Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Report Team

This report has been compiled by Elsa Araya, Governance Global Practice; Verena Fritz, Governance
Global Practice; Daniel Rogger, Development Impact Evaluation and Development Research Group
World Bank; Kerenssa Kay, Development Research Group and Governance Global Practice; Ravi
Somani, University College London; Tesfaye Digie and Adebabay Abay, consultants advising the team.

Hisham Waly (Practice Manager, Governance Global Practice), Renaud Seligmann (Practice Manager,
Governance Global Practice) and Carolyn Turk (Country Director for Ethiopia, Sudan and South Sudan)
provided overall guidance.
ii

Acknowledgements
Ethiopia’s Ministry of Public Service and Human Resource Development (MPSHRD) requested this
analysis, and the team wishes to especially thank H.E. Ato Tagesse Chaffo, Minister; H.E. Ato Adamu
Ayana, State Minister; and Ato Tesfaye Atire, Director of the Federal Reform Program Monitoring and
Support Directorate, for their guidance. The team also appreciates the engagement of members of the
special task force formed to support the process: Tesfaye Belachew, Melaku Bizualem, Taye Wakuma,
Jemila Ahmed, and Mekonen Gashu.

We also wish to thank Berhanu Legesse, Stefan Dercon, and Jaydeep Biswas for their support during
the development of the survey instrument as well as the survey team members who collected the
data for this project: Girma Shegute, Seife Fekade, Molla Tedla, Sileshi Seneshaw, Habtu GebreHiwot,
Abiy Tessema, Mandefro Bekele, Mohammed Umer, Eskinder Tesfaye, Gashaw Yetwale, Naga Tesfaye,
Wubliker Teklu, Mitselal Gebre, Hailu Alemsellassie, Yared Abraham, and Tesfakiros Arefe, as well as
data encoders Fikir Temtem, Kidist Tesfasilasie, and Haimanot Zeleke. Administrative support was
provided by Hanna Ketselamaryam Hailu, and document editing and formatting support was provided
by Candace Gebre and Cybil Maradza, respectively.

We gratefully acknowledge financial support from the World Bank, UK Department for International
Development, and International Growth Center.
 iii

Abbreviations
ATA Agricultural Transformation Agency
BA Bachelor of Art
BPR Business Process Reengineering
BSC Balanced Scorecard
CSR Civil Service Reform
CSRP Civil Service Reform Program
DFID Department for International Development
DG Director General
EMIS Education Management Information System
ERCA Ethiopian Revenue and Customs Authority
FAO Food and Agriculture Organization
FY Fiscal Year
GDP Growth Domestic Product
GTP Growth and Transformation Plan
HRM Human Resource Management
IMF International Monetary Fund
IT Information and Technology
KII Key Informant Interviews
MAB Ministries, Agencies and Bureaus
MDG Millennium Development Goal
MOH Ministry of Health
MOT Ministry of Trade
MPRSHRD Ministry of Public Service and Human Resource Development
NCBP National Capacity Building Program
NPM New Public Management
SWOT Strength, Weakness Opportunity and Thereat
TGE Transitional Government of Ethiopia
UK United Kingdom
UNDP United Nations Development Program
UNECA United Nations Economic Commission for Africa
US United States
USAID United States Agency for International Development
USD United States Dollar
iv

Table of Contents
Report team ............................................................................................................................................................... i
Acknowledgements .............................................................................................................................................. ii
Abbreviations .............................................................................................................................................................. ii
List of Figures ........................................................................................................................................................... vi
List of Tables ............................................................................................................................................................. viii
List of Boxes ............................................................................................................................................................. viii
Executive Summary .............................................................................................................................................. ix

1 Introduction .................................................................................................................................................................... 1
1.1 Motivation for this report .................................................................................................................................. 2
1.2 Methodological approach: qualitative analysis, survey, and administrative data .............. 4

2 Recent Evolution of the Civil Service ......................................................................................................... 7

2.1 Overall staff and wage-bill size ..................................................................................................................... 8
2.1.1 Size and staff composition of the five sectors covered ........................................................................... 12
2.2 Wage and non-wage benefits .......................................................................................................................... 13
2.2.1 Pensions ...................................................................................................................................................................... 15
2.2.2 Health insurance .................................................................................................................................................... 16
2.2.3 Allowances ................................................................................................................................................................. 16
2.2.4 Transportation ......................................................................................................................................................... 17
2.2.5 Preferential participation in the housing lottery ....................................................................................... 17

3 Civil Service Reform Efforts .............................................................................................................................. 19

3.1 Brief overview of global and regional experiences with public service reforms to situate
Ethiopia’s reform approaches ........................................................................................................................ 21
3.2 Initial reform period in Ethiopia .................................................................................................................... 22
3.3 Reform processes since 2005/2006 ........................................................................................................... 24
3.3.1 Business process reengineering (BPR) ......................................................................................................... 24
3.3.2 After the BPR - current reforms ........................................................................................................................... 27
3.4 Civil servants’ perceptions of current reforms ...................................................................................... 31
3.4.1 Perceptions of the BSC ....................................................................................................................................... 33
3.4.2 Perceptions of the Change Army ................................................................................................................. 34
3.4.3 Perception on the Citizens Charter ............................................................................................................. 34
3.5 Regional comparison of government effectiveness ................................................................................ 35
3.6 Management practices within Ethiopia’s civil service ............................................................................. 36
3.6.1 Targeting ..................................................................................................................................................................... 37
3.6.2 Monitoring ................................................................................................................................................................. 38
3.6.3 Roles ............................................................................................................................................................................... 41
3.6.4 Flexibility ......................................................................................................................................................................... 41
3.6.5 Staffing .............................................................................................................................................................................. 42
3.6.6 Staff involvement ................................................................................................................................................... 43
 v

4 The State of the Service Today ....................................................................................................................... 45

4.1 Recruitment and management of staff ...................................................................................................... 46
4.1.1 Recruitment practices in the civil service ................................................................................................. 47
4.1.2 Treatment of staff ................................................................................................................................................. 47
4.1.3 Is staff promotion merit based? .................................................................................................................... 48
4.1.4 Merit-based advancement: a cross-country comparison ................................................................ 52
4.2 Drivers of staff turnover ..................................................................................................................................... 53
4.3 Staff capacities - diverging perceptions .................................................................................................... 56
4.4 Use of databases and client feedback ....................................................................................................... 58
4.4.1 Do staff make use of available databases? ............................................................................................ 58
4.4.2 Attention to clients’ feedback and complaints ........................................................................................ 61

5 Constraints to Public Service Delivery Identified by Civil Servants ............................................ 63

5.1 Five highest ranked challenges ...................................................................................................................... 65
5.1.1 Inadequate resources .......................................................................................................................................... 65
5.1.2 Motivation .................................................................................................................................................................. 71
5.1.3 Inadequate leadership ......................................................................................................................................... 74
5.1.4 Last minute/ad hoc work requests .............................................................................................................. 74
5.1.5 Delays receiving inputs from others ............................................................................................................ 75
5.2 Other constraints and challenges .................................................................................................................... 75
5.2.1 Rent seeking and corruption ................................................................................................................................. 75
5.2.2 Effective use of staff time ................................................................................................................................. 79
5.2.3 Satisfaction and perceived advantages and disadvantages/satisfaction/experience ...... 83

6 Options for the Way Forward ......................................................................................................................... 87

