeholders 
expressed a need for additional technical assistance.
72
GAI also raised to the sectoral policy agenda the need to further strengthen and institutionalize 
protection mechanism for whistleblowers, especially in LSGs and public enterprises.73 However,
some of identified challenges that affect whistleblower protections have been poorly understood.74
Additionally, there are limited capacities to implement the Law on whistleblower protection at the 
local level, a lack of appropriate models to establish internal whistleblowing systems, and a lack of 
complaints management applications. 
● POLICY DEVELOPMENT. The primary and secondary data collected by the evaluation team 
indicated that GAI used some solutions that were useful to enhance transparency and accountability. 
These tested and positive solutions could be codified and used to prepare specific sectoral policies. 
For example, GAI supported LSGs in Serbia to prepare internal acts, procedures, and regulations, to 
implement the new Law on Public Procurement. The Activity published recommendations based on 
international best practices to improve the internal rulebook on whistleblowing in LSGs.75 GAI also 
developed a draft model for cooperation between high school student parliaments and the city 
administration, describing roles, responsibilities, processes, and a timeframe to facilitate regular 
annual consultation with high schools.76
GAI supported ACA with the development of three model acts regulating donations, conflict of 
interest, and complaints. GAI also prepared two analytical documents as the basis for the unified and 
integrative ERCC.77 These analytical documents provided an overview of the existing case 
management systems (AVP and SAPO) to help members of the Working Group for the ERCC 
69 Reference to Guy Peters “Policy Making Cycle- steps and approaches,” from “Capacity for Policy- how can we 
do it better” (prepared by Tomislav Novovic, UNDP Serbia, 2010)
70 Priority setting is inherently complex and involves making choices about what issues are most pertinent and time 
sensitive. After identifying priority issues, government and other governance actors must set the agenda to address 
the issues. "Framing" priority issues involves defining a problem in terms that can mobilize political support to find 
a solution. More available at Jon Pierre, B. Guy Peters “Governance, Politics and the State,” Macmillan, USA, 2016
71 The interviews with different LSGs additionally confirmed some differences among LSGs regarding their 
capacities. 
72 KII with the national stakeholders KII_07, KII_06, KII_14. Also, details are provided under the part Decision on 
instruments and implementation. 
73 KII notes KII_15
74 KII notes KII_15, FGD_03
75 Recommendations for Rulebooks on Internal Whistleblowing Procedures for Local self-governments, 
https://www.odgovornavlast.rs/wp-content/uploads/2019/09/USAID-GAI-and-Pistaljka_Recommendations-for￾rulebooks-on-Whistleblowing-Framework-for-LSGs-1.pdf
76 KII notes KII_15
77 These documents are the Automated Case Management System for Courts (AVP) and the Software Application 
to the Prosecutor’s Office (SAPO) Database model and Data Dictionary along with a report on the compatibility 
of catalogues in both case management systems.
11 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
determine and select the appropriate set of data to be used in the ERCC database.78 The evaluation 
team found that the existence of a functional ERCC is inextricably linked with evidence-based 
policymaking. The partners stated that the ERCC “provides solid and reliable information on the 
effectiveness of implemented measures, while also enabling monitoring of activities.”79
Regarding the upcoming amendment of the Law on Free Access to Information, GAI prepared two 
analytical documents with recommendations for the government’s consideration. GAI prepared the 
Post-Mortem Analysis of Corruption Cases for judges and advisors from specialized anti-corruption 
court units.80 The team’s analysis of primary and secondary data confirmed that this document 
provided a comprehensive learning-oriented analysis of select corruption cases and a basis for 
evidence-based policy on capacity building and improvement for specialized anti-corruption court 
units and public prosecutor’s offices.81,82
● DECISION ON INSTRUMENTS AND IMPLEMENTATION. GAI has designed tools and 
instruments that, according to the stakeholders, facilitated policy implementation, including the 
following: 
a) The Performance Audit Policies and Procedures Manual outlines the methodology to conduct a 
performance audit, including audit design, creation and use of sampling, application of audit 
criteria, questionnaire design, content analysis, and data reliability during audits. 
b) The Plea Agreement Guidelines harmonizes ACA’s filings with the courts’ consideration of plea 
agreements.83
c) The Practicum for Investigation and Adjudication of Corruption Offences provides guidance and 
practical advice to investigate and prosecute criminal offenses that include elements of 
corruption.84 This publication represents an initial step in promoting efficient and innovative 
methods to fight corruption and adopting good practices through implementation.85 In addition, 
GAI prepared a Practical Guide and FAQs for LAPs.86
d) The Conflict of Interest Manual serves as a guide for officials to use in managing ethical standards 
and conflicts of interest in the workplace.
87
78 Members of the group included representatives of the Ministry of Justice (including representatives of the Justice 
Division of the MOJ), Public Prosecutor’s Office, the Court of Cassation; GAI experts facilitated meetings of the 
WG).
79 KII notes, especially KII_29, KII_06
80 Details are available at https://www.4digitsconsulting.com/post-mortem-analysis-of-anti-corruption-cases/
81 GAI Quarterly Report- Q1 2020, also KII_06, KII_12
82 Key informant interviews- KII_01 and KII_06, KII_12 
83 These guidelines should enable agreement between ACA and Misdemeanor Courts on the use of plea 
agreements to adjudicate misdemeanor case filings ACA presents to the courts.
84 Practicum for Investigation and Adjudication of Corruption Offences, prepared by Biljana Sinanović, Brankica 
Marić, Vlatko Božović, Dragana Stojadinović; Goran Nikolić; Jelena T. Jakovljević and Jelena Kostić, PhD￾https://www.odgovornavlast.rs/wp-content/uploads/2019/11/Practicum-for-Investigation-and-Adjudication-of￾Corruption-Offences-ENG.pdf
85 Some of the topics include: initial information and access to open data; financial investigation; analysis of money 
flow and accounts; and drafting and execution of an investigation plan (indictment and asset recovery).
86 http://www.birodi.rs/wp-content/uploads/2019/08/LAP-cesto-postavljana-pitanja-i-odgovori-2.pdf
87 Details available in GAI Quarterly Reports- for example, Q2-Year 3 (May 1 – July 31, 2020)
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 12
e) The Guide for Persons Authorized for Receiving and Investigating Whistleblowers Reports in Local Self￾Government includes the necessary information and instructions to organize an efficient internal 
whistle-blowing system in LSGs.88
f) Local Transparency and Accountability Instruments. GAI provided instruments to enhance 
transparency and accountability at the local level, such as a guide to improve the communication 
of LSGs on social networks or the Local Government Model Website.89
g) Tools for Implementing the Public Procurement Law. GAI is working to provide tools to implement 
the new Public Procurement Law and develop model internal acts for public procurement to 
align with this new law. 
h) Tools for Financing Youth-Selected Projects. GAI prepared tools to facilitate student engagement in 
the planning of the local budget, and for budget allocation and transfer of funds to finance the 
selected projects (cases from Sombor and Šabac).90
2.2.2 EFFECTS ON PARTICIPATING INSTITUTIONS
According to the key informants, GAI has influenced the overall administrative culture among 
the partner institutions through the application of tools and methods to enhance 
transparency and accountability. 
● GAI guided and supported participating LSGs to identify priority anti-corruption actions in their 
LAPs and establish Local Anti-Corruption Councils as independent public monitoring mechanisms, 
according to key project participants.91 GAI-supported LSGs have also established internal capacities 
for personal data protection and mechanisms for whistleblower protection. In terms of delivery of 
services, LSGs recognized they “enhanced their capacities (e.g., clarified issues, prepared templates 
and practiced on real-alike situations) to implement the new Law on Public Procurement.”92 Two 
GAI-supported LSGs (City of Sombor and City of Šabac) have established single administrative 
points, integrating several administrative procedures into one, thus, working towards more efficient 
service delivery. 
● SAI management and staff recognized critical assistance from GAI to identify strategic priorities and 
prepare a plan to improve the delivery of core services.93 The key informants recognized that 
support from GAI contributed to SAI’s organizational development primarily by introducing 
performance audits.94
● The Activity is working to strengthen the organizational capacities of ACA. However, political 
influences and the newly appointed leadership’s lack of willingness to implement ACA’s core 
mandate limits the evaluation team’s ability to assess the effects of GAI assistance on the 
88 Prirucnik za postupanje lica ovlašćenih za prijem informacija od uzbunjivaca u jedinicama lokalne samouprave, 
prepared by Milan Pavlica, Dragana Matović, Tom Divajn, Duško Kovacević, Vladimir Radomirović, Belgrade, 2019
89 https://www.transparentnost.org.rs/images/dokumenti_uz_vesti/LTI_2020_final_report_ENG.pdf
90 KII notes KII_14, KII_19, FGD_04
91 KII notes- LSGs and FGD_01, FGD_02, FGD_03
92 KII_014, KII_019, FGD_01, FGD_03
93 KII_07
94 KII_07
13 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
organizational development of ACA.95 Specifically, the evaluation team was not able to assess 
whether trained ACA employees contributed to improved ACA services. 
● The media’s increased attention to and reporting on subjects under the transparency and 
accountability framework is one of GAI’s indirect results. For example, GAI has brought the topics 
related to LAPs or performance audit-reports to the attention of the media in some participating 
LSGs.96
2.2.3 MANAGEMENT PRACTICES
The evaluation team’s desk review and primary data analysis indicates that GAI has been 
implemented efficiently and that GAI has exercised adaptive management. Partners stated 
that GAI has demonstrated flexibility throughout implementation.
97 Management mechanisms, 
including implementation modalities, organizational structure, and the role of each partner under GAI’s 
components have facilitated GAI’s ability to achieve planned results.98 Also, the timely implementation of 
activities and the availability of financial and human resources to implement these activities have 
contributed to overall efficiency. 
Implementing partners appreciate the GAI team’s flexibility and limited bureaucracy.99 Partners indicated 
that GAI has dealt with their requests in a timely manner, and the relationship with GAI management 
has been based on mutual trust and respect.100
GAI has established a clear work plan and is on target to complete the Activity on time 
and within the approved budget.101 GAI’s work plans included results-oriented annual plans (and 
more detailed quarterly plans), providing a sound basis for scheduling, resource allocation, budget 
control, and IR attainment. The evaluation team found that the work plan development process has 
been participatory, through the joint work of the GAI team. The partners stated that the GAI work plan 
also served to coordinate activities, especially highlighting its importance as a tool for coordinating 
efforts related to capacity development and training.102
The main factor that contributed to GAI’s flexible and agile response during the implementation is the 
proactive role that GAI management and technical staff played in responding to the changes, challenges, 
and emerging priorities under the Activity’s three IRs. This, combined with the established partnerships 
with the main stakeholders, created an efficient implementation process. The partners highlighted that 
GAI’s responsiveness and flexibility during implementation is attributed to its “proactive and swift 
decision making based on needs of” participants.103
The financial and human resources for the implementation of GAI have been well-planned 
and available, according to the work plans. At the operational level, GAI established an Activity 
95 Different analytical reports from the Freedom House- Nations in Transit 2020, 
https://freedomhouse.org/country/serbia/nations-transit/2020 to EU Progress Report for 2020, indicate the 
challenging situation of ACA, with no autonomy in performance of ACA. 
96 KII_10
97 KII_02, and other KIIs with national partners 
98 KII notes KII_01.
99 KII notes 
100 KII notes 
101 KII notes KII_01 
102 KII notes with the national stakeholders 
103 The training program was modified to online delivery during the COVID-19 pandemic. Following protection 
measures, GAI prepared online programs for all subjects planned during the last quarters of the project. Delivering 
online training programs is associated with numerous challenges, from the availability of quality internet 
connections to ensuring active participation. Still, the participants have been active, expressing positive opinions 
about the approach, and have received trainings.
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 14
team with the Chief of Party (COP) as its head. The COP's prime responsibility has been to plan 
efficiently and deliver results, including to ensure that deliverables are of a high quality, and delivered 
within the approved work plan and budget. The COP also serves as the technical expert for the 
Activity’s component under IR2, related to supporting IOIs. The team included component managers to 
oversee progress towards intermediate results. Each component has its own staff to provide technical 
and on-demand support during implementation. The Activity, through the COP, facilitates constant 
communication and information exchange regarding strategic priorities with the partners and authorities 
at the national and local level. GAI did not originally plan to establish a steering committee or advisory 
structure. However, the existence of a high-level steering mechanism could be beneficial to set 
institutional and sectoral priorities, while also facilitating commitment and ownership over results. 