6.1 Recognize strengths and preserve what works ....................................................................................... 89
6.2 Options for addressing the challenge of low motivation and low prestige of the civil service ... 89
6.3 Options for further improving management practices ...................................................................... 90
6.4 Options for incentivizing performance and professional growth ............................................... 93
6.5 Ensuring that tools and systems work as intended ............................................................................. 93
6.6 Demand-side accountability for reforms and further service delivery improvements .... 94

References ................................................................................................................................................................. 96

Annex 1 Specific Issues in Focus Sectors ................................................................................................ 98

Annex 2 Methodology Notes for the Civil Service Survey .............................................................. 113
Annex 3 Methodology for Graphs and the Management Practice Indices ................................ 141
A 3.1 Methodological notes on individual graphs ........................................................................... 141
A 3.2 Management practices ..................................................................................................................... 153
vi

List of Figures
Figure 2.1 Population growth and civil service expansion .................................................................. 9
Figure 2.2 Wage bill to GDP ratio ......................................................................................................................... 10
Figure 2.3 Comparative data on wage bills relative to GDP, and general government
revenue and expenditures ............................................................................................................. 10
Figure 2.4 Annual expansion of the civil service in Ethiopia 1992-2014 .................................... 11
Figure 2.5 Public versus private employees by education levels ....................................................... 12

Figure 3.1 Timeline of reform tools ................................................................................................................ 24

Figure 3.2 The core functions of the BPR .................................................................................................... 25
Figure 3.3 Perceptions of successful reforms by tier ........................................................................... 32
Figure 3.4 Perceptions of successful reforms by sector ...................................................................... 32
Figure 3.5 Staff perception of unfair bias inherent in the subjective component of the BSC ... 33
Figure 3.6 Government effectiveness by income blocks ...................................................................... 35
Figure 3.7 Changes in government effectiveness in selected African countries – 1998-2015 .... 35
Figure 3.8 Average management scores across management dimensions ................................ 36
Figure 3.9 Average quality of management (composite measure of all management
dimensions) ........................................................................................................................................... 36
Figure 3.10 Relationship between the intensity of BSC reform implementation and the
use of performance targeting by management .................................................................. 37
Figure 3.11 Targeting ................................................................................................................................................. 38
Figure 3.12 Relationship between the intensity of BSC reform implementation and the
use of performance monitoring by management ............................................................. 39
Figure 3.13 Monitoring ............................................................................................................................................. 39
Figure 3.14 Incentives ............................................................................................................................................... 40
Figure 3.15 Roles management ........................................................................................................................... 41
Figure 3.16 Flexibility ................................................................................................................................................ 42
Figure 3.17 Staffing ..................................................................................................................................................... 42
Figure 3.18 Staff involvement ............................................................................................................................... 43

Figure 4.1 Top determinants of how staff are treated ........................................................................... 48

Figure 4.2 Importance of merit as a criterion for promotion. By tier and sector. By region ... 49
Figure 4.3 Importance of merit as a criterion for promotion. By tier and sector ................... 49
Figure 4.4 Importance of length of service in the civil service. By tier and sector ................ 50
Figure 4.5 How confident are you that you will get promoted if you perform well? ............ 51
Figure 4.6 Top 5 drivers of turnover (directors/heads) ......................................................................... 54
Figure 4.7 Reasons for wanting to change jobs (staff) ........................................................................... 54
Figure 4.8 Do you have a Management Information System (MIS) program in place and
functioning? ......................................................................................................................................... 58
Figure 4.9 “What would be the main challenges to getting the data?” ........................................ 59
Figure 4.10 To what extent do you agree that the introduction of the MIS increased the
information employees of their organization had access to regarding the state
of service delviery in the jurisdiction? .................................................................................... 60
 vii

Figure 4.11 To what extent do you agree that the introduction of the MIS has made individuals
work harder because they had more information about what works? .................... 61

Figure 5.1 Top 5 most common challenges to completing tasks effectively ............................ 65
Figure 5.2 Success rate of large ICT public sector projects .............................................................. 66
Figure 5.3 Out of every 10 experts how many have access to a computer. By tier and sector .. 67
Figure 5.4 Out of every 10 experts how many have access to a computer. By region ......... 67
Figure 5.5 Out of the five working days, how many days is the internet access good enough
to check email? ................................................................................................................................... 68
Figure 5.6 Please indicate the extent to which you agree with the following: Network
interruptions are not a challenge towards me being able to complete my tasks ... 68
Figure 5.7 Please indicate the extent to which you agree with the following: software and
applications are not a challenge towards me being able to complete my tasks ... 69
Figure 5.8 During a typical working day, how many hours is there electricity. By tier and sector .. 70
Figure 5.9 During a typical working day, how many hours is there electricity. By region ... 70
Figure 5.10 ‘Imagine when you started your motivation was 100. What number would you
say it is now relative to that?’ By tier of government ...................................................... 71
Figure 5.11 Proportion of civil servants in that organization that are neutral or positively
satisfied with their job ...................................................................................................................... 73
Figure 5.12 What percentage of your time is used on tasks which are not part of your
main job .................................................................................................................................................. 75
Figure 5.13 Comparative progress in control of corruption, Ethiopia and Rwanda ................ 76
Figure 5.14 How does the Change Army deal with rent seeking? Heads of organization only .. 77
Figure 5.15 Which of these tools do you think are most effective at limiting rent seeking? .. 78
Figure 5.16 Hours worked by civil servants per week (staff and directors only) ...................... 80
Figure 5.17 “Which of these groups regularly give you tasks as part of your formal work duties?” ... 81
Figure 5.18 Perception of proportion of meetings that stick to a fixed agenda ........................ 81
Figure 5.19 Perception of proportion of meetings that achieve their goals in minimum time .. 82
Figure 5.20 Perception of proportion of meetings that are of substantive use to daily tasks .. 82
Figure 5.21 “To what extent would you say you are satisfied with your experience of the
civil service?” .......................................................................................................................................... 83
Figure 5.22 What would you say are the advantages of working in the civil service relative
to the private sector? Job security. By tier and sector ................................................... 84
Figure 5.23 What would you say are the advantages of working in the civil service relative
to the private sector? Opportunities for learning. By tier and sector .................... 84
Figure 5.24 What would you say are the advantages of working in the civil service relative
to the private sector? Career opportunities. By tier and sector ................................ 85
Fgiure 5.25 What would you say are the advantages of working in the civil service relative
to the private sector? Prestige/ social status. By tier and sector ............................. 85
Figure 5.26 Favored aspects of working in the civil service ................................................................. 86

Figure A.1.1 Increase in cereal yields, Ethiopia compared to East Africa, 1992-2014 ............. 98
viii

List of Tables
Table 2.1 Public and private sector pension plans in Ethiopia .................................................................. 15

Table 4.1 Perceived criteria for advancement in the civil service across countries ...................... 52

Table 5.1 Level of satisfaction with job in the civil service: a cross-country comparison ............ 72
Table 5.2 Factors that stop civil servants to fight rent seeking. All civil servants ........................... 79

Table A1.1 Percentage of farmers’ qualitative rating on the extent of DA-FES helping farmers
to address technical problems .............................................................................................................. 101
Table A2.1 Total number of woredas per region and number of woredas sampled ......................... 127
Table A2.2 Pre-survey changes to woredas based on identified security risks or infeasible logistics .. 128
Table A2.3 Woredas dropped during the survey period due to security challenges ........................ 129
Table A2.4 Final sample list of woredas and original itineraries ................................................................. 129
Table A2.5 Original team itineraries ........................................................................................................................... 131