GAI established a monitoring system and a team responsible for data-collection and 
analysis. This monitoring system served, to a certain extent, as the reference during 
reporting. The Monitoring, Evaluation, and Learning Plan (MELP) set the performance monitoring 
framework. It includes tailor-made data collection protocols and instruments. The evaluation team found 
that the adopted indicators and the overall MELP have served to facilitate tracking of performance under 
all Activity components and overall progress reporting, despite some weaknesses (elaborated under 
EQ1). The indicators presented in the MELP have been mostly satisfactory; still, the evaluation team 
found gaps within the existing indicators. The existing indicators were not gender sensitive. Also, they 
could not capture progress under some important GAI areas. For example, indicators could not 
adequately measure GAI’s performance in the areas, such as right to information, civic participation, and 
complaint mechanisms. GAI did not introduce appropriate indicators to measure development of 
capacities of the target groups or the effects on anti-corruption education programs on the selected 
youth (e.g., Anti-Corruption Legal Clinic and its participants). Also, GAI did not measure the effects of 
awareness campaigns and assess if the targeted audience has enhanced understanding of a specific topic. 
The MELP also did not include more specific indicators related to CPI’s assistance to LSGs.
104 The 
proposed means of verification and data sources provided were appropriate to validate progress and 
achievements. 
GAI has established a regular reporting practice, but results-oriented reporting has been 
limited. Concerning reporting frequency, GAI prepares regular quarterly reports, providing a 
substantive presentation of implemented activities, and presenting plans for the next quarter. However, 
results-oriented reporting and references to the MELP indicators to measure the progress towards the 
achievement of targets in the quarterly reports are limited.105
2.3 EVALUATION QUESTION 3: SCOPE OF WORK EXPANSION
What other areas of interventions could fit under the SOW beyond what is already planned?
Building on the evaluation team’s assessment of GAI’s current SOW, through the findings under EQ3 
and the desk review and interviews with key informants, the team sought to identify additional areas of 
interest that would complement GAI’s intervention to further achieve the Activity’s desired results and 
promote its sustainability and scalability. 
2.3.1 ASSESSMENT OF THE CURRENT SOW
GAI fully aligns with governance priorities related to transparency, accountability, and the 
prevention of corruption in Serbia. GAI is well-aligned with the 2013-2018 National Anti￾104 The MELP included indicator IR2.1.5. Number of USG-supported anti-corruption campaigns and did not provide 
other indicators to measure effects of the assistance to CPI and LSG. Still, the LTI could serve to some extent as 
proxy indicator to measure results at this level 
105 Based on GAI quarterly reports
15 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
Corruption Strategy (NACS), its Action Plan (AP), and its general objective “to eliminate corruption, as 
an obstacle to the economic, social and democratic development of the Republic of Serbia.” GAI's focus 
on enhancing government accountability at the local and national levels through the capacity 
development of key stakeholders and strengthening their connections is instrumental in implementing 
the NACS objective to “ensure good governance through the inter-institutional partnerships as the 
vehicle for its proper and effective implementation.” 
There are clear links between GAI's areas of intervention and the NACS AP, with its focus on the most 
vulnerable areas, including strengthened participation, oversight mechanisms, and improvement of 
investigation capacity, among others.106 GAI is relevant for the implementation of Chapter 23 of the AP, 
concerning justice, fundamental rights, and the fight against corruption. It reflects all mitigation measures 
recommended through the European Commission (EC) Screening Report.107 The relevant priorities for 
GAI are: i) clarify the cooperation between the stakeholders in charge of implementing anti-corruption 
measures; ii) clarify the mandate of ACA (including on a legal basis) to ensure that the staffing level 
matches the tasks; iii) improve free access to information rules and their practical implementation; iv) 
strengthen transparency and integrity of governing structures, including through strengthening internal 
control and audit bodies; v) adopt and implement the law on whistle-blowing and take the necessary 
steps to make the whistleblower protection system more effective in practice; and, vi) ensure 
independent, effective, and specialized investigation/prosecution bodies.108
GAI is aligned with the action point under Chapter 23 that envisaged “preparation of LAPs,” to identify 
corruption-prone areas and processes and define preventative and sanctioning mechanisms.109 Chapter 
23 stated that “by implementing LAPs and establishing adequate mechanisms for their enforcement, local
governments will enhance their adherence to good governance principles.”110
The partners were positive about the Activity’s responsiveness and involvement from 
design and planning to implementation, while re-stating that GAI provided the requested 
technical assistance. The Activity design process was consistent, based on a sound programming and 
formulation approach.111 The Activity identified problems using a bottom-up process of collaboration, 
and involved policy makers and representatives for the Activity’s main partners.112 GAI’s design of 
technical assistance and transfer of knowledge addressed the major challenges facing the independent 
regulatory bodies, including their limited technical capacities to deliver some of their core services, and
106 National Strategy for Fight Against Corruption for the period 2013-2018, more at: http://www.acas.rs/wp￾content/uploads/2010/06/Nacionalna_strategija_za_borbu_protiv_korupcije.pdf
107 https://www.mpravde.gov.rs/tekst/30402/revidirani-akcioni-plan-za-poglavlje-23-i-strategija-razvoja-pravosudja￾za-period-2020-2025-22072020.php
108 Ref to Chapter 23 - https://www.mpravde.gov.rs/tekst/30402/revidirani-akcioni-plan-za-poglavlje-23-i-strategija￾razvoja-pravosudja-za-period-2020-2025-22072020.php
109 Ibid.
110 Ibid.
111 KII notes KII_01, KII_02
112 KII_01 KII notes, national partners
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 16
exercise monitoring and oversight proactively and effectively.113,114,115 These challenges have continuously 
undermined mechanisms to hold the government accountable for the delivery of services.116
The core areas of GAI assistance respond to the needs for adoption, implementation, and 
monitoring of LAPs; improved transparency in local governance processes; enhanced civic 
participation; and provided the prerequisites for whistleblower protection and 
improvement of service delivery. The GAI objective, as well as its activities, are relevant to the 
needs of local authorities: the interviewed representatives stated that “GAI helps them to respond to 
citizens’ needs more effectively in line with law and assigned services.”117 The Activity adopted a tailor￾made approach, responding to the needs of each LSG and taking into consideration their unique 
characteristics.118 All municipality representatives who participated in the online survey administered as 
part of the evaluation recognized that GAI assistance is well-aligned with local governance reform 
priorities in Serbia: 50 percent of survey respondents reported a complete alignment and 50 percent 
reported a sufficient alignment. 
The survey participants also indicated that the Activity addressed the needs of IOIs: the 
majority (95 percent) indicated that they positively regarded GAI’s efforts to address the 
needs of key Serbian government institutions to ensure greater accountability. The focus of 
GAI is to enhance the capacities of government institutions to deliver services that ensure oversight of 
the operations of other government bodies and hold them accountable for their work. However, 
despite improvements and support from GAI, these institutions lack many of the tools and technical 
capacities to fully and competently fulfil their mandates.119
The IOIs report annually to the NA on challenges, successes, and opportunities to improve 
governmental operations; however, a lack of follow up by the NA and the Government has prevented 
the enforcement of the oversight bodies’ findings. This situation is undermining the public’s perception 
of the oversight institutions’ work and is discrediting citizens' faith in solving corruption in the long run. 
The absence of follow-up on findings and recommendations contributed to a general lack of 
understanding of oversight institutions’ mandates. 
The representatives of the judiciary recognized that “GAI played a pivotal role in responding to the 
immediate need to enhance capacities of specialized anti-corruption prosecutors and court units, 
established with the adoption of the new Law on Organization and Jurisdictions of State Bodies in 
Charge of the Prevention of Organized Crime, Terrorism, and Corruption.”120 The Activity also 
113 For example, implementation of the new Law on Private Data Protection, implementation of performance 
audits for the State Audit Institution and follow up on recommendations, mandate of ACA. 
114 For example, ACA needed to improve risk-based oversight of asset and income declarations and build the 
institution’s analytical and research skills to enhance its performance.
115 The EC Progress report for 2018 highlighted that the Parliament’s practice not to discuss any annual reports by 
independent bodies in plenary sessions has continued from 2015. Very limited discussion and follow up to the SAI’s 
audit reports and a lack of parliamentary discussion on the implementation of major laws and policies have been 
also recorded. 
116 More details in the EC Progress Report 
117 KII_12, KII_21, KII_20 and focus group discussions
118 KII_01
119 The recent EC Progress report highlighted these challenges, also emphasizing other challenges: to ensure more 
effective cooperation and channels of communication (with government institutions) and enhance credibility and 
understanding of mandates and roles of these institutions.
120 KII_06
17 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
addressed the need to establish an integrative electronic system to register corruption cases and 
increase follow-up proceedings.121
GAI has remained well-aligned with governance related strategic priorities during its 
implementation thus far. The evaluation team assessed GAI’s relevance at all points during its 
lifecycle to test and analyze whether the intervention logic remained true in the context of possible 
changes on the demand side, and, if not, to assess the response made to adjust and fine-tune the 
intervention as set out in GAI’s intervention logic and its logical framework. The evaluation team has 
focused on whether the IRs and Activity objective, as the main references under the intervention logic, 
remained valid and achievable or whether there was a case for revision to take account of changes in 
the external environment, demand, or speed of delivery. 
The “real time” assessment of GAI’s relevance confirmed that the intervention logic remained true 
throughout its implementation, as justified in some of the key governance-related strategic and policy 
documents. The 2020 Serbia Progress Report issued by the EC recognized limited progress with the
implementation of anti-corruption measures.122 At the same time, the report highlighted that 
“corruption remains a pressing problem for the country,” together with insufficient transparency and 
accountability of public institutions.123
GAI under its components is addressing the identified needs and priorities of Activity 
stakeholders. The overall priority has been to work on “establishing an effective coordination 
mechanism to operationalize prevention policy goals.”124 At the institutional level, ACA is the key 
institution in the fight against corruption, and requires development of its capacities to “effectively 
implement the law on the prevention of corruption in line with the European Union (EU) acquis, 
international agreements and Council of Europe’s Group of States against Corruption (GRECO) 
recommendations.”125 In addition, IOIs require “systematic support for their work, to regularly monitor 
the implementation of their recommendations and hold the executive bodies accountable.” Their 
priority is to “improve data management, including registration of corruption cases, considering different 
units and stages.”126 At the local level, the need remains to ensure tangible improvements in relation to 
anti-corruption efforts and enforcement of implementation of the LAPs.127 The revised Action Plan for 
Chapter 23 also restated these priorities, proposing more actionable measures and indicators of 
success.128
The stakeholders and online survey participants that the evaluation team interviewed recognized 
USAID’s “substantive insight and adequate approach to development needs and challenges in the 
governance area” that contributed to the lasting relevance of GAI.129 Furthermore, achieved results and 
121 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
122 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
123 European Commission- Serbia 2020 Report, more at https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
124 Ibid., European Commission- Serbia 2020 Report
125 Ibid., European Commission- Serbia 2020 Report, and also https://www.coe.int/en/web/greco/evaluations/serbia
126 https://www.coe.int/en/web/greco/evaluations/serbia
127 The report recognized that “out of the 102 local self-governments that adopted anti-corruption plans, 28 
established a body to monitor their implementation, mostly in line with the Anti-corruption Agency’s model.” 
However, the actual benefits achieved through implementation of LAPs could not be assessed.
128 https://www.mpravde.gov.rs/en/tekst/30881/booklet-on-project-achievements-eu-for-justice-support-for￾chapter-23.php
129 KII notes KII_15 and KII_07
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 18
progress in the specific areas of the intervention are inherently linked to the “enhanced relevance of 
GAI and its intervention.”130
GAI has accurately identified components and clearly established areas of intervention 
under these components. GAI’s already tested services need to continue, with a focus on 
root causes of challenges. The scope of GAI is well defined, with adequately identified 
priorities and areas of intervention. The prevailing opinion among the stakeholders is that GAI 
should “continue with the assistance/activities;” stakeholders state that “models and packages have 
already been developed and tested, effectively producing visible and concrete results.”131 The 
interviewed partners, namely those holding senior level positions, and survey participants, recognized 
that “GAI has confirmed its validity to the needs of respective institutions and LSGs in Serbia,” also 
proposing to “further fine-tune and focus support to address root causes of challenges.”132
The participants in the online survey confirmed the relevance of GAI and the need to continue with its 
activities, prioritizing capacity development, inter-institutional cooperation for anti-corruption, and the 
improvement of delivery of services at the local level.133 Survey participants suggested the following areas 
of proposed support, among others: capacity development or training, mentoring assistance aimed at 
improving service delivery, and developing specific anti-corruption guidebooks and other tools.
The broad area of capacity development included proposals for “organization of priority training 
programs and workshops to improve good governance related capacities for holding the government 
accountable at the local and national level” and “further strengthen the capacity of specialized units, 
prosecutors, and judges in the areas of anti-corruption.” Survey participants recognized mentoring 
support as an important capacity development tool that should be used for the preparation of LAPs, 
establishment of public participation practices, and more transparent budgeting processes at the local 
level.134 In addition to strengthening cooperation among IOIs, survey participants highlighted the need to 
increase cooperation with the civil sector and media on specific anti-corruption topics. Improvement of 
the delivery of services, especially at the local level, should continue, with advancement of e-government 
opportunities, simplification of procedures, and more citizens-oriented processes.