List of Boxes
Box 3.1 The Indian experience with a ‘Right to Public Services Act’ ................................................ 30

Box 6.1 Problem Driven Iterative Adaptation (PDIA) as a way to generate specific reform options .. 91
 ix

Executive
Summary

1. This report identifies key challenges facing the While the past two decades have seen important
Ethiopian civil service today and suggests avenues achievements made, the government seeks to
for further reforms. Ethiopia’s civil service has further strengthen its economic governance and
so far developed at a rapid pace and faces a service delivery.
number of constraints, challenges, and new or
expanding demands. Continuing reform efforts 2.The analytic effort that this report reflects
and building on past achievements is critical was triggered by a request from the Ministry of
for the successful implementation of Ethiopia’s Public Service and Human Resource Development
Growth and Transformation Program (GTP II). (MPSHRD). The Ministry sought to identify
x

challenges and options to target and pursue concerning the number of civil servants, as well
further reforms and efforts to strengthen the civil as salaries and benefits.
service and its contributions to service delivery.
The report is based on several sources: (i) the 3. The focus is on five key service delivery sectors:
Ethiopian Civil Servants Survey (ECSS), which agriculture, education, health, revenue collection,
was undertaken in June through September and trade—all critical areas for state activity. The
2016 and targeted professional level staff and analysis encompasses an assessment of previous and
directors or heads of federal ministries or ongoing reforms, and takes a look at the constraints
agencies, regional bureaus, and woreda offices; and overall quality of current management practices.
(ii) a series of key informant interviews with Where possible, the report also compares the civil
officials and stakeholders, as well as focus group service in Ethiopia to that of other countries, based
discussions with members of ‘Public Wings’; and on information gathered from similar surveys and
(iii) administrative data and information, e.g. other comparative information.

Recent evolution of the civil service

4. The civil service has expanded rapidly, with an and expenditures, wages account for about 25
annual average growth rate of over 7 percent in percent of overall spending. This is in line with
the number of employees during the five-year regional comparators and signals limited room
period between 2008/09 and 2013/14. This was for expansion, at least prior to an increase in the
motivated by the government’s decentralization, level of revenue collected. Civil service wages have
and its fast-tracked approach to enhancing been adjusted at irregular intervals for inflation,
service delivery as part of its pursuit of the MDGs. and remain among the lowest in the region. Non-
A swift expansion of post-secondary education wage benefits (pensions, travel allowances) exist,
that aimed to increase the number of qualified but are relatively small, and in the case of travel
entrants to the civil service also underpinned its allowances, tend to be below actual costs for
growth. regular civil servants.

5. Ethiopia has managed to expand civil service 6. Surveyed civil servants and key informants
numbers without an excessive expansion of the (including directors) indicate that salaries are an
overall wage bill. The wage bill has been contained important issue of concern. About 70-80 percent
at around 5 percent of GDP, which is at the low end of respondents are dissatisfied with their salaries;
observed internationally. At the same time, given and wages and benefits are among the most
the country’s still low overall revenue collection frequently cited reasons for wanting to change
 xi

jobs. Managers in ministries and agencies in time. While public sector staff earn more
turn are concerned about their ability to attract than private sector employees on average, the
qualified staff, especially in sectors where pay is difference is limited and has been shrinking when
comparatively lower. comparing government and permanent private
sector employees. Moreover, when controlling
7. Comparisons with average private sector for job characteristics, the private sector has
wages suggest that Ethiopia, as many other surpassed public sector pay in recent years. It is
countries, has a public-sector wage premium, also important to consider that both public and
but this has narrowed or even reversed over private sector wages in Ethiopia are among the
lowest globally.

Civil service reform efforts and perceptions of current

management practices
8. Since the 1990s, the Government of Ethiopia 10. Among the reform tools deployed, the Change
has deployed a series of reform tools, aimed at Army – which was first established at local levels
improving the civil service and service delivery in the health care sector – is perceived, relatively,
performance. In particular, these include the as the most successful, in particular at woreda
Business Process Reengineering (BPR), the and regional levels (during qualitative discussions,
introduction of a Balanced Scorecard (BSC) system however, it was noted that it may be politicized.)
to set targets and monitor performance, and the At the federal level, the BSC registers as the most
roll-out of a Change Army.1 The government has successful reform tool. Among sectors, the Change
also directed all public sector institutions to adopt Army is particularly utilized by health sector staff,
and publish Citizens Charters. These tools have and relatively less so by staff working with the
been embedded into the wider Civil Service Reform Ministry or Bureaus of Trade. However, Citizens
Program pursued since the mid-1990s. Charters are seen as relatively less successful
thus far. Citizens Charters are more frequently
9. Among low income countries, Ethiopia stands raised as a successful reform by staff in the trade
out as having followed civil service reforms and revenue organizations, which interface with
intensively over the past two decades, and with citizens and businesses around specific ‘business
some success. The wide-ranging changes carried processes’, such as paying taxes or issuing a license.
out include the large-scale decentralization
and strengthening of sub-national levels of 11. The active participation of high-level political
government, as well as the successive roll out leadership has made a major contribution in
of reforms across all levels of government. driving the reform agenda throughout the civil
Governance indicators suggest that government service in a relatively short period, while also
effectiveness – as a proxy measure of civil service posing challenges. The urgency with which the
performance – improved between the late 1990s government has pursued reform has at times
and the early 2000s; however, since then, further overloaded the civil service with reform tools.
progress on this measure appears to have stalled. Furthermore, reforms have tended to be rolled-

1 Please refer to Chapter 3 for a detailed description of each reform.

xii

out in a uniform manner, with limited means for enable civil service organizations to achieve their
flexibility, adaptation, and integration. Reforms were objectives. The two stronger performing areas are
not complemented with the greater involvement of monitoring and targeting (on average rated 3.5 and
professional staff; use of incentives to encourage 3 (on a scale from 1 as the weakest, to 5 as the best),
performance improvements; and strong demand- while ‘staffing’2 is intermediate (2.7), and other scores
side pressure for reforms or improvements. – ‘roles’,3 ‘flexibility’4 (2.5), and ‘staff involvement’5 (2)
are all assessed as relatively low. While just a snapshot
12. As a complement to tracing the public sector in time, these survey results suggest that the BSC
reform efforts that have been made, the survey reforms have had traction, while those aspects of
gauged staff perceptions of current management the BPR reforms aiming to strengthen autonomy and
practices. Management practices are methods that flexibility for staff have had less impact.