2.3.2 ADDITIONAL AREAS OF INTERVENTION THAT COULD FIT UNDER THE SOW
The evaluation team used the data gathered from KIIs, FGDs, the online survey, GAI documentation, 
other sources, and analytical documents to identify possible areas of interventions that could fit under 
GAI’s SOW, beyond already planned activities.135,136 The evaluation team presents these additional areas 
by levels of intervention: systemic, institutional, and individual. 
SYSTEMIC LEVEL
● ENGAGE IN POLICY MAKING. Stakeholders from the oversight institutions recognized that GAI 
could engage more actively in the policy making process related to anti-corruption, transparency, 
and accountability. According to these key informants, the expertise and experience gained from 
130 KII_07
131 KII notes KII_16, KII_15, KII_07 
132 KII notes- partners 
133 Results from the online survey, prepared and analyzed by the evaluation team.
134 Results from the online survey
135 The revised Action Plan for Chapter 23, the annual reports of the National Assembly of Serbia, and IOIs.
136 The main references are the Serbia 2020 Progress Report of the European Commission, The Freedom House 
Report on Serbia, 2020, and the Council of Europe analytical reports on Serbia. 
19 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
GAI could be beneficial to prepare and adopt a “new anti-corruption strategy together with a 
credible and realistic action plan.”137
● SUPPORT OVERSIGHT INSTITUTIONS DURING AND AFTER COVID-19 PANDEMIC.138
Stakeholders recognized that GAI could provide important support to IOIs to respond to the 
corruption risks and challenges posed by COVID-19.139 Oversight is an issue of relevance, including 
access to information of public interest, protection of personal data, and also the protection of 
whistleblowers (e.g., if issues arise related to the procurement of medical equipment and supplies).140
INSTITUTIONAL LEVEL
● EXPAND SUPPORT TO LOCAL AUTHORITIES. GAI’s support to local authorities should 
continue and expand. For example, GAI provides specific assistance to LSGs in designing sound and 
realistic monitoring systems for their LAPs. In parallel, there is a need to enhance capacities of local 
independent anti-corruption bodies in charge of LAP monitoring. GAI could further support these 
bodies to assess progress, analyze results, highlight challenges, propose areas for additional 
attention, or promote best practices.141
The local level partners and FGD participants recognized GAI as an important support for 
preparation of local communication strategies and to improve communication. At the same time, 
they observed insufficient capacities for strategic communication especially at the level of its 
leadership and senior level officials. Considering GAI’s achieved results, GAI is well-positioned, 
according to them, to address these needs and facilitate development of communication skills at the 
local level.142
In some supported LSGs, GAI assisted with the analysis of finances and the budget implementation 
process, preparing plans with recommendations for improvement. However, there is no evidence 
that LSGs are progressing with implementation of these plans. GAI could build on existing support 
to budgeting activities and provide next-level assistance, focusing on implementation of 
recommendations and monitoring the entire improvement process. 
● ENGAGE WITH PUBLIC AND PRIVATE COMPANIES. The need for oversight agency 
engagement with the public and private companies is highlighted as an issue in the most-recent EU 
country progress report.143 The mentioned areas of potential support include strengthening 
capacities to manage public interest data and building mechanisms and capacities to protect 
whistleblowers, among others.144 The partners recognized that GAI could support inclusion of 
private sector interests into good governance and anti-corruption priorities for the country. 
According to partners, GAI could also engage with public companies, including public utility 
137 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
138 The COVID-19 outbreak increases corruption risks. The health sector is particularly exposed, because of the 
simplified procurement rules as a response to the immediate need for medical supplies, non-transparent spending 
of public funds, overcrowded medical facilities, and overburdened medical staff. As such, potential corruption and 
procurement issues, including bribery in medical-related services, conflicts of interest, and lobbying are possible.
139 KII_12, KII_07 and KII_15
140 Corruption Risks and Useful Legal References in the context of COVID-19, issued by Mr. Marin Mrčela, 
GRECO President, Strasbourg, 15 April 2020, GRECO (2020)
141 Based on the analysis from the EC Progress Report for Serbia, 2020- https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
142 FGD_01, FGD_02, FGD_03, KII_25
143 EC Progress Report for Serbia, 2020. https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
144 KII_15 and EC analysis 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 20
enterprises and other direct public budget users. Initial activities could include establishing and 
strengthening whistleblower protection mechanisms for the private sector, which would enable 
employees to report corruption.
● STRENGTHEN NATIONAL ASSEMBLY. The main analytical documents and information from 
the independent institutions showed the need to strengthen the capacities of the NA and its 
relevant committees to read, understand, and comment on the oversight institution reports, and 
present findings and recommendations during plenary sessions.145 The newly constituted Serbian 
parliament is dominated by an overwhelming majority of the ruling coalition and the absence of a 
viable opposition, a situation which is not conducive to political pluralism in the country. 
Nevertheless, members of the NA and especially members of the main committees would benefit 
from capacity development in the core areas of their work to exercise their functions more 
competently.146
● STRENGTHEN PROSECUTORS. The recent EC Progress Report for Serbia and other references 
on good governance in Serbia show the need to strengthen the technical capacity of prosecutors.147
GAI could build on the positive results so far, including the Manual for Whistleblower Protection, 
the exposure to best practices, and experience exchange, by facilitating the visit of a career 
prosecutor from the U.S. GAI could consider intensifying these efforts, by presenting, for example, 
the American experience on following up with and acting on whistleblower reports. Particularly, 
GAI could provide technical assistance to prosecutors regarding implementation of the relevant anti￾corruption laws. 
● STRENGTHEN DATA PROTECTION OFFICERS. GAI could explore opportunities to strengthen 
the network of data protection officers (DPOs) according to CPI.148 A total of 3,000 DPOs from 
public institutions and private enterprises have been engaged in Serbia.149 There is a need to provide 
sound capacity development support to the DPOs. GAI could work with CPI to prepare an 
interactive online platform with basic and advanced training for DPOs. 
INDIVIDUAL LEVEL
● OFFER LEARNING OPPORTUNITIES TO YOUTH. The partners observed that youth activism, 
through participation in local planning and decision-making processes, has been one of GAI’s 
important areas of support, and according to them, the experience from the City of Sombor could 
be beneficial to other participating LSGs.150 The partners also shared that the implementation of the 
Anti-Corruption Legal Clinic is a format that could be built upon.151
● OFFER LEARNING OPPORTUNITIES TO JOURNALISTS. GAI could explore opportunities to 
develop the capacity of journalists to better understand the anti-corruption framework and audit 
reports, including performance audit reports and clarify issues of personal data protection, especially 
in media. There is a need for local media to oversee the implementation of LAPs as well.
145 EC Progress Report for Serbia, 2020. https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
146 The main committees relevant to GAI oversight are the Committee on Constitutional and Legislative Issues, the 
Committee on Finance, State Budget and Control of Public Spending, and the Committee on the Judiciary, Public 
Administration and Local Self-Government.
147 https://www.coe.int/en/web/greco/evaluations/serbia
148 KII_15
149 All large companies must appoint a representative for Serbia for personal data protection. Some of the large 
companies such as Google, Viber, now Facebook have started this process. 
150 KII_14, KII_08, FGD_04
151 KII_011; FGD_04 
21 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
● RAISE AWARENESS. Citizens’ participation remains critical for transparent and accountable local 
governance. GAI’s primary focus has been on the capacities of local authorities to establish systems 
and mechanisms for citizens’ participation. However, GAI could encourage citizens to take a 
broader interest, and increase their willingness to participate in local decision-making processes. It 
will be especially important to engage the public in anti-corruption efforts and activities at the local 
level. Youth activation remains highly important. GAI could work to build awareness and interest by 
increasing youth’s understanding of their rights and encourage their participation in local processes. 
2.3.3 FLEXIBILITY TO CHANGING NEEDS AND PRIORITIES
GAI has been flexible in responding to challenges that have occurred during its 
implementation to date, particularly to the emerging priorities of the participating 
institutions. The GAI implementing environment is challenging and is characterized by limited progress 
in the fight against corruption. The analytical reports on the governance situation in Serbia highlight 
some of the systemic difficulties, including the weakening of institutions, and the executive branch’s 
interference with institutional mandates, which compromise the institutions’ fragile independence.152
There is an emphasis on the need to strengthen the capacities of ACA to implement its mandate. 
However, ACA’s implementation of the new anti-corruption law is lagging behind due to limited 
commitment of the new management and declining capacities of ACA.153 ACA must fully acknowledge 
the role of independent bodies and follow up on their recommendations promptly.154 The changes 
brought by the law on prevention of corruption (in force since March 2018), produced some results in 
increasing the number of closed cases. However, the 2020 EU progress report concludes that 
"corruption remains an issue of concern,” and there is still “no effective prevention coordination 
mechanism in place.”155
The GAI work plans are critical to ensuring timely and successful implementation of GAI activities.156 In 
addition, the Activity’s mitigation strategy focused on delivering technical level activities, ensuring that 
strategic level decisions are aligned with national priorities and identified needs.
157 Nevertheless, the 
Activity is responsive and flexible to the changes in the external environment within the agreed plans 
and implementation framework. For example, GAI demonstrated this flexibility in its response to the 
limited commitment of the leadership of ACA. GAI adjusted its capacity development approach, helping 
with strategic planning and development of analytical capacities.158 Also, GAI’s mechanism to assess the 
effectiveness of the training programs enabled GAI to “institutionalize” flexibility. Practically, GAI 
responded to the comments and reactions from trainees and adjusted its delivery approach to address 
needs.159 In addition, GAI responded to the COVID-19 pandemic effectively, adjusting its training 
programs to an e-learning format. Within a short period of time, GAI trainers and instructional 
designers reconfigured in-person classroom training programs for online learning.160 Implementing 
partners appreciated GAI’s level of flexibility and limited bureaucracy.161 Partners have indicated that 
they generally have good relations with GAI, the COP, and component heads. GAI responded to 
152 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf and 
https://freedomhouse.org/country/serbia/nations-transit/2020
153 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
154 European Commission- Serbia 2020 Report, https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
155 Ibidem, European Commission- Serbia 2020 Report
156 Notes from KII_01
157 Notes from KII_ 01, project progress reports also reflected on these changes and reconfirmed the approach.
158 GAI Progress Reports and KII_01
159 KII_07, KII_15 and KII with representatives of LSGs. The results of the survey also indicated appropriateness. 
160 KII notes- KII_01, KII_07, KII_15
161 KIIs notes 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 22
requests from partners in a timely manner and cultivated relationships based on mutual trust and 
respect.162
2.4 EVALUATION QUESTION 4: FUTURE PROGRAMMING
What are the interventions with greatest traction, unintended outcomes, and/or lessons learned from 
USAID assistance that the Mission can use in decision making for future programming?
In response to EQ4, the evaluation team analyzed primary and secondary data sources to identify GAI’s 
interventions that demonstrated the greatest traction to date and those that resulted in unintended 
outcomes, as well as to document lessons learned. 
2.4.1 INTERVENTIONS WITH GREATEST TRACTION
GAI has consistently responded to Serbia’s national capacity development needs and is 
steadily strengthening national and local level capacity to improve government 
accountability. Future progress and its sustainability, however, rests on the continued 
commitment and capacity of Serbia’s governance system, institutions, and citizens. GAI has 
provided capacity development assistance to employees and stakeholders involved in governing at the 
local and national levels.163 Of the surveyed employees and stakeholders who participated in GAI’s 
intervention, all stated they benefited from this assistance.164
The partners recognized that GAI has provided "timely, effective and highly demanded” technical 
assistance to the specialized anti-corruption court units, in line with the law on prevention of 
corruption.165 In addition to improvements in case management, monitoring, and adjudication of 
corruption cases, through the functioning of the GAI-supported ERCC system, these results and 
capacities should facilitate future evidence-based policy and decision making in the sectors of concern.166
KIIs with IOI stakeholders suggest that assistance available through GAI has been instrumental in 
providing demanded technical assistance to these institutions to establish capacities and systems to 
better deliver their core services.167 Still, the sustainability of progress relies on the continued genuine 
commitment of management teams to implement assigned functions. Authorities in Serbia must
recognize the role of independent and regulatory bodies, strengthen their capacities, including their 
budgets and skilled staff, and guarantee their functional independence in practice.168 There are still 
challenges, such as dominance and influence of the executive branch over the work of IOIs, slow 
implementation of anti-corruption laws, politicized leadership of ACA, among others. These and other 
factors prevent Serbia from promoting and ensuring credible, professional institutions that follow the 
principles of transparency and accountability, which may present possible implications for 
sustainability.169
162 KIIs notes 
163 A total of 81 percent of the participants in the survey participated in the trainings that GAI organized. 
164 According to the results of the online survey, 93 percent of the participants stated they benefited from the 
capacity development support from GAI
165 The Law on the Organization and Competences of State Organs in Fight Against Organized Crime, Terrorism 
and Corruption, adopted by the National Assembly in 2016, and then came into force in March 2018.
166 KII_29 and KII_06
167 KII_15, KII_07, KII_06. These services include performance audits and follow up on recommendations (SAI), 
systematic protection of personal data (CPI) or prosecution of corruption cases
168 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
169 Ref to progress under the Chapter 23, in the report https://ec.europa.eu/neighbourhood￾enlargement/sites/near/files/serbia_report_2020.pdf
23 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
GAI has been working to enhance policy processes by designing instruments and tools and implementing 
technical assistance to enhance transparency and accountability at the national and local levels. Further 
details are provided under Section 2.2.1. of this report. These efforts have been important and indicate 
that progress is likely to be achieved in some areas, including more transparent and accountable local 
governance, improved state audit practices, and strengthened frameworks for implementation of 
personal data protection. However, there are still challenges to ensuring the successful implementation 
of the policies, largely the coordination of horizontal and vertical policy. There are also remaining 
impediments to translate policies into actions, connecting adequately with public funds and ultimately 
monitoring and reporting on progress.170 The lack of country and sector specific targets and indicators 
under the anti-corruption framework and absence of regular monitoring and reporting mechanisms have 
potentially limited GAI’s progress to date and may affect the sustainability of GAI’s interventions.
GAI has remained a valuable partner in supporting the achievement of transparent and 
accountable governance in Serbia, establishing strong and effective partnerships with the 
national institutions. Stakeholders from partner organizations recognized “strong comparative 
advantages” of GAI, stating that GAI (and USAID) remain “an important, credible and widely accepted 
partner in supporting transparency and accountability of the governing institutions” and contributing to 
“achieving priorities for the establishment of functional system to prevent and punish corruption in 
Serbia.”171 The backbone of GAI’s comparative advantage is its technical capacities and profound 
understanding of the organization-specific constraints and development needs, as well as its proven 
focus on the needs and priorities of national institutions.172 In addition, “USAID through GAI and 
previous interventions in the area of building transparent and accountable institutions in Serbia 
demonstrated strong abilities to establish and maintain effective partnerships based on trust, 
responsiveness, and mutual respect.”173 Stakeholders from participating organizations recognized the 
value of GAI’s technical support and enhanced efforts to ensure compliance with anti-corruption laws in 
Serbia, EU accession requirements (derived from Chapter 23), and other international standards.174
GAI’s demand-driven approach and strong national participation in all GAI activities has 
additionally contributed to a strong sense of stakeholders’ ownership of the achieved 
results.175 National partners have been involved in conceptualizing and implementing activities under 
GAI, through forums for example.176,177 GAI has employed a consultative and inclusive approach to 
formulating work plans and Activity design, ensuring stakeholders' involvement, contributing to an 
increased sense of local ownership and a foundation to promote the sustainability of results. The 
favorable environment for the Activity’s implementation can be attributed to long-lasting and mostly 
positive experience with USAID’s support to the establishment of good governance in Serbia. 
Likewise, GAI followed a participative partnership-building approach throughout the implementation of 
activities. National partners have been involved in the activities, either as participants, implementers, or 
members of different forums, to create awareness of the achievements and results in their respective 
areas of work. At the same time, the evaluation team found, through KIIs with key stakeholders, that 
170 OECD-SIGMA “Monitoring Report: The Principles of Public Administration-Serbia,” May 2019, more 
http://www.sigmaweb.org/publications/Monitoring-Report-2019-Serbia.pdf
171 KII notes- national partners
172 KII notes- national partners
173 Notes from KII_02, KII_03, KII_07 and KII_08
174 Notes from KII_07, KII_14, and KII_08. 
175 Notes from KII_02 and the results presented under the previous part of this report.
176KII_15. CPI and GAI have jointly developed and implemented a program on personal data protection for local 
governments 
177 SAI Management has been participating in different activities for its organizational strengthening; at the same 
time, SAI was involved in delivery of mentoring support to LSGs and budget users. 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 24
some stakeholders have been familiar with GAI components only directly related to them, with limited 
awareness of the other areas of the Activity’s intervention.178 This situation has been somewhat different 
among the senior-level representatives of partner organizations, who confirmed a more in-depth 
understanding of the Activity and its scope and results.179
2.4.2 UNINTENDED RESULTS
GAI has produced important, yet unintended, results in different areas of its intervention, 
including increased communication, coordination, and cooperation among LSGs and 
national governing institutions. The evaluation team lists these unintended results below:
● Delivery of training programs addressed knowledge gaps and enhanced capacities and skills among 
the staff from participating LSGs. In addition, the trainings favored horizontal knowledge 
sharing amongst the employees from different LSGs, helping employees to connect and unify 
their understanding and application of procedures. 
● Similarly, GAI has helped SAI to establish effective, although informal, communication 
with participating LSGs. During GAI events, the SAI management team participated as guest 
speakers and agreed on regular communication channels with LSGs.180 According to SAI 
representatives, this channel of communication has enabled LSGs to resolve some of the critical 
audit-related findings and recommendations and improve the internal management processes.181
● GAI provided comments and interpretations on the relevant policies and laws, for 
example, the Public Procurement Law, the Law on Personal Data Protection, and the Law on 
Whistleblower Protection. GAI prepared guidelines or manuals that were broadly disseminated and 
shared with other LSGs.
● Stakeholders, with support from GAI, have been exposed to international best 
practices and models tested in the United States. Besides contributing to the improvement 
of professional capacities, these events created additional positive attitudes and enhanced Serbian 
public institutions’ commitment and readiness to implement practices for greater transparency and 
accountability.182
● The Activity has been active in advising and advocating for transparency and accountability at 
different levels. The active role of GAI staff in implementing activities and building 
partnerships generated additional readiness and willingness among participating 
institutions to propose reviews of legal and policy documents to better reflect 
principles of transparency and accountability. Partners recognized that GAI provided 
important support to introduce participatory budgeting in the city of Sombor. These efforts included 
increasing youth participation to select priority initiatives for their schools and communities. 
According to the partners, this support resulted in improved youth civil activism in the City of 
Sombor, creating a favorable environment and developing the capacities of youth to explain their 
needs, to make questions and demands, and to become more socially active.183
178 Refer to KII notes with the national partners
179 The senior level representatives of the partner organizations have participated in GAI programming phase and 
have been involved in discussions during the conceptualization of GAI. They have also been involved in activities 
under other components. For example, SAI management took part in the workshops with LSGs, took part in GAI 
workshops and joined other events that enabled them to grasp the full scope of the Activity. 
180 KII_07
181 KII_07
182 FGD_01, FGD_02, FGD_03, KII_08, KII_14
183 FGD_04
25 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
2.4.3 BEST PRACTICES
While the team’s priority in conducting the evaluation was to analyze GAI’s expected impact and 
sustainability prospects, the team identified best practices from GAI’s implementation based on several 
criteria, such as the effects on participating institutions, the durability of established mechanisms, and the 
partnerships. This section presents the following best practices: using the LTI, engaging youth in 
budgeting processes, supporting SAI, and strengthening cooperation between CPI and LSGs.
USING THE LTI. According to the representatives of participating LSGs and based on primary data 
sources, the evaluation team found that the LTI, together with the scoring system and GAI’s technical 
assistance, have contributed to long-lasting improvements and greater transparency.184,185,186 Based on 
LTI, the average score for 2020 reached a total of 46 out of 100, improving from 40 in 2019. The main 
advances are in the areas of publishing budget information and developing and publishing local anti￾corruption action plans.187 The representatives of the participating LSGs stated that GAI provided needs￾based support to them and directly contributed to their overall improvement of the LTI score.188 GAI 
prepared a detailed analysis of the LTI score for each of the participating LSGs and continued with 
recommendations for improvement with specific measures that LSGs might implement.
BEST PRACTICE: USING THE LTI
Participation: Representatives of all 13 GAI-selected LSGs (heads of administration, communication managers, and IT 
experts) participated in GAI workshops and follow-up activities.
Effects (and Impact): All GAI partners among the LSGs have increased transparency as measured by the LTI.189
Furthermore, GAI partners, the City of Novi Pazar and the City of Sombor, recorded the most significant improvements, 
reaching second and third position among all LSG units.190 Longer-term effects could be recognized at the level of developed 
systems, as they generated awareness and commitment among the LSGs to work on and ensure more transparent 
functioning. 
Replicability: The LTI scoring involves all municipalities and cities in Serbia; GAI’s step-by-step approach for the 
improvement of transparency is available to all LSGs. 
Sustainability: The achieved results likely will be sustained by GAI-selected LSGs, as the leadership and professionals have 
been strongly engaged to achieve and sustain transparency.191
ENGAGING YOUTH IN BUDGETING PROCESSES.192 Youth engagement is an added value of GAI’s 
support to budgeting in Sombor. GAI designed activities enabled young people to provide input to 
budget planning by suggesting projects to include in the city budget. The Activity, together with the 
administration of the City of Sombor and the high school student parliaments, enabled voting on 
184 The LTI is a sum of positive points obtained from the analysis of web sites, documents, responses to requests 
for free access to information and direct insight into the functioning of the service centers and premises of the 
local administrations. https://www.transparentnost.org.rs/index.php/en/research-on-corruption/lti
185 The officials from participating LSGs expressed their commitment to continue working for the improved 
position on the LTI list, stating numerous activities beyond the scope of the LTI-assessment: publishing of all 
procurement data and contracts, meeting minutes from the local assembly meetings; decisions from the municipal 
council’s meetings, etc. 
186 FGD_01, FGD_02, FGD_03, KII_08, KII_14, KII_21, KII_22
187 https://www.transparentnost.org.rs/LTI2020/ and https://www.transparentnost.org.rs/LTI2019/
188 FGD_01, FGD_02, FGD_03, KII_08, KII_14
189 https://www.transparentnost.org.rs/index.php/sr/istraivanja-o-korupciji/lti
190 https://www.transparentnost.org.rs/images/dokumenti_uz_vesti/LTI_2020_final_report_ENG.pdf
191 KII notes LSGs and FGD_01, FGD_02, FGD_03
192 With GAI's support, the City of Sombor successfully implemented an innovative public engagement action plan 
for budget public hearings, organizing discussions in all local communities on topics specifically related to these 
communities. These events enabled citizens to discuss the proposed budget and provide suggestions directly to 
city representatives.
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 26
potential projects to take place. GAI created a campaign and developed voting materials for schools to 
raise youth awareness about the importance of participation in local decision making.
BEST PRACTICE: ENGAGING YOUTH IN BUDGETING PROCESSES 
Participation: More than 2,700 students from 5 high schools participated, representing more than 80 percent of Sombor's 
high school student population.193
Effects (and Impact): The approved and implemented projects have improved educational infrastructure, such as the 
renovation of school toilets, purchase of air conditioners, and acquisition of equipment for classroom training and student 
leisure activities. These projects have generated additional trust and commitment of the youth to participate in municipal 
affairs and solve immediate problems.194
Replicability: The model from Sombor has high potential for replication. GAI has begun similar activities in the City of 
Šabac, facilitating sustainable cooperation between the city administration and high schools.195
Sustainability: The City of Sombor has adopted student participation in budgeting as a regular practice and will continue 
to support students in formulating project proposals to be funded from the local budget.196
SUPPORTING SAI. Support to SAI to introduce and implement performance audits included a 
comprehensive capacity development support, training for specialized performance audit teams, and 
mentoring assistance. The objective is to assess if public system organizations and their operations, 
programs and activities are operating in accordance with the principles of economy, efficiency, and 
effectiveness. GAI supported SAI’s 13 performance audit teams to finalize initial performance audit 
designs for the 2020 performance audits. GAI also facilitated the teams’ preparatory work for the 
presentation of their audit objectives, research questions, selection of relevant methodology, and 
selected auditees during the preliminary design summits. GAI supported SAI in codifying the knowledge 
by releasing the Performance Audit Policies and Procedures Manual.197
BEST PRACTICE: SUPPORTING SAI
Participation: GAI has provided comprehensive support to SAI related to performance audits, especially for the special 
performance audit teams.198
Effects (and Impact): The approved and implemented activities have demonstrated an impact on improved capacities of 
SAI to independently implement performance audits, bringing financial benefits to the country.199
Replicability: The capacity development program could be replicated and shared with SAI public auditors.