State of the civil service today

13. This analysis sets out to assess the degree to criteria for how staff are treated by only 37 percent
which civil service staff perceive performance of respondents, followed by 26 percent who
management and promotions within the civil identify political connections as most important.
service to be merit based; feel capable of carrying In general, opportunities for promotion remain
out their duties; and use databases and client relatively scarce, which is one of the factors that
feedback to improve their performance. A merit- drive staff turnover.
based civil service has long been seen as critical
to effective public sector performance, while staff 15. Whether civil service staff have adequate
capacity and attention to client feedback are capacity becomes a notable point of divergence
other important ingredients for delivering services between the perceptions of focus group discussants
reliably and effectively. and key informants, and the self-perceptions
of civil servants as captured in the survey.
14. Staff perceived their treatment to be relatively Ethiopia has rapidly expanded post-secondary
merit and performance based, but not uniformly education, and has also invested in the training
so. Regional variation exists: While 91 percent of civil service rules and processes. According
of respondents in Tigray consider merit and to survey results, a large majority of bureau and
performance to be the key factor that determines ministry staff see their skills as adequate for their
their treatment, only 23 percent of Gambella- jobs. However, discussions with key informants
based respondents do. Other perceived factors suggest that the very rapid expansion of higher
relevant to how staff are treated include their education has resulted in shallower education
relationship with managers; political, educational, received, including a lack of practical training, e.g.
and ethnic connections; and length of service. in fields such as agriculture, health, and education.
Notably, while Addis Ababa performs well on most Public Wing members also express concerns that
metrics, merit and performance are seen as key federal ministry staff often do not have strong

2 Measuring management attention to attracting and developing talent.

3 Measuring the level of autonomy staff are given in undertaking tasks.
4 The ability to confront new demands and/ or ways of working.
5 The ability of staff to become involved in operations and contribute towards decisions.
 xiii

knowledge of their respective fields. Relatively use and usefulness of management information
low salaries combined with perceptions of low systems (MISs) is still very uneven. Civil servants
prestige make it difficult to attract and retain report that the three most important sources of
high caliber professional staff in the civil service; feedback are colleagues, citizens, and supervisors
while political appointees rotate too frequently (all above 90 percent). Information from MIS
and among different technical fields. At the systems is more widely available at the federal
implementation level, competency assessments and regional levels than for woreda level offices.
have been deployed for some front-line service Importantly, for agriculture and education sectors,
staff, and these also indicate challenges. staff observe that the data captured in the systems
is not useful. A further challenge is that even in
16. Regarding the importance of feedback offices with MISs, network problems often disrupt
mechanisms and information on service delivery access (this is especially noted by health sector
performance, civil servants emphasize the staff at the regional level).
importance of feedback from clients, while the

Constraints to working effectively

17. The analysis further explored the current In addition, phone service is uneven, further
constraints to effective service delivery from the limiting the ability of staff to communicate and
perspective of civil servants and managers, and transmit information and data. Staff perceptions
several findings stand out. Firstly, staff struggle about inadequate resources are consistent with
with inadequate resources to do their work. fiscal data, which indicates that operations and
Secondly, staff are demotivated to a significant maintenance spending in Ethiopia is low.
extent. Demotivation is more widespread in
Ethiopia than in other countries where similar
surveys have been carried out, while it is unevenly 19. The second main constraint identified is a lack
distributed across organizations. A third obstacle of motivation; and this is particularly pronounced
is inadequate leadership. Further constraints are at the federal level (20 percent of staff cite it
posed by ad hoc work requests and staff having as the most important constraint). For regional
to wait for inputs from others. staff, a lack of motivation is the most important
constraint (14 percent), as well as at woreda
18. Inadequate resources to carry out assigned level (18 percent). Responses to an open-ended
tasks are the most frequently identified obstacle question about what contributes to declining
by a wide margin. Resources are particularly motivation suggest that contributing factors
constrained at the woreda level (48 percent), and are dissatisfaction with salaries and benefits,
somewhat less so at the regional and federal frustration over a lack of operational resources,
level (23 percent). For the federal level, staff in the limited prospects for professional development
agriculture sector more frequently raise resource and promotion, as well as inadequate leadership
constraints as an issue, more so than health sector and a perception of the civil service’s low level
staff. The survey provides a drill down, in particular of prestige (which is also greater at the federal
into the degree to which IT systems are available than at the sub-national level). Importantly,
and function as intended. Usage is limited and levels of motivation vary significantly between
frequently disrupted, especially at woreda levels. organizational units.
xiv

20. Inadequate leadership is raised most higher for the health sector at the federal level
frequently at the woreda level (by 19 percent), (20 percent), and particularly low for the revenue
followed by the federal (15 percent) and then the sector. Among the hallmark features of earlier
regional level (9 percent). The issue appeared in successful developmental states has been a
both the key informant interviews and survey. At well-regarded civil service, able to attract ‘the
the federal level, inadequate leadership was flagged best and brightest’ through competitive entrance
especially in the trade and education sectors; and assessments. While Ethiopia has been relatively
at the woreda level also for the revenue sector. It successful in making the civil service merit and
is rarely flagged among regional-level staff. The performance based, its exceptionally low social
practice of selecting decision-making staff as status is an issue that needs to be addressed.
political appointees from the director level upwards
may play a role in this regard, in particular when 24. Most civil servants do not consider rent
incentives are not fully aligned, and managers may seeking to be a major constraint to their ability
be oriented more towards their political roles and to work effectively, and they point to the Change
opportunities rather than viewing performance as Army as one of the most effective mechanisms
a key criterion for further promotion. for addressing corruption. Less than 3 percent
of civil servants expressed that corruption or rent
21. Ad hoc work requests, and having to wait for seeking hinders their ability to work effectively.
inputs from others pose some burden. Around The survey did not entail questions about whether
30 percent of civil servants’ time is used for ad civil servants had observed or suspected any
hoc tasks that are not part of their main job, with corruption in their offices or in other parts of the
variation mainly across regions rather than sectors. civil service. The Change Army is considered to be
effective in identifying solutions to rent seeking
22. Regarding other aspects of time management, in the civil service by 65 percent of organization
meetings appear to not be well planned and used heads, and 55 percent of all civil servants. One
in a targeted way, especially at woreda levels. At key exception is the agricultural sector at the
the woreda level, 52 percent of respondents think federal level, where the Change Army is perceived
meetings achieve their goals using the minimum to be ineffective in this regard. Somewhat
amount of time in less than a quarter of all disconcertingly, a very low share of civil servants
meetings. Regarding the usefulness of meetings, 37 (3 percent) see Federal and Regional Ethics and
percent of federal respondents believe meetings Anti-Corruption Commission investigations as
are useful in more than 75 percent of cases, but effective for limiting rent-seeking. Furthermore,
45 percent of woreda staff believe meetings are consultations with Public Wings – i.e. the main
useful in less than 25 percent of cases. Meetings interaction between government officers and
that take excessive amounts of time or are not organized citizens – are not widely seen as
useful to the tasks staff are expected to perform addressing corruption. This is consistent with
amount to lost resources. the perceptions of Public Wing members who
observe that there seems to be a limited appetite
23. A key challenge for the civil service, to discuss corruption challenges openly.
particularly at the federal level, is the low level
of prestige. Only 8 percent of federal-level civil 25. Additional sector-by-sector findings are
servants view the civil service as prestigious, presented in Annex 1 of this report, covering
increasing to 11 and 14 percent at the regional and the five sectors of focus—agriculture, education,
woreda level respectively. Ratings are somewhat health, revenue, and trade.
 xv

Emerging recommendations

26. Based on the discussion in the preceding collaboration with other ministries, can build on
chapters, Chapter 6 outlines options for pursuing the significant reform efforts that have already
further reforms. The suggestions made are mindful been made, and that indicate Ethiopia is well
of resource constraints and seek to minimize prepared to pursue further reforms once their
suggestions that would have a substantial fiscal direction has been agreed upon. The proposed
cost. In moving forward, the Government of implications draw on the results of the undertaken
Ethiopia, and specifically the Ministry of Public quantitative and qualitative surveys.
Service and Human Resource Development in

Recognize strengths and preserve what works

27. Ethiopia’s public sector stands out for two important assets that many other low income
main strengths: (i) it has achieved a lot with countries have struggled with. Thus, in seeking to
very limited resources, and (ii) there is a fairly address challenges such as staff motivation and
good alignment between its ability to make and reducing delivery bottlenecks, attention should
implement decisions, so that its deployment also be given to maintaining and further
of practices such as targeting and monitoring developing these strengths.
is relatively effective. These two aspects are