Sustainability: SAI has adopted performance audits as its regular practice and will continue.200
STRENGTHENING COOPERATION BETWEEN CPI AND LSGs. GAI, together with CPI, trained 
LSGs to implement the Law on Personal Data Protection, with the focus to encourage local authorities 
to be more open in response to public, CSO, and media requests for public information. These capacity 
development efforts have been especially relevant, considering that local administrations, including their 
direct and indirect budget users, handle large amounts of personal data, while simultaneously facing 
193 KII_08
194 FGD_04
195 KII_08 and KII_14 and FGD_03
196 KII_08
197 The manual incorporated performance audit design and message agreement practices, as well as methodology 
guides on specific aspects of performance audits. This included topics such as design, creation and use of sampling, 
audit criteria, questionnaire design, case studies, content analysis, and data reliability during audits. 
https://www.odgovornavlast.rs/knowledge-sharing/
198 KII_07
199 According to unofficial data from SAI, around 800 mil dinars have been saved. KII_07
200 KII_07
27 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
implementation challenges.201 For example, during the COVID-19 pandemic, all private data was 
protected, although there was media pressure to report on the identity of infected persons.202
2.4.4 LESSONS LEARNED
The evaluation team identified the following lessons learned at GAI’s mid-term: 
● Providing focused expertise and policy advice is an added value of GAI. Providing focused 
expertise and policy advice on specific priorities, GAI added value and contributed to building 
institutions to ensure more transparent and accountable governance in Serbia. The real sustainability 
of these results at the level of institutions and the overall legal and public management systems 
depends on the government’s commitment to accountable and transparent governance. However, 
interference of the executive and legislative branches with the mandates of IOIs could undermine 
the credibility of their performance, especially of their reported findings and recommendations. 
● Long-term support is crucial to ensure positive institutional development. The long-term 
USAID support to good governance in Serbia that has continued through GAI, together with 
technical assistance, has been instrumental for the institutional development of IOIs and local 
authorities in Serbia. It is critical USAID to work closely and collaboratively with its counterparts in 
Serbian institutions to achieve national ownership and, ultimately, sustainable change. Focusing on 
strengthening the technical capacity of analytical staff in government institutions is a strong approach 
to improve the delivery of its functions. However, the commitment of senior management in IOIs 
remains critical to ensure that developed capacities are embedded in organizational structure and 
contribute to the broader good governance agenda. The example of ACA shows that newly 
appointed senior management remained reluctant to utilize established capacities or participate in 
organizational development efforts, despite a declared commitment.203 The capacity building efforts 
of international development partners may legitimize senior management without achieving 
institutional impact or improved organizational performance.
● Incorporating technology-based efforts allows for greater participation. GAI tested 
innovative approaches to deliver training programs, benefiting from web-based opportunities. This 
experience demonstrates that the combination of traditional development interventions with the 
introduction and use of tailored new technologies can enhance the effectiveness of development 
efforts, increase results, and contribute to greater involvement of targeted groups.
201 The Law became effective in August 2019, but its successful implementation requires training, especially for 
administration and public bodies who are unaware of the new requirements and, according to the new CPI 
Commissioner, LSGs are not prepared to implement the Law.
202 KII_15
203 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf and 
https://freedomhouse.org/country/serbia/nations-transit/2020
BEST PRACTICE: STRENGTHENING COOPERATION BETWEEN CPI AND LSGs
Participation: GAI, in partnership with CPI, provided training to selected LSGs.
Effects (and Impact): Local authorities have an enhanced understanding of the Law on Personal Data Protection and its 
importance. GAI LSGs followed the law during the COVID-19 pandemic, although some LSGs have been disclosing the 
number of people that have been infected.
Replicability: CPI continued with the delivery of training programs to other LSGs in Serbia, recognizing financial 
constraints.
Sustainability: CPI adopted a training program as part of its regular services, however, sustainability will depend on 
participating LSGs’ readiness and commitment to implement the Law on Personal Data Protection.
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 28
● It is necessary to focus on LSG ownership. Selecting LSGs through a nation-wide open call 
generated commitment among these LSGs to participate in the Activity and to work and implement 
activities for enhanced transparency and accountability. These initial results together with the GAI 
partnership-building approach and demand-driven assistance, provided a platform for LSGs on which 
to build a strategy to progressively devolve responsibility to local authorities to lead and manage 
anti-corruption activities.
● Two-way communication is essential to establish trust between local authorities and 
Serbian citizens. Communication by the local authorities is crucial to increase transparency, and 
successful two-way communication remains critical for citizens and their meaningful involvement. If 
LSGs consider communication as a strategic tool, it will increase transparency and trust among its 
citizens. It is critical that LSGs continue to focus on unifying internal communication among 
employees and engaging in external communication with citizens to ensure sustainable transparency.
● Monitoring systems should incorporate results-based management principles. Although 
narrative reporting provides an overview of processes, it is time consuming and often misses links 
with broader changes at organizational or systemic levels. A comprehensive monitoring system must 
exist to guarantee rigorous control of activities and assessment of progress, while also contributing 
to effective reporting of results. Results-based management principles enabled greater transparency 
and partner awareness of achieved progress. The use of well-known and respected think-tanks and 
CSOs have been important factors to support LSGs. For example, Transparency Serbia and 
Pistaljka’s capacities and experience have been instrumental in ensuring progress and enhancing 
understanding of local governance structures towards greater transparency and accountability. 
2.5 EVALUATION QUESTION 5: GENDER EQUALITY
Did the Activity affect gender equality within the area(s) of focus?
Most stakeholders perceived USAID as steadily promoting gender equality in Serbia, particularly through 
assistance to LSGs.204 However, the evaluation team found that GAI has given less attention to 
gender equality and gender mainstreaming, although corruption “presents a barrier for 
women to gain full access to their civic, social and economic rights."205
The evaluation team found that the GAI MELP does not adequately integrate gender 
considerations. The MELP is void of gender sensitive indicators and minimally draws on the use of 
sex-disaggregated data. Furthermore, the GAI team did not design data collection methods to 
systemically disaggregate information by gender and monitor the meaningful engagement of women 
across activities. 
204 An example of a reference is the USAID-funded Community Development through Democratic Action (CRDA) 
Program implemented from 2001 – 2006. CRDA programs prescribed to include at least 30 percent of women in 
local commune councils, bringing change in the gender patterns in participation. Programs also enabled higher 
activism of women in areas with a long tradition of women’s grassroots initiatives (i.e. Vojvodina).- ref to “The 
Role of Community Development and Citizen Engagement Activities in Strengthening Civic Engagement and 
Government Responsiveness in Serbia,” https://secons.net/publications.php?p=75
205 Gender, Equality and Corruption: What are the Linkages? Policy Brief 01/2014, Transparency international 
https://images.transparencycdn.org/images/2014_Policybrief1_GenderEqualityCorruption_EN.pdf
29 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
For example, under the capacity development components, the 
focus has been only on gender participation (e.g., number of 
women participating in capacity development events). Although the 
opportunity existed to identify perceptions, understanding, and 
challenges to the achievement of gender equality in the broader 
anti-corruption framework (for example, through post-training 
evaluation questionnaires or through surveys on government’s 
effectiveness to tackle corruption), feedback questionnaires did not 
include options to receive gender-sensitive information, as there 
were no gender-specific questions.206 GAI still needs to work on 
and measure “lasting changes in the power and choices of women 
over their own lives, while tackling the root causes of inequality.”207
Exhibit 3 displays survey responses related to opinions about GAI’s 
support for gender equality during its implementation.
Most survey participants stated either that GAI insufficiently (25 percent) or partially (25 percent) 
addressed gender, while many (31 percent) did not have that information. However, a total of 10 
percent declared that GAI did not address gender equality at all; opposite to this, 10 percent of 
participants in the survey stated that GAI addressed gender in a satisfactory manner. Unfortunately, 
participants in the survey who expressed negative or relatively negative opinions did not justify or 
explain these views and perceptions. In addition, interviewed partners highlighted the importance of the 
technical support that GAI provided; at the same time, the partners were unaware of the links between 
gender equality and anti-corruption, expressing only limited knowledge of gender-sensitive practices.208
This lack of justification makes it difficult for the evaluation team to make specific recommendations or 
elaborate on the lack of gender inclusion. 
3. CONCLUSIONS AND RECOMMENDATIONS 
3.1 CONCLUSIONS
EVALUATION QUESTION 1: GAI ACCOMPLISHMENTS
Conclusion 1. GAI has been effective in delivering results and meeting targets, creating a 
solid basis to continue activities at the national and local levels to further strengthen good 
governance. GAI has achieved concrete and visible results during this period of implementation. Most 
notably, the results related to enhancing transparency at the local level and developing technical 
capacities of SAI. The evaluation team found that GAI has strengthened the capacities of the critical 
institutions responsible for improving government accountability, particularly at the local level and for 
participating institutions, such as SAI and LSGs. Despite GAI’s efforts and results, the indicators showed 
a decline in government accountability at the national level due mainly to lack of commitment.209
EVALUATION QUESTION 2: LINKING THE ACTIVITY’S OBJECTIVES AND RESULTS
206 The analysis of the MELP. 
207 https://www.undp.org/content/undp/en/home/sustainable-development-goals/goal-5-gender-equality.html
208 KII notes- national partners 
209 The analytical reports such as Nations in Transit 2020 highlighted the “country’s increasing slide towards 
authoritarianism,” with raising concerns over media freedom, political control, and the free and fair conduct of 
elections, among other issues. The European Commission’s 2020 Report on Serbia recognized limited efforts or 
absence of any progress in the core governance areas. Specifically, the country is lagging behind with the reform of 
the judiciary and fight against corruption, with numerous challenges to ensure respect of fundamental rights and 
freedom of expression- https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf
EXHIBIT 3: PERCEPTIONS OF GAI’S 
GENDER INTEGRATION (N=72)
HAS GAI EFFECTIVELY PROMOTED
AND ADDRESSED GENDER
EQUALITY DURING
IMPLEMENTATION OF ACTIVITIES?
Not at all 10%
Insufficiently 25%
Partially 25%
Completely 10%
Don’t have information 31%
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 30
Conclusion 2. GAI’s results show that it is meeting its objective, with notable results at the 
policy and institutional levels. GAI’s results have contributed to sectoral policies. The policy cycle
model confirmed that positive results from GAI have influenced the designing of policy tools. GAI 
results also contributed to organizational developments, particularly of SAI and participating LSGs. 
Despite this, GAI has faced challenges in cooperating with ACA and strengthening its capacities and 
delivery of services.
Conclusion 3. GAI has been implemented efficiently, in line with the work plans, and 
demonstrated flexibility to stakeholders’ needs. GAI established a clear system of work planning, 
to complete the Activity on time, within the approved budget and standard of quality. GAI also 
established an effective monitoring, evaluation, and learning system, with a team responsible for data￾collection and analysis. The proposed indicators are relevant to measure progress, but important 
elements of GAI remain outside of the monitoring system without appropriate indicators to measure 
progress and results.210 The reporting practice has been solid, but a specific results-oriented and gender 
sensitive approach, including references to adopted indicators, has been limited. 
Conclusion 4. The GAI team, including its senior management, contributed to the effective 
implementation and achievement of results. GAI has been a flexible and reliable partner, 
accountable for the achievements of results. These elements, together with the strong partnership and 
technical capacities of GAI staff, have been critical factors contributing to GAI’s effectiveness and its 
flexible and responsive implementation. Implementing partners commented positively about GAI’s 
responsiveness and involvement with design, planning, and implementation, while restating the relevance 
of the technical assistance that is available through the Activity. Yet, GAI did not have a steering or 
advisory mechanism to provide strategic guidance and coordinate development interventions in specific 
areas. This body could be an effective forum for greater involvement of the high-level national partners 
to genuinely drive development efforts, ensuring strong national leadership in the future.
EVALUATION QUESTION 3: SCOPE OF WORK EXPANSION
Conclusion 5. GAI’s scope is well-defined and already tested services need to continue. In 
Section 2.3, the evaluation team describes potential areas for additional GAI involvement at the sectoral 
level (for example, the overall anti-corruption strategy), institutional level (for example, strengthening 
local anti-corruption councils’ capacity to monitor plan implementation or strengthening mechanisms for 
private sector whistle-blowers), and individual level (for example, addressing the needs of youth or 
enhancing understanding of participation benefits for citizens) to complement already planned and 
implemented activities.
EVALUATION QUESTION 4: FUTURE PROGRAMMING
Conclusion 6. GAI’s intervention logic has been well-established, based on a sound TOC, 
ensuring its alignment with national strategic priorities under the good governance 
framework. GAI effectively responded to the needs to establish more transparent and 
accountable institutions. GAI remained relevant during the entire period of implementation thus far 
(2018-2020) and has aligned itself with the needs of the country. The responsiveness of GAI to the 
changing environment in Serbia and the needs of the IOIs, specialized anti-corruption court units, and 
LSGs (as the main partners) are critical factors to its relevance. 
210 For example, indicators could not adequately measure GAI’s performance in the areas, such as right to 
information, civic participation, complaint mechanisms. GAI did not introduce appropriate indicators to measure 
development of capacities of the target groups or the effects on anti-corruption education programs on the 
selected youth (e.g. Anti-corruption Legal Clinic and its participants). Also, GAI did not measure the effects of 
awareness campaigns and assess if the targeted audience has enhanced understanding of a specific topic. The MELP 
also did not include more specific indicators related to assistance to CPI and LSGs.
31 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
Conclusion 7. GAI has been effective in responding to national capacity development needs 
in areas of intervention. GAI has steadily and successfully developed the capacities of national 
partners to improve functioning and delivery of services (under all components). The actual sustainability 
of these results depends upon the continuing availability of these capacities and that they will continue to 
be requested upon the completion of GAI.
Conclusion 8. The sense of national ownership over the achievements under GAI is due to 
effective partnerships and the involvement of the national stakeholders in its design and 
implementation. The participatory approach for the design and implementation of GAI 
has contributed to an enhanced sense of ownership, setting the ground for sustainability of 
results. The sustainability of GAI’s achievements, however, require additional attention and further 
efforts. Several external factors may undermine the sustainability of GAI interventions, such as the 
Government’s continued lack of commitment to implement reforms and anti-corruption measures, the 
deteriorating role of independent oversight institutions; unpredictable political developments; and the 
recent economic decline largely due to the COVID-19 pandemic.211
EVALUATION QUESTION 5: GENDER EQUALITY
Conclusion 9. GAI has given relatively less attention to gender equality and gender 
mainstreaming. The MELP is not gender sensitive. The use of sex-disaggregated indicators has been 
limited, and GAI team did not design methods of data collection and analysis for systematization of 
gender-disaggregated information. 
3.2 RECOMMENDATIONS 
Through its analysis of primary and secondary data, the evaluation team identified challenges to the 
implementation of GAI. The evaluation team proposes the following main recommendations to USAID 
and GAI to resolve these problems:
Recommendation 1. The evaluation team recommends that GAI focus on the next stage of 
strengthening government accountability at national and local levels. The next stage should focus on 
implementation of identified organizational strategic priorities for the SAI, CPI, ACA, specialized anti￾corruption court units, and LSGs and follow up on improvement of services, by helping to build 
capacities, systems, and connections among the key partners/institutions.
Recommendation 2. Development of research and analytical capacities of ACA staff remains a 
dominant priority for this organization and GAI is well-positioned to continue building on the existing 
achievements and extend its technical assistance. However, if the ACA's leadership continues to resist 
implementing its oversight mandate and the new procedures and policies that GAI supports, GAI should 
reconsider supporting ACA. 
Recommendation 3. GAI should remain flexible and responsive to the needs and priorities of the 
good governance framework. Concerning responsiveness, emerging priorities include:
● Systemic Level. To engage more actively in the policy making process related to anti-corruption, 
transparency, and accountability, GAI should codify and summarize knowledge in the core areas of 
its work and share it with the government as the basis for or evidence to create a more favorable 
legal and policy environment in this area. GAI also should provide specific support to IOIs to 
respond to the corruption risks and challenges posed by COVID-19.
● Institutional Level. GAI should: 1) continue support to local authorities – priorities could include: a) 
designing sound and realistic monitoring system for LAPs; b) enhancing the capacities of local 
211 https://ec.europa.eu/neighbourhood-enlargement/sites/near/files/serbia_report_2020.pdf and 
https://freedomhouse.org/country/serbia/nations-transit/2020
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 32
independent anti-corruption bodies, in charge of LAPs monitoring; c) enhancing the capacities to 
implement recommendations to improve budgeting process and local finances; and d) enhancing 
strategic communication skills of local officials in senior positions; 2) explore opportunities to 
engage with public enterprises, including public utility companies, and enhance their transparency, in 
line with the internal rules, legal provisions, and requirements; 3) expand the reach of and embrace 
private sector interest, their roles and responsibilities in advancing good governance, and to fight 
corruption efforts in the country; 4) strengthen and establish whistleblower protection mechanisms 
for the private sector, enabling them to report on corruption; 5) strengthen capacities of the NA 
and its relevant committees to read, understand, and comment on the oversight institution reports, 
and present findings and recommendations; 6) strengthen the technical capacity of prosecutors; and 
7) Explore opportunities to support the network of DPOs.212
● Individual Level. GAI should: 1) invest in additional efforts to reach youth, offering capacity 
development programs to prevent corruption, including anti-corruption clinic models, on-line 
training programs, and the use of social networks and on-line discussion forums; 2) prepare and 
implement programs on anti-corruption framework for the journalists; and 3) increase “demand” for 
participation, by working with the public to create broader interest and willingness to take part in 
local decision-making processes.
Recommendation 4. GAI should define appropriate qualitative and quantitative indicators to enable 
GAI to adequately measure progress under components and capture GAI’s progress towards its 
objective.213 The evaluation team recommends that GAI provide a well-balanced combination of 
qualitative and quantitative indicators to capture changes and results attributable to the Activity, using 
national indicators and targets to the extent possible. The evaluation team also recommends including 
gender-sensitive indicators with a focus on “gender transformation,” identify assumptions and risks, and 
implement a sound risk management system. 
Recommendation 5. In addition to including gender-sensitive indicators, GAI should consider options 
to better integrate gender priorities in its activities and the overall anti-corruption framework. 
Additionally, GAI should consider activities and approaches that go beyond gender participation.
Recommendation 6. GAI should continue implementing its systemic approach to capacity 
development, following needs assessments of the main stakeholders and partners. Considering the 
current stage of the governance systems, structures, and capacities of partner institutions in Serbia, GAI 
should consider a long-term, needs-based capacity development approach. The evaluation team
recommends that GAI design a comprehensive training evaluation approach based on the Kirkpatrick 
model, assessing four levels of learning: 1) participant satisfaction with the training; 2) immediate change 
in individual knowledge and skills; 3) change in individual performance back in the workplace; and 4) 
change in the overall performance of the institution. Finally, GAI should prepare a plan to hand-over 
training programs to the relevant national institutions and prioritize preparing them to assume these
responsibilities.
212 GAI’s role has been to support analysis of the budgeting processes and prepare recommendations for 
improvements; however, the Action did not provide follow up support, to implement recommendations.
213 For example, indicators could not adequately measure GAI’s performance in the areas such as right to 
information, civic participation, and complaint mechanisms. GAI did not introduce appropriate indicators to 
measure development of capacities of the target groups or the effects on anti-corruption education programs on 
the selected youth. Also, GAI did not measure the effects of awareness campaigns and assess if the targeted 
audience has enhanced understanding of a specific topic. The MELP also did not include more specific indicators 
related to assistance to CPI and LSGs.
33 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
ANNEX A. EVALUATION STATEMENT OF WORK
The United States Agency for International Development (USAID) in Serbia seeks the services of a 
Contractor to conduct a performance evaluation of the USAID-funded project “Government 
Accountability Initiative” (GAI).
C.1. BACKGROUND 
The goal of this activity is to strengthen the capacities and connections of key Serbian stakeholders to 
increase government accountability at the national and local levels. The development hypothesis is that if 
the skills and connections of government institutions and oversight bodies are increased, then Serbia will 
have a stronger foundation for more responsive and accountable governance, and in order to achieve its 
purpose, this activity focuses on three objectives: local government is more accountable; oversight of 
government institutions strengthened; and adjudication of corruption cases increased. 
Within these three areas, the Contractor must: a) support efforts to increase the inclusiveness of public
decision making regarding public resource allocations and to encourage and incentivize local 
governments to operate in a more transparent and responsive manner; b) provide assistance to 
oversight institutions to strengthen monitoring of government performance and public officials, and 
exercise mandates as defined per relevant legislation; c) provide assistance to specialized anti-corruption 
court units and help establish a system for the collection of unified statistics on corruption, distinguishing 
clearly between different types of criminal offences, the length of cases, outcome of the proceedings. 
Detailed background information on the SOW for this activity can be found in Attachment #1 of this 
RFTOP (“GAI SOW”).
C.2. PURPOSE OF THE EVALUATION 
USAID/Serbia requires a full, evidence-based and independent mid-term performance evaluation of the 
USAID/Serbia Government Accountability Initiative (GAI) Activity, and a report that provides a 
qualitative as well as quantitative evaluation of the progress made towards achieving expected results 
per the contract. Specifically, the purpose of the evaluation is to: 
• Examine the effectiveness of project activities implemented to date; 
• Determine whether the project has achieved planned results; 
• Identify gaps in performance against targets; and 
• Provide recommendations on closing these gaps in the upcoming years of the project. 
The Contractor must not simply provide an accounting of performance against targets but provide 
independent analysis on why targets were realized, met, or significantly surpassed. The Contractor must 
also provide an analysis of the reasons behind gaps between expected and actual performance, including 
an identification of the likely source(s) of these gaps (design, project management, changes in operating 
environment, change in context, success factors, if contractor’s approach has led to intended results and 
goals, etc.). 
C.3. EVALUATION QUESTIONS, METHODOLOGY AND DATA SOURCES 
Evaluation Questions: 
Evaluation questions that Contractor must address are as follows: 
1. To what extent is the program on track to accomplish the inputs, outputs and likely outcomes 
expected under all three objectives of the activity? 
2. What link can be determined between progress toward the activity's objectives and the activity's 
results? 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 34
3. What other areas of interventions could fit under the SOW beyond what is already planned? 
4. What are the interventions with greatest traction, unintended outcomes and/or lessons learned 
from USAID assistance that the Mission can use in decision making for future programming? Did 
the activity affect gender equality within the area(s) of focus? 
Evaluation Methodology: 
In developing an appropriate methodology, the contractor must consider potential limitations and 
incorporate best-practice mitigation measures. The evaluation must be supported by relevant data and 
information gathered from meetings with beneficiaries and all the relevant stakeholders. The evaluation 
team should clearly demonstrate links between the data collected and the subsequent findings, 
conclusions, and recommendations drawn. Where applicable, responses to the questions above and 
resulting data must be dis-aggregated by sex and gender. 
Available Data Sources: 
Suggested data sources include, but are not limited to: (a) secondary data/background documents, (b) 
activity plans, outputs, and reports, (c) key informant interviews, (d) focus group discussions and (e) 
survey(s) of project stakeholders and participants. Contractor must place emphasis on the collection of 
reliable empirical data and/or objectively verifiable evidence, as opposed to anecdotal evidence. Where 
surveys or interviews are used, appropriate sampling and questioning techniques must be utilized to 
ensure representative results; where references are made to data generated by GAI and their partners, 
it must be complemented by references to independent data sources and any significant data differences 
must be explained. 
The contractor will have access to all performance data, M&E database, surveys, and other deliverables, 
that have been collected and reported by the implementing partner as part of their Performance 
Monitoring and Evaluation plan for GAI. The contractor must also review the following documents in 
preparation for the GAI performance evaluation: 
• Relevant project reports, assessments, annual and life of project work plans. 
• Full data set in their formats available and collected by the implementing partner. 
• Annual performance reports of GAI partners and sub-partners. 
• Anti Corruption Assessment conducted prior to the GAI design. 
• USAID/Serbia and Montenegro’s 2011-2015 Amended Strategy and 2013-2017 CDCS. 
In addition, the evaluation team must consult with key informants and stakeholders, including:
• Meetings with USAID/Serbia. 
• Meeting with Checchi’s home office key staff and Belgrade office GAI staff. 
• Meetings with key current/former GAI implementing partners, sub-contractors, and sub￾grantees in Serbia. 
• Meeting with relevant officials of the Government of Serbia (such as the Ministry of Justice), 
oversight institutions (such as SAI, ACA), and anti-corruption court units. 
• Meetings with other donors, including the EU Delegation, and those donors engaged in this 
area. 
C.4. GENERAL CONTRACTOR REQUIREMENTS – STATEMENT OF WORK 
The Contractor must conduct a methodologically sound evaluation which must be fully compliant with 
USAID’s Evaluation Policy and Automated Directives System 
(https://www.usaid.gov/sites/default/files/documents/1870/USAIDEvaluationPolicy.pdf), and must utilize 
tested and scientific evaluation methodologies.
35 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
C.5. TEAM COMPOSITION
The Contractor must utilize a team with the experience and skills necessary to complete this task as 
specified in Section F.4. 
C.6. SCHEDULE AND LOGISTICS 
The schedule (with precise due dates for deliverables) will be defined in the work plan before the start 
of the in-country field work. Prior to commencing the field work, all team members will review 
background program documents to gain a firm understanding of the situation in Serbia and the activity. 