Options for addressing the challenge of civil service’s low motivation and
prestige

28. The survey indicates that low motivation meant to play an important role in the economy
and low prestige are widespread. For a country and society. The following are potential options
pursuing a ‘developmental state’ model, such for improving motivation:
low motivation is concerning, since the state is

i Establishing a system based on competitive ii Allocating greater funds for key operational
entry exams. Such processes have been tasks and systems. The findings presented
used in Asian developmental states such in Chapter 5 indicate that having sufficient
as South Korea and Japan, as well as in funds and systems (such as functioning IT
parts of continental Europe. They create a systems) are the most important constraint
greater sense that only the best and brightest to performance and the effective delivery of
can join key positions in the civil service, expenditures), as it can contribute both to
and hence contribute to greater prestige. improving motivation as well as strengthening
Specialized competitive exams can be used service delivery.
for recruitment into sectors.
xvi

iii Leadership selection and leadership training iv Seeking to understand further why motivation
focused on motivating staff. The importance varies so greatly between organizations
of good leadership to motivate performance within Ethiopia. As addressed in Chapter
is well recognized. Inadequate leadership 5.1, the dispersion of motivation among
appears to be an important constraint, staff in different organizational units is very
especially in the health and revenue sectors noticeable. This merits further exploration,
where concerns about leadership are voiced and a targeting of those organizations within
more frequently. each sector where performance is significantly
weaker than the average.

Options to further improve management practices

29. The relatively standardized reforms staff involvement lower. Going forward, a more
pursued over the past 10 to 15 years (e.g. BPR, specific approach that addresses the particular
BSC, Change Army) were rolled out across all constraints of sectors and organizational units
sectors. With regard to the implemented current would be useful, and contribute to empowering/
management practices, civil service staff tended motivating staff without losing the benefits of
to rate target setting and monitoring higher, target setting and monitoring.
while rating practices that promote flexibility and

i Moving from general reform models to ii Consider rebranding the Change Army as
identifying and enabling sector and task- quality circles: The case team workers seem
specific improvements: As is described comparable to “quality circles” in that they
further in Annex 1, the specific challenges are a formal, institutionalized mechanism
faced and the role and functions of staff for participatory problem solving, peer to
differ across sectors. Making public service peer learning and continuous improvement
reform efforts more specific can entail at of performance. However, the qualitative
least two avenues: (a) developing further interviews and dissemination discussions
reform efforts in close collaboration indicate that the system is perceived to be
between the Ministry of Public Service and politicized and there is too much emphasis on
specific sector ministries, while at the same attitudes and not enough on knowledge and
time keeping an eye on overall coherence, constructive problem solving. Rebranding the
e.g. in terms of staff grading, pay scales, and Change Army will encourage team members
recruitment standards; (b) within individual to cooperate with one another to focus
sectors, eliciting ideas from staff of potential on finding solutions for improving service
organizational improvements could result delivery and to steer away from fault finding.
in valuable contributions to improved
service delivery. In addition, this may also iii Facilitating learning across organizational
contribute to an increased perception of units in the same sector/level of
staff involvement. Within sector ministries, government: The Change Army addresses
Reform Directorates can play a critical role peer learning and the dissemination of
in shaping public sector changes. good practices within organizational units.
 xvii

However, to address the issue of the significant v Better use of staff time should receive
Differences that exist between organizational attention to improve overall public sector
units within the same sector in terms of efficiency. The wage bill pays for a set of
motivation and management scores, greater staff time, skills, and efforts. As the survey
learning across units should be facilitated. indicates, ad hoc tasks and meetings that
are not sufficiently pertinent and efficient
iv Consider enhanced staff involvement in reduce public sector efficiency, in particular
policy formulation and monitoring functions. at woreda levels. Investments in leadership
Greater autonomy and flexibility, as well as training can include guidance on how to use
greater involvement in policy and monitoring time and how to conduct meetings efficiently
functions can contribute to staff satisfaction, and in a way that supports organizational
and facilitate the attraction, retainment, and performance.
motivation of well-performing staff.

Options for incentivizing performance and professional growth

30. Findings both from the survey and the while a performance management system is in
qualitative analysis suggest civil servants perceive place, it does not offer much in terms of rewards
limited opportunities for professional growth. for good performance, and poor performance is
Staff believe that opportunities for promotions not dealt with systematically.
are slim, as well as for training and learning. Also,

i Provision of more training and learning as poor performers, and recognition should
opportunities for staff should be a primary be given to good performers. While the BSC
consideration. Staff should be more widely reforms have put in place regular performance
encouraged to scale up their skills, knowledge, assessments, survey responses suggest that
and competencies. This can take various follow-up remains limited. A lack of follow
forms, e.g. class-room based learning, but up may make staff reluctant to make strong
also temporary assignments of different tasks, investments in better performance. At the
as well as on the job mentoring. As the report same time, it is important to ensure fairness
notes, staff seem to be reluctant to admit to and objectivity in assessing performance.
skills gaps, despite these being widely noted
by key informants, as well as by Public Wing iii Considering some expansion of promotion
representatives. An approach that combines opportunities. As is discussed in Chapter 4,
competitive entry, stronger assessments of while staff believe promotions are largely
actual skills, and greater opportunities for merit-based, there is also a perception
learning can contribute to an upgrade of staff that promotion opportunities are extremely
competencies, as well as contribute limited, and this contributes to declining
to job satisfaction and motivation. motivation and turnover. Some broadening
of promotion opportunities should be
ii Action needs to be taken for staff identified considered in pay and grading exercises.
xviii

Ensuring that tools and systems work as intended

i Pay attention to deployment and usability staff to provide trouble-shooting and updates
of ICT systems. Staff clearly struggle with to software to ensure continuous usability.
systems being available and useable; and the
latter includes the ability to reliably make ii Simplify reform tools. Interventions that are
phone calls. Investing in functioning ICT simple are more likely to be implemented
systems may have the greatest pay-off for sustainably. The BSC is a good tool in
revenue collection, followed by agriculture principle, but it seems complex and difficult
as a key productive sector. The required to use effectively. There appears to be some
investments are not just in hardware and scope for simplification on the one hand, and
software, but also in the provision of requisite for automatization on the other.

Demand-side accountability for reforms and further service delivery

improvements

31. Citizens’ feedback mechanisms that work are survey results suggest that the level of activity,
important for strengthening the service delivery seriousness of engagement, and usefulness of
orientation of the public sector. As discussed in Public Wing consultations is still variable and not
Chapter 3, Citizens Charters, i.e. the tool that was fully developed.
expected to provide citizens’ power to know and
ask for their rights, is considered to be a relatively 32. The following avenues could be pursued to
less successful reform tool by implementers. further strengthen accountability and effective
Furthermore, the discussions held as well as feedback mechanisms:

i Surveys of civil servants such as those carried ii Continuing efforts should be made to
out for this report should be complemented develop Citizens Charters and effective
by surveys of service users. This would complaints mechanisms. Citizens Charters
be extremely valuable for understanding appear to be less effective than desirable,
whether mismatches in perceptions exist. while some progress appears to have been
Findings could then be fed back to public achieved already on which further efforts
sector managers to take action and monitor can be built. Charters should be clear and
for future improvements. Such surveys could realistic, e.g. in terms of processing times.
also serve as an incentive for identifying and The government can also consider making
rewarding particularly well performing teams. Citizens Charters legally binding, as has
Efforts could be piloted in selected urban been done in India through the adoption of
areas to limit costs. a ‘Rights to Public Services Act’.6 Monitoring
patterns of complaints and how complaints