The Contractor will be responsible for logistical arrangements for conducting the evaluation, including 
scheduling meetings, making copies of documents and drafts, making travel arrangements, etc. Field visit 
logistics will be organized and managed by the Contractor. This includes travel in-country 
transportation. No international travel is anticipated. USAID/Serbia will provide the team with input and 
guidance in setting up a schedule of interviews and site visits, but the responsibility for the schedule for 
the evaluation resides with the Evaluation Team.
The Contractor is responsible for ensuring the safety and well-being of Contractor’s staff 
and shall apply accepted practices in accordance with the local laws and health authorities’ 
directives.
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 36
ANNEX B. ACTIVITY BACKGROUND
USAID/Serbia established “Accountability of Key Democratic Institutions Strengthened” as the first of its 
two Development Objectives (DOs) in its Country Development Cooperation Strategy (Fiscal Years 
2013 – 2020), to address the challenges in sustaining accountable and representative governance in 
Serbia. These challenges result in part from the lack of adequate checks on central authority and 
sufficient opportunities for citizens to hold the Serbian government and elected officials accountable. 
Under this DO, USAID/Serbia provided on-going support to strengthen the accountability and 
representativeness of key democratic institutions in Serbia, with the goal of improving governance and 
furthering Serbia’s efforts towards EU accession. Directly tied to this DO, USAID/Serbia funded the 
GAI, a four-year activity with the overall objective to strengthen the capacities and connections of key 
Serbian governmental institutions and stakeholders, resulting in increased government accountability at 
the national and local levels. To achieve this objective, the Activity is structured according to three IRs: 
1) increase public participation in local government decision-making and oversight; 2) increase follow-up 
to oversight institutions’ recommendations; and 3) increase adjudication of corruption cases. This $8.5 
million, four-year activity started in February 2018 and will end in February 2022. In this section, the 
evaluation team gives an overview of the Activity’s key components, organized according to each IR.
Corruption is principally a governance issue which indicates failure of institutions and a lack of 
institutional capacity to manage society by means of a framework of social, judicial, political and 
economic checks and balances.214 When these formal and informal systems break down, it becomes 
harder to implement and enforce laws and policies that ensure accountability and transparency. From an 
institutional perspective, corruption arises when public officials have wide authority, little accountability, 
and perverse incentives, or when their accountability responds to informal rather than formal forms of 
regulation. Considering this, the focus of GAI has been to strengthen capacities and organizational 
development of the main IOIs, as well as strengthen the operational and functional links between them. 
Furthermore, GAI assistance at the local level strives to enable targeted LSGs to fulfil their 
commitments and obligations by promoting constructive interaction among local authorities and citizens. 
These activities have been infused with the principles of anti-corruption and good governance. 
The GAI TOC provided an overview of these and other challenges and issues, enabling the evaluation 
team to analyze systemically the formulation of GAI’s intervention logic and establishment of its chain of 
results.215 We present the GAI TOC in the box below.
GAI Theory of Change216
“The GAI TOC is that if select branches and levels of Serbian government are provided tools for good 
governance, they will have the will and ability to apply them constructively to result in improved governance for 
all Serbian residents. 
● Specifically, we believe that sound examples of good governance at the local level, applied in ways 
that show clear benefit to the governments themselves, will prove appealing and durable, so that 
counterpart LSGs continue to apply them, and non-counterparts are interested in adopting them. This in 
turn will increase opportunities for members of the public to participate in and evaluate decisions by LSGs. 
214 The principles at the heart of good governance are the same as those at the heart of democracy: Equity; Participation; 
Pluralism; Partnership; Subsidiarity; Transparency; Accountability; Rule of law; Effectiveness; Efficiency; Responsiveness; and 
Sustainability.
215 OECD DAC definition: “Chain of results is the causal sequence for a development intervention that stipulates the necessary 
sequence to achieve desired objectives beginning with inputs, moving through activities and outputs, and culminating in 
outcomes, impacts, and feedback”
216 USAID/Serbia GAI MELP. 6 February 2018 – 5 February 2022 (October 19, 2018)
37 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
● We believe that support to independent oversight bodies can improve the capacities of some, and 
slow or arrest the declining role of others, such that the oversight bodies continue to act as meaningful 
checks on government corruption. 
● We believe that building the capacity of new anti-corruption courts and PPOs and identifying 
bottlenecks in corruption case processing will give counterparts the ability to increase the 
adjudication of corruption cases if they are motivated to do so. 
Overall, these results will lead to an increased ability of independent and judicial bodies and of the public to 
check malfeasance by government actors.”
IR1: INCREASED PUBLIC PARTICIPATION IN LOCAL GOVERNMENT DECISION-MAKING 
AND OVERSIGHT
Under IR1, the Activity is supporting efforts to increase the inclusiveness of public decision making 
regarding public resource allocations and to encourage and incentivize LSGs to operate in a more 
transparent and responsive manner. GAI anticipates that the adoption of good governance principles will 
offer members of the public a better understanding of how government works and enable them to 
participate in government decision-making processes, which will in turn reduce opportunities for 
corruption, while increasing transparency, accountability, and inclusiveness.217 At the Activity mid-point, 
GAI has iteratively provided technical assistance to 13 LSGs in areas in which they show interest and 
willingness to implement changes and new policies.218 Notably, the Activity provided assistance in 
developing LAPs and establishing of independent monitoring bodies (IMBs), and, in 2020, is transitioning 
to supporting LAP implementation and capacity building of IMBs.219 The evaluation team summarizes the 
key activities under IR1, as described in the GAI work plans for project years 1-3:
LAP DEVELOPMENT AND OVERSIGHT. GAI provided technical assistance to 13 LSGs to develop 
and/or revise LAPs and establish independent monitoring bodies. GAI also developed and proposed 
mechanisms and tools to strengthen oversight of LAP implementation to be used primarily by IMBs. 
Together, this work will be used to raise public awareness in the LSG communities about the 
importance of creating and implementing sound local anti-corruption policies. GAI also focused on 
identifying and documenting experiences and successful practices to share with other LSGs, as well as 
with relevant stakeholders to raise awareness and provide guidelines to others. In 2020, GAI began to 
focus its technical assistance on LAP implementation and capacity building of IMBs and also began 
providing capacity building workshops for GAI-supported IMBs to share experiences with other IMBs 
that do not receive donor-funded technical assistance.
TRANSPARENCY. In addition to supporting the development of LAPs, GAI focused on specific topics 
such as transparency, public participation, budgets and spending, public services, and whistleblower 
protection. GAI work closely with counterpart LSGs to make clear how and why these interventions 
can have both long and short-term benefits for the LSGs and the public they serve. Specifically, GAI 
developed a model LSG website aligned with legal requirements and good governance standards, as well 
as an open budget portal in line with international best practices. GAI also recommended necessary 
improvements of individual LSG websites, to ensure relevant information is made available to the public 
in a transparent and user-friendly manner. Supporting this work, GAI also promoted LSGs use of social 
media platforms by suggesting content and training them in their use. In partnership with Transparency 
Serbia (TS), the Activity also provided LSGs with action plans that included concrete steps required to 
improve the LTI survey scores. GAI is also assisting LSGs in aligning their internal procedures with the 
217 USAID/Serbia GAI MELP. 6 February 2018 – 5 February 2022 (October 19, 2018)
218 LSGs were added to the Activity in two groups: Group One – Cities of Vranje, Šabac, and Sombor; and the municipalities of 
Raška, Vrnjačka Banja, Dimitrovgrad, and Sjenica; Group Two – Cities of Niš, Kragujevac, Sremska Mitrovica, Novi Pazar, the 
municipality of Žabalj, and the city municipality of Stari Grad. The combined population of the 13 LSGs is 1,046,000, 
representing 14.8 percent of the total population in Serbia.
219 USAID/Serbia GAI Work Plan: Year 3 – February 06, 2020 – February 05, 2021 (December 20, 2019)
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 38
new law, partnering with the Public Procurement Office to provide all counterpart LSGs training on the 
use of its procurement portal, and providing tailor-made technical assistance to interested LSGs to 
develop and implement communication strategies for better interaction with the public.
PUBLIC PARTICIPATION. GAI is supporting LSGs’ efforts to increase public involvement in 
government decision making, particularly public policy development and resource allocation. Specifically, 
to improve the supply side of public participation, the Activity aims to build capacity of LSGs to improve 
initiatives or develop new initiatives to improve communication strategies, use appropriate channels to 
facilitate public participation, and provide adequate feedback to citizens. This includes LSG capacity to 
engage the public in formulating, monitoring, and measuring the impact of public policies. To strengthen 
the demand side, GAI seeks to create interest and demand among citizens, youth organizations, 
businesses, and CSOs to take an active part in making decisions on locally important topics, notably 
through developing and implementing public awareness campaigns.
PUBLIC SERVICES. GAI is implementing technical assistance targeting specific services in each LSG 
participating in GAI, responding to local needs, which will provide opportunities to significantly reduce 
the prevalence of corruption and improve governments’ efficiency and the quality of services provided 
to citizens and businesses. Specifically, GAI is assisting LSGs in optimizing procedures and introducing 
public quality monitoring for selected services in each LSG, as well as developing new tools and 
procedures to enable LSGs to receive information and feedback on the quality of services and better 
respond to the legitimate public needs and expected level of quality of delivered services.
WHISTLEBLOWER PROTECTION. GAI is providing both basic and more practical and focused 
trainings to designated persons in LSGs on protecting whistleblowers, as well as creating a network for 
sharing experiences and cooperation among the trained individuals. In addition, GAI supports 
counterpart LSGs with direct technical assistance for defining the most appropriate systems for 
whistleblowers protection.
IR2: INCREASED FOLLOW-UP TO OVERSIGHT INSTITUTIONS’ RECOMMENDATIONS
Under IR2, the Activity is working with the Serbian government oversight institutions to strengthen the 
ability of the ACA and SAI to monitor the performance of government and public officials, and to 
exercise their mandates as defined in relevant legislation. GAI expects that, by increasing the focus of 
IOIs’ resources on risk-based prevention and detection of corruption and improving communications 
and coordination with the NA, the public, and counterparts, IOIs will achieve more successful outcomes 
and greater public visibility and respect. Improved interface among the institutions and stakeholders will, 
in turn, contribute to more significant response by the NA and enforcement bodies to IOI 
recommendations.220 The evaluation team summarizes the key activities under IR2, as described in the 
GAI work plans for project years 1-3:
ASSISTANCE TO SAI. GAI is helping SAI to enhance strategic planning and management, performance 
auditing, reporting and oversight processes, with the goal of increase SAI’s ability to execute its 
mandate. Notably, GAI staff are mentoring performance audit teams to enhance audit design and 
implementation, integrate financial and compliance audit expertise, and enhance reporting and 
recommendation follow-up policies and processes.
ASSISTANCE TO ACA. GAI is providing ACA with assistance in strategic planning, risk-based 
monitoring of government officials’ conduct, and reporting with effective follow-up from action agents, 
with the goal of increasing ACA’s ability to execute its mandate. Particularly, GAI has focused on helping 
ACA to implement strategies to prevent and detect corruption; strengthen ACA’s oversight of asset and 
220 USAID/Serbia GAI MELP. 6 February 2018 – 5 February 2022 (October 19, 2018)
39 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
income declaration statements; and support ACA’s training for government institutions to implement 
codes of ethical conduct, among other areas.
SUPPORT TO THE COMMISSIONER FOR FREE ACCESS TO INFORMATION OF PUBLIC 
IMPORTANCE AND PERSONAL DATA PROTECTION (CPI). GAI is implementing activities to 
support CPI’s two key mandates regarding access to information of public importance and personal data 
protection. GAI is also working to raise awareness of citizens’ rights to access information from the 
government, including by collaborating with CSOs to identify strategies to improve their access to 
information. GAI also partnered with CPI to provide training at the local level on the requirements in 
the new Law on Personal Data Protection. Overall, the focus of GAI’s support to CPI has been 
sustaining the attention of all institutions on freedom of access to public information, particularly by 
ministries that propose laws related to freedom of information. 