6 This would require ensuring that resources are available to service delivery units to stay reliably within targets set in Citizens
Charters, or else such an approach could trigger substantial litigation.
 xix

are addressed can help to strengthen these way as prescribed by Citizens’ Charters and
and to incentivize effective follow-up. effectively in terms of costs and benefits.

iii Opportunities for greater public-private iv Improved communication: The government

partnership for effective service delivery. may consider the need to communicate more
A number of opportunities for enhanced explicitly about what public sector reforms
partnership between public and private are being pursued, why, how, and with what
sectors exist. This includes opportunities to expected results. This will help stakeholders to
further strengthen Public Wing consultation be more attentive to what is being done, and to
mechanisms, including in terms of the agreed provide feedback and proposals accordingly.
upon decisions and their implementation. It
may be relevant to consider arrangements that v Broaden opportunities for feedback: The
enable more issue-specific discussions, either main feedback channel from citizens/
through the formation of sub-committees service users to specific organizational units
(e.g. on cancer care in the health sector), or currently are complaints mechanisms, e.g.
through the dissemination of agendas for in the form of boxes to collect complaints,
upcoming Public Wing meetings. Moreover, and then follow up by staff and managers
there are areas in which the government as appropriate; as well as Public Wing
can consider to ‘do less’ or crowd in private meetings. It may be helpful to broaden the
sector contributions more, e.g. through feedback being solicited from ‘complaints’
involving the private sector in the continual to ‘suggestions, ideas, and complaints’. This
training of specialists. Revisiting earlier BPR way, service users would feel encouraged to
efforts towards bureaucratic streamlining also contribute ideas of what could be done
and simplification – in consultations with better, and to provide additional channels for
citizens and the private sector – can also ‘bottom-up’ communication.
enable services to be delivered in a timely
1

1
Introduction
 2

1.1 Motivation for this report

33. This report seeks to identify what factors central planks of the development strategy. The
emerge as obstacles to further service delivery public service stands at the center of it, having a
improvements, and how these might be major role in how services are delivered to citizens
addressed. Ethiopia is pursuing shared growth and and how growth is accelerated and shared.
poverty reduction through a developmental state
model. Accordingly, transforming and building the 34. Much has been achieved over the past
capacity of the state has been a core development two decades. Ethiopia has either met or made
priority of the Government of Ethiopia (GoE) significant progress towards all eight MDGs.
since the mid-1990s. The current development Access to education has expanded rapidly, with
strategy of Ethiopia, the second phase of the gross enrollment at the primary level reaching
Growth and Transformation Plan (GTP II), aims 100 percent by 2014. The number of agricultural
to propel Ethiopia into a middle-income country and health extension workers has significantly
by 2025 through a focus on agricultural and increased, and development outcomes have
rural development, industrial base expansion, improved. Moreover, Ethiopia has enhanced
government-led infrastructure development, and service delivery while still having very limited fiscal
investment in social and human development. resources at the government’s disposal, and while
Good governance and state transformation remain pursuing a significant expansion of infrastructure.
3

The government has also taken up a number of Reengineering (BPR), Balanced Scorecard (BSC),
public sector reform initiatives (discussed in Citizens Charter, and the Development Change
Chapter 3), showing an impressive capacity to Army (hereafter referred to as Change Army).8
translate policies into action, in particular when These have been successively introduced to
compared to other neighboring countries where improve service delivery, and enhance transparency
‘implementation gaps’ have been diagnosed as and accountability in the civil service.
very significant.7
37. The Ministry of Public Service and Human
35. Still, more remains to be done. The Resource Development (MPSHRD) has been
reforms and resulting social and economic given the mandate to lead further reform
development have increased the demand for efforts. In collaboration with other relevant
even better service delivery, transparency, and institutions, it seeks to further develop human
accountability. Furthermore, the rapid scaling up resources, strengthen processes and systems,
of services has affected quality, such as actual and address rent seeking and corruption risks in
learning outcomes in schools. Thus, while the the civil service. The MPSHRD, along with others,
progress made thus far is impressive in many spearheaded past reform efforts such as the Civil
ways, the government believes that public sector Service Reform Program in 1996, which has now
capabilities should continue to strengthen and been implemented over a period of twenty years.
improve in line with its own expectations and
plans for reaching middle-income status. In 38. The MPSHRD requested that the World Bank
addition to capabilities, strengthening integrity is conduct an independent review of the different
a key concern: following the elections in 2015, the civil service reform tools,toassesswhat has been
GoE has highlighted the fact that rent-seeking achieved, and what to consider next. The review
attitudes and behavior in the civil service are will take stock of what has been done, identify
increasingly becoming binding constraints for its remaining and potential new challenges, and draw
effective delivery of services and the achievement lessons, as well as suggest recommendations on
of the country’s development goals. how to move further ahead in the coming years
to foster a fair, responsible, efficient, ethical, and
36. The civil service reforms pursued in Ethiopia transparent civil service that promotes accelerated
to date have been wide ranging, and have and sustained economic development and
included the adoption and implementation contributes to the building of democracy and
of a number of tools. Reforms have included good governance.
significant decentralization, first to regions and
then to districts (woredas), alongside core Human 39. In response to this request, a World Bank
Resource Management (HRM) and Public Finance Group team, led by staff from the Governance
Management (PFM)-focused reforms, and efforts Practice in collaboration with staff from the
to strengthen leadership as well as orientation Research Department (DEC), have undertaken
towards serving the needs of citizens. The civil a multi-pronged analysis. This has included
service has also seen a rapid expansion to achieve qualitative work, as well as the implementation
a significant broadening of service delivery. Key of a survey of civil servants at federal, regional,
reform tools deployed include Business Process and woreda levels. In line with the request from

7
See e.g. the SCD for Uganda (2015).
8
Another reform being pursued is Kaizen, i.e. the originally Japanese approach to effective production. This is mainly applied
to enterprises (private and public) and is not included in the discussions in this report.
 4

MPSHRD, the focus focused on five sectors, stakeholders, as well as an analysis of primary
namely, agriculture, education, health, revenue and secondary data and information, followed by
administration, and trade. The review involved a survey of civil servants, based on a questionnaire
key informant interviews (KIIs) with relevant developed in close collaboration with the Ministry.

1.2 Methodological approach: qualitative analysis,

survey, and administrative data

40. This report is based on the following main data on the perceptions and experiences of civil
sources: (i) a survey of Ethiopian civil servants, servants, and on the key obstacles civil servants
undertaken in summer 2016, (ii) interviews with face as they perform their duties to the best of their
a range of stakeholders – including heads of abilities, and ultimately on the provision of public
agencies, officials from the Ministry of Public services. The survey utilized random sampling
Service, researchers, and members of Public techniques at all levels of sample selection to
Wings, and (iii) administrative data and selected gain a representative sample of officials across the
secondary sources. The report also highlights government sectors studied. This type of primary
some comparative aspects regarding the size of data provides an unfiltered snapshot of the day-
the public service and the current wage bill, as to-day operations of the civil service at all three
well as on the civil service surveys undertaken in tiers of government. The questionnaire for the
several other countries. survey was developed based on initial discussions
and in close collaboration with the Ministry of
41. As part of the analysis, the report triangulates Public Service and Human Resource Development.
information from the different sources (for Interviews were carried out in June to September
example, the views of civil servants about the 2016 at the federal level, in all 11 regions, and in
availability of working resources, fiscal data on a selection of woredas.
allocations for operations, and maintenance
expenditures across all levels of government). 43. Survey coverage and limitations. The survey
Furthermore, as Ethiopia has undertaken extensive from which this report draws from covers all
civil service reforms over the past 20 years, the three levels of government (federal, regional,
analysis focuses on perceptions and views about and woreda). All 11 regions (including Addis
what these reforms have delivered, and how Ababa), and 66 woredas were covered. 9 The
they contribute to the various objectives of a focus of the analysis is on five sectors as outlined
well performing public service, such as ensuring above, which include two service delivery sectors
effective processes, transparency for citizens, and (education and health), two growth-related sectors
promoting integrity. (agriculture and trade), and one core institution of
the state (revenue administration).10 The detailed
42. The purpose of conducting a survey of the methodology used to develop the questionnaire,
Ethiopian civil service has been to gather micro- woreda selection, and other methodological

9
Woredas were sampled for eight regions, but not for Harari, Addis Ababa, or Dire Dawa.
10
Of course revenue administration can also have important growth implications, but its primary function is to enable a
financing of public institutions and tasks – while minimizing the negative impact on growth.
5

aspects are reflected in Annex 2. In addition to 47. Initial consultations on the findings of the
the survey, key informant interviews and focus report were held with the Minister of Public
group discussions were carried out at the federal Service on March 10, 2017. The team will be keen
level and for the city of Addis Ababa. to receive comments on this draft full report.
Further consultations and dissemination efforts
44. The survey included modules and questions will be organized in consultation with the Ministry.
directed at the three types of civil servants
sampled: (i) civil servants in non-managerial 48. This report summarized the findings from
positions (employee-level track); (ii) civil servants the work undertaken and proceeds as follows:
in managerial positions (director-level track); and Chapter 2 covers the overall evolution of the civil
(iii) organization heads (political appointee-level service, provides information on the wage bill,
track). While some modules were administered to as well as individual-level wages and non-wage
all three tracks, some were designed specifically to benefits, and discusses public versus private
be administered to one of the three types of civil sector wages. Chapter 3 covers the main civil
servants to ensure questions remained relevant service reform tools that have been deployed
to each broad level of staff within the Ethiopian over the past two decades, including the BPR,
civil service. Throughout the report, it is noted BSC, Change Army, and Citizens Charters,
which responses come from the directors and as well as associated changes, such as the
organization heads specifically, and which are establishment of ‘Public Wings’. Section 3.6 then
responses from the sampled non-managerial staff. turns to how current management practices are
assessed by civil servants. Chapter 4 considers
45. Several limitations should be noted: while the the extent to which treatment, and advancement
survey includes woreda-level sector offices, front- of staff are perceived to be merit based, as well
line service delivery staff such as teachers or nurses as issues of turnover, staff capacities, and use
were not included. An important further limitation made of existing databases and other feedback
is that the survey needed to be limited to civil mechanisms. Chapter 5 discusses key constraints
servants and could not also collect mirror data from to public service delivery as perceived by
service delivery users. Such an approach would be civil servants; as well as other issues such as
desirable, but carries significant additional costs. perceptions about how rent-seeking can best
Finally, a number of caveats apply to comparative be constrained, how well staff-time is managed,
information about civil service dimensions and as well as levels and aspects of satisfaction.
perceptions in other countries; these are noted in Based on this range of issues discussed, Chapter
the respective sections. 6 highlights key implications and options for
further reform resulting from the analysis. Annex
46. Despite these limitations and caveats, the 1 provides specific sector-by-sector information
analysis adds significant new information and and results, while Annex 2 provides a detailed
insights into service delivery challenges in Ethiopia, overview of the methodological approach for
which can be utilized for efforts at strengthening the survey.
civil service performance going forward.
6
7

2
Recent Evolution
of the Civil Service
 8

2.1 Overall staff and wage-bill size

49. Following the enactment of the new urban areas. As a result, the implementation of a
Constitution in 1994, the government shifted to wide range of policies and programs was assigned
a federal state system by sharing powers between to the regions. As regionalization and later woreda
the federal government and the ethnically decentralization were being carried out, it was
organized regional state governments. This was clear that the capacity to deliver public services
initiated to promote equity across regional states varied widely. All regions had huge staffing and
and rectify the neglect of rural areas relative to communication deficits. The government’s strategy
9

for addressing these issues centered mainly on the from the mid-1990s to the mid-2000s (see Figure
Civil Service Reform Program. 2.1). This rapid expansion of the civil service was
associated with a very large volume of recruitment
50. The adoption of these decentralization policies to support regional and woreda decentralization
increased the need to expand the number of civil and the expansion in service delivery. As reflected
service staff. As a result, over the past decade the in Figure 2.4 below, in some recent years, the civil
civil service in Ethiopia has seen a significant and service has absorbed around 150,000 new recruits.
rapid expansion in staff numbers, at rates higher While spread across different functional areas and
than population growth in particular. This follows levels of government, this is still a large number
a decade of relatively more gradual expansion of staff to absorb into any civil service.

Figure 2.1: Population growth and civil service expansion

100,000,000 1,400,000
90,000,000
1,200,000
80,000,000
70,000,000 1,000,000
60,000,000 800,000
50,000,000
40,000,000 600,000
30,000,000 400,000
20,000,000
200,000
10,000,000
0 0
1991/92

1993/94

1995/96

1997/98

1999/00

2001/02

2003/04

2005/06

2007/08

2009/10

2011/12

2013/14

Population Civil service

Source Government of Ethiopia and WDI/World Bank (population).

Note The red line indicates population growth, with a corresponding axis on the left; the blue line indicates the growth
in civil service staffing, with a corresponding axis on the right.

51. While expanding the civil service at a rapid focusing available fiscal (and lending) resources
pace, the government has sought to contain on capital expenditures. Despite a period of rapid
the overall wage bill. The overall development expansion in the number of civil servants through
model being pursued has strongly emphasized 2012/13, the wage bill declined but has again
strengthening infrastructure. Doing so has required increased somewhat in recent years.
 10

Figure 2.2: Wage bill to GDP ratio

6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
1991/92
1992/93
1993/94
1994/95
1995/96
1996/97
1997/98
1998/99
1999/00
2000/01
2001/02
2002/03
2003/04
2004/05
2005/06
2006/07
2007/08
2008/09
2009/10
2010/11
2011/12
2012/13
2013/14
2014/15
Source World Bank staff calculations based on the Government of Ethiopia’s fiscal data (data not available for the
years 2005 to 2008).

52. The share of the wage bill relative to GDP is at percent of GDP, and most recently just above. This
the low end compared to other countries, while is consistent with the government’s intention to
relative to total expenditures it is on the higher use scarce fiscal resources as much as possible to
end due to the low overall levels of spending. fund investments that can help to propel growth.
Most countries for which data is available have Given Ethiopia’s still rather low revenue collection
wage bills exceeding 5 percent of GDP (Figure and overall low ratio of total public expenditures
2.3 – grey bars), reaching 10 percent of GDP in the to GDP, the wage bill takes up 25 percent of total
African countries for which comparative data is public expenditures, which is in line with cross-
available. For Ethiopia, the share has been below 5 regional averages (Figure 2.3 – blue bars).