IR3: INCREASED ADJUDICATION OF CORRUPTION CASES
Finally, under IR3, GAI is working through the Serbian Judiciary’s structures to provide technical 
assistance to specialized anti-corruption court units and PPOs. The Activity’s work is designed to 
facilitate the implementation of the Law on Organization and Jurisdiction of State Organs in the Fight 
against Organized Crime, Corruption, and Terrorism Financing, which is also known as the Law on 
Organized Crime and Corruption. Specifically, GAI committed to providing technical assistance to anti￾corruption specialized court units to help them to implement the new law. This assistance will help them 
establish a system to collect unified statistics on corruption, distinguishing clearly between different 
types of criminal offences, the length of cases, and outcome of the court proceedings. This in in turn 
should enable justice system leaders to pinpoint obstacles and bottlenecks that impeded the successful 
adjudication of corruption cases.221 The evaluation team summarizes the key activities under IR3, as 
described in the GAI work plans for project years 1-3:
SUPPORT TO SPECIALIZED ANTI-CORRUPTION COURT UNITS. To support the court units, 
GAI conducted an initial needs assessment which examined the institutional capacities, IT infrastructure, 
and utilization of existing case management systems of the PPOs and court units, given the required 
cooperation of state organs as stated in the Law on Organized Crime and existing interdependencies 
among the institutions. Informed by this assessment, among other activities, GAI organized two 
roundtables on the adjudication of corruption cases, which were facilitated by a U.S. District Court 
judge, to introduce U.S. best practices in adjudicating financial crimes and corruption. The Activity also 
organized two roundtables with judges of specialized anti-corruption court units to harmonize 
procedures. GAI also is exposing judiciary and prosecutorial communities to practices used in other 
countries, such as the use of forensic experts to reinforce the benefits of such interventions newly 
introduced to Serbia by the 2018 legislation. 
ELECTRONIC REGISTER OF CORRUPTION CASES. GAI is supporting the development of a new 
ERCC that will enable the anti-corruption court units to track and report on cases, with the goal to 
improve information sharing among institutions pursuing prosecution and adjudication of corruption 
cases. GAI completed an initial needs assessment and mapped key stakeholders and best practices in 
case management systems. The Activity also identified the main interlocutors who are currently 
amending regulations to establish standardized methodology for data collection, record keeping, and 
statistical reporting on criminal offences of corruption. GAI is now finalizing its database model and 
database dictionary to ensure the compatibility of catalogues with existing case management systems, 
including AVP and SAPO. GAI is identifying an IT company to develop the defined hardware and 
software architecture solutions for the ERCC.
221 USAID/Serbia GAI MELP. 6 February 2018 – 5 February 2022 (October 19, 2018)
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 40
ANNEX C. ANALYSIS OF KEY PERFORMANCE INDICATORS
INDICATOR DEFINITION FINDINGS
OBJECTIVE: TO STRENGTHEN THE CAPACITIES AND CONNECTIONS OF THE KEY STAKEHOLDERS TO INCREASE 
GOVERNMENT ACCOUNTABILITY AT NATIONAL AND LOCAL LEVELS
OI-1. Open 
Government – WJP
Baseline: 0.49 (2017/18)
Target: 0.51 (2020) 
The World Justice Project (WJP) Open 
Government Index (OGI) uses 4 dimensions to 
measure government openness, publicized laws 
and government data, right to information, 
civic participation and complaint mechanisms. 
A score from 0 to 1, with 1 indicating the 
highest level of openness, based on household 
surveys and in-country expert questionnaires 
collected for the WJP’s Rule of Law Index.222
The issue remains if and to what extent the 
Objective 1ndicator (OI)-1 can help measure 
attainment of GAI’s objective. GAI has been 
working in some of the WJP OGI related areas 
(e.g., right to information, civic participation, 
complaint mechanisms). The evaluation team 
found that this indicator could serve to satisfy a 
broad contribution claim, without clear 
justification between the results of GAI and 
progress measured by this indicator. 
OI-2. LTI (TS)
Baseline: 40.35 (2015)
Target: 44 (2020)
Average numerical score, on a scale from 0 to 
100, on the LTI developed by Transparency 
Serbia, for Local Governments assisted by GAI. 
The initial baseline is the average score of all 
LGs covered by the 2015 Index (the LTI 2015 
is more complete than the 2017 Index). As the 
project progresses, GAI will report on 
progress of LSGs GAI selects to support as 
well as the entire universe of LSGs covered in 
each year’s Index.
The OI-2 LTI measures “government 
accountability at local levels,” and it is relevant 
to measure progress under the Objective. 
Reflecting different dimensions such as the 
analysis of web sites, documents, responses to 
requests for free access to information and 
direct insight into the functioning of services, 
this indicator directly correlates with the core 
areas of GAI support at the local level.
OI-3. Number of 
public-sector 
employees and 
representatives of 
CSOs trained on good 
governance principles 
and accountability
Baseline: 0
Target: 1,865 (2020)
Training is defined as in-class or in-service 
technical training and assistance to impart 
knowledge and skills, as well as an exchange of 
ideas, values or culture, for public-sector 
employees and representatives of CSOs. Good 
governance principles are understood as 
defined by UNDP governance principles, and 
include accountability, transparency, rule of 
law, responsiveness, equity and inclusiveness, 
effectiveness and efficiency, and participation. 
Training intends to assist in the 
implementation of existing laws, enhance 
counterpart communication when performing 
mandates and tasks, strengthen inclusiveness 
and transparency in processes, and creating 
more accountable government on all levels. 
The training topics may include, but are not 
limited to, performance auditing, 
whistleblower protection, strategic 
communication, access to information, 
citizen/public participation and inclusiveness, 
and use of information system. The 
methodology will be adjusted to the topics of 
trainings and may include analysis and 
application of relevant professional standards 
and legislation, review and analysis of practical 
cases, presentation of basic concepts, roles, 
responsibilities, and international practices.
The OI-3 indicator measures the dimension 
“strengthen capacities of the key stakeholders.” 
However, counting “number of public-sector 
employees and representatives of CSOs 
trained on good governance principles and 
accountability” is not a valid reference to 
strengthen capacities. This indicator could be 
useful at the result-level, but not at the 
objective level.
222 The World Justice Project® (WJP) is an independent, multidisciplinary organization working to advance the rule of law 
worldwide. The World Justice Project has three primary goals: i) Increased understanding of the rule of law and its foundational 
importance; ii) Greater rule of law adherence by governments; and iii) Multidisciplinary, home-grown cultures of the rule of law. 
More at: https://worldjusticeproject.org/ The effective rule of law reduces corruption, combats poverty and disease, and 
protects people from injustices large and small. It is the foundation for communities of justice, opportunity, and peace—
underpinning development, accountable government, and respect for fundamental rights.
41 | MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA USAID.GOV
INDICATOR DEFINITION FINDINGS
IR1. INCREASED PUBLIC PARTICIPATION IN LOCAL GOVERNMENT DECISION-MAKING AND OVERSIGHT
Number of policies or 
mechanisms adopted in 
line with procedures to 
increase public 
engagement in selected 
LSGs
Baseline: 0 (2017/18)
Target: 16 (2020) 
The number of policies or mechanisms that 
have been adopted in counterpart LSGs to 
increase citizen engagement. Policy is defined 
as a principle or rule, or set of thereof, 
adopted by counterpart institution to guide 
decisions and achieve rational outcome(s) in 
the area of good governance. Mechanisms 
includes citizen advisory committees, open 
public hearings, formal consultations such as 
comment periods, and other strategies that 
solicit and respond to public input.
The evaluation team found that the proposed 
indicator for IR1 is valid and useful to measure 
and communicate the Activity’s progress in 
developing the capacities of government to 
engage citizens. The indicator is quantitative, 
measuring the number of policies or 
mechanisms that have been adopted in 
counterpart LSGs to increase citizen 
engagement. However, the indicator did not 
capture qualitative dimensions of these policies 
and mechanisms. Some of the examples of 
these dimensions could be the capacities for 
preparation and implementation of policies, 
functional aspects, changes or improvements 
that have occurred through implementation of 
polices and mechanisms. 
IR2: INCREASED FOLLOW-UP TO OVERSIGHT INSTITUTIONS’ RECOMMENDATIONS
IR2.1.1. Financial benefit 
provided to the 
Government of Serbia 
and its citizens as a 
result of SAI audits
Baseline: 671 million 
(2018)
Target: 738 million 
(2020)
The measure here is the amount of financial 
benefits identified through SAI audits, 
presented in Serbian Dinars.
Indicators I IR2.1.1. and I IR2.1.2. refer to the 
improvement of functioning of SAI, are highly 
relevant to measure progress under GAI. 
Indicator I IR2.1.2 reflects on the improved 
institutional and technical capacities of SAI to 
oversee the performance of government 
institutions at all levels, measured by the 
number of mechanisms and practices adopted 
to implement policies (e.g., methods for 
conducting audits and communication with 
external stakeholders). Complementary to this, 
indicator I IR2.1.1. could be at the “outcome” 
level, measuring the effects of improved 
capacities through the amount of financial 
benefits identified through SAI audits. 
IR2.1.2. Number of 
mechanisms adopted by 
SAI to improve its 
capability to oversee 
government 
performance
Baseline: 0
Target: 12 (2020)
The measure here is the number of 
mechanisms SAI adopts, thus improving its 
capacities to oversee the performance of 
government institutions at all levels. 
“Mechanisms” are defined as practices adopted 
to implement policies and may include, for 
example, methods for conducting audits and 
communication with external stakeholders.
IR2.1.3. Number of 
ACA referrals to PPO 
and Misdemeanor 
courts
Baseline: 414 (2017)
Target 525 (2020)
The measure here is the number of ACA 
referrals that are made to PPOs and 
misdemeanor courts.
Indicators IR2.1.3. and IR2.1.4. are relevant to 
measure GAI efforts to enhance capacities of 
ACA, by measuring its organizational 
development (e.g., established mechanisms to 
improve oversight capability of ACA), while 
also assessing its performance and functioning 
within the anti-corruption system. 
IR2.1.4. Number of 
mechanisms adopted by 
ACA to improve its 
capability to conduct 
oversight
Baseline: N/A
Target: 5 (2019); 5 
(2020)
The measure here is the number of 
mechanisms ACA adopts to improve its 
capacity to conduct oversight. “Mechanisms” 
are defined as practices adopted to implement 
policies and may include, for example, 
processes for reviewing asset disclosure 
statements and controlling conduct of officials 
and civil servants.
IR2.1.5. Number of 
USG-supported anti￾corruption campaigns
Baseline: 0
Target: 14 (2020), with 3 
in 2018, 4 in 2019 and 7 
in 2020.
This indicator measures the number of 
campaigns supported by GAI. A campaign is 
defined as an activity aimed to raise awareness 
of the importance of accountability of 
government on all levels, which can be 
delivered through, but is not limited to 
supporting CPI and LSGs in media events, 
promotional material distribution and media 
Indicator I IR2.1.5. is relevant for GAI, 
measuring the number of campaigns it 
supports. The quantitative side of campaigns is 
important but without measuring a campaign’s 
qualitative components it is difficult to assess 
or quantify the effects of a campaign. It would 
be, for example, important to know if the 
targeted audience has an enhanced 
USAID.GOV MID-TERM EVALUATION OF THE USAID’S GAI ACTIVITY IN SERBIA | 42
INDICATOR DEFINITION FINDINGS
appearances (TV, radio, newspaper) and online 
media. The campaign methodology and 
medium of delivery will be adjusted to suit the 
selected target audience. Key messages will be 
formed to enhance target audience reception 
and understanding.
understanding of a specific topic. The MELP 
also did not include more specific indicators 
related to assistance to CPI and LSGs.
IR3: INCREASED ADJUDICATION OF CORRUPTION CASES
IR3.1.1. Number of 
corruption cases 
registered by the ERCC
Baseline: N/A
Target: 780 courts and 
4725 PPP 
(2020 total) 
The number of corruption cases for which a 
case number is assigned by the prosecutor or 
judiciary and is registered by the ERCC, 
annually. For judicial cases, the figure counts 
original filings in the four special anti￾corruption units of higher courts (Novi Sad, 
Belgrade, Kraljevo, and Niš), but not appeals. 
“Corruption” is defined as cases falling under 
Article 2, clauses 3) and 4) of the 2016 Law on 
the Organization and Competencies of State 
Authorities in Combating Organized Crime, 
Terrorism, and Corruption. To avoid 
duplication of cases counted, the data will be 
reported separately for the PPOs and the 
courts. ERCC is expected to be available to 
users in FY2019.
The proposed indicators are useful to measure 
some of the dimensions of IR3, as defined here. 
This is particularly the case to measure if and 
to what extent the ERCC has proven 
functional and delivered results. The 
assumptions under these indicators have been 
that the “ERCC is expected to be available to 
users in FY2019;” however, the ERCC was not 
established at the time of this evaluation 
(October 2020).
GAI did not introduce indicators to measure 
other dimensions and results under IR3, such 
as enhanced capacities of the target groups or 
the effects on anti-corruption education 
programs on the selected youth (e.g. Anti￾IR3.1.2. Number of user corruption Legal Clinic and its participants).
requests of ERCC data
Baseline: 0
Target: 70 (2020 total)
The number of user requests for corruption 
case related data from ERCC. ERCC is 
expected to be available to users in FY2019.