Figure 2.3: Comparative data on wage bills relative to GDP,

and general government revenue and expenditures

South Asia (n=4)

OECD (n=30)
Middle East & North Africa (n=10)
Latin America & Caribbean (n=13)
Eastern Europe & Central Asia (n=22)
East Asia & Pacific (n=10)
Sub-Saharan Africa (n=6)

0% 5% 10% 15% 20% 25% 30% 35% 40%

2000-2013 Average

Wage Bill % GDP Wage Bill % Revenue Wage Bill % Expenditure

Source World Bank Database on public sector wage bills (2015).

11

53. A further important feature is the fact that 150,000 new recruits, which is a large number.
the rapid expansion of the civil service has been Furthermore, consistent with the fact that the civil
associated with a very large volume of recruitment service has seen a recent strong expansion, the
in recent years. As reflected in Figure 2.4, in age profile is that of a predominantly young civil
recent years the civil service has absorbed around service, especially at subnational levels.

Figure 2.4: Annual expansion of the civil service in Ethiopia 1992-2014

Number of additional civil servants per year

180,000

160,000

140,000

120,000

100,000

80,000

60,000

40,000

20,000

0
1992/93
1993/94
1994/95
1995/96
1996/97
1997/98
1998/99
1999/00
2000/01
2001/02
2002/03
2003/04
2004/05
2005/06
2006/07
2007/08
2008/09
2009/10
2010/11
2011/12
2012/13
2013/14
Source Government of Ethiopia, data on the civil service.

54. At sub-national levels, 61 percent of the civil 55. As has been explored in a recent study on
service are 32 years old or younger, with the urban labor markets, public sector employees
largest age group being those in their mid-20s tend to be educated at the secondary or post-
(23 to 27 year olds account for 29 percent of civil secondary level to a much higher degree than
servants at sub-national levels). At the federal those working in the private sector. Among
level, 38 percent of civil servants are below the urban workers surveyed, 87 percent of those
age of 32. As in many other countries, the single working in the public sector had secondary or
largest cohort of public servants are teachers and post-secondary education, compared to only 39
health professionals. Across government levels, percent of those working in the private sector
federal civil servants account for about 125,000, (including both formal and informal private sector
and sub-national civil servants for the remainder employment). The educational qualifications of
of 1,200,000. Just over one third of public sector public sector employees have increased over the
employees are female. past decade.
 12

Figure 2.5: Public versus private employees by education levels

100%
12%
80%
27%
62%
60%

40% 42%

20% 25%

18% 11%
0% 2%

Private Public

No Education Primary Secondary Post-secondary

Source World Bank staff calculations using UEUS 2014 (WB Urban Labor Market study)

2.1.1 Size and staff composition of the five being regions with large staff complements, and
sectors covered the others being much smaller in terms of staff.
Public sector reforms are spearheaded at the
56. The approximate staff sizes for each of the five federal level by dedicated reform directorates.
sectors covered in this analysis are the following:
58. As noted in Section 1.2., the survey undertaken
(i) Education – with over 420,000 staff, it is the for this report has mainly focused on civil
largest sector in terms of staff size service staff in policy making and supervision
(ii) Health – 130,000 staff functions, but also includes the perspectives
(iii) Agriculture – 80,000 staff of staff from just two sectors who interface
(iv) Trade – 19,000 staff directly with clients. The survey covers ERCA
(v) Revenue – 15,000 staff11 and trade staff at regional and woreda bureaus
who, in their day-to-day activities, have direct
57. Each sector comprises of a federal-level contact with clients. However, the civil servants
ministry or headquarters, as well as regional interviewed from the agriculture, education, and
bureaus and woreda-level offices. At the federal health sectors did not directly engage in frontline
level, organizational size ranges from 441 staff activities such as extension services, teaching,
at the Federal Ministry of Trade to 1,401 at the or health care. Instead, staff from these sectors
Federal Ministry of Agriculture.12 As Ethiopia’s played a role in the transmission of policies and
regions are highly unequal in size, sub-national the monitoring, supervision, and support of front-
staffing numbers vary substantially from region to line implementation.13
region – with Oromia, SNNPR, Amhara and Tigray

11 In addition to staff directly mapped to the federal level ministries and to regional bureaus there are additional staff in
sector specific colleges and research institutes (e.g. for agriculture).
12 As of 2014, the most recent year for which detailed statistics were available.
13 For cost reasons, front line staff in distributed service delivery units could not be covered.
13

59. The survey findings are complemented by sector issues and challenges related to staff capacity,
qualitative analysis. This has included the review management, and performance, as well as focus
of background materials for each of the five sectors group discussions with ‘Public Wing’ representatives.
covered, discussions with key informants about The findings, sector-by-sector, are set out in Annex 1.

2.2 Wage and non-wage benefits14

60. One key implication of rising staff numbers the upper single digits, and are projected to stay at
combined with the government’s efforts to keep around 8 percent annually over the coming years.
the overall wage bill in check is that remuneration
levels have remained low. The minimum salary 62. The government has been cautious about
has been gradually increased from ETB 300 to wage increases, out of a concern to (i) contain
ETB 615 in recent years, and the maximum salary recurrent expenditures, (ii) avoid fueling inflation,
from ETB 3,000 to ETB 5,780. According to and (iii) contain private sector wage pressures.
exchange rates as of March 2016, salaries range Recurrent expenditures were reduced in recent
from USD 30 to around USD 300 per month for years from close to 15 percent of GDP to 7.5
civil servants. This presents challenges for civil percent of GDP by 2013/14.15 Most recently,
servants, particularly in urban areas where the recurrent expenditures re-increased somewhat to
cost of living has accelerated more rapidly than 10.8 percent of GDP for 2015/16. The main thrust
salaries have been adjusted. Salary levels are has been towards capital expenditures to finance
periodically reviewed by the MPSHRD, and any infrastructure expansion. To address the fear that
changes have to be decided upon by the Council wage increases for civil servants can further fuel
of Ministers. inflation, the government has explored options for
non-wage benefits such as a transport system for
61. According to the Civil Service Proclamation, civil servants in Addis Ababa (see below).
civil servants are eligible for wage increments
every two years. However, the government 63. Instead of regular wage increments, the
stopped providing such increments in 2003. Since government has provided wage adjustments
then, the practice has been to announce larger every couple of years, but these have generally
scale wage increases on an occasional basis, only partially compensated for purchasing power
sometimes for all civil servants, and sometimes for losses due to inflation. Earlier wage increases
specific employment categories, such as teachers, were approved in 2002 and in 2007, followed by
who received a specific wage increase in 2016. an increase in 2011 (30 percent), in 2014 and again
Relatively high inflation means that without wage in 2017. However, these increases only (partially)
increases, civil servants experience a continuous compensate for losses in purchasing power due to
loss in purchasing power. This was particularly inflation, rather than representing a real increase in
pronounced in 2012 when inflation ran at over 30 wage levels. By regional comparison, professional-
percent. Inflation rates have declined since then to level civil service salaries in Ethiopia still remain low.

14
This discussion refers only to staff employed in the civil service, and not to staff employed in state owned enterprises, i.e.
the ‘wider public sector’.
15
The concern for private sector wages is to avoid increases ahead of productivity improvements. However, productivity may
be affected by a variety of policies such as barriers to entry and exit.
 14

64. Comparisons with average private sector 66. In the survey, a substantial share of staff
wages suggest that Ethiopia, as with many other responded that salaries are better in the private
countries, has a public sector wage premium, but sector than in the public. For example, for the
this has narrowed or even reversed over time.16 federal level, around 60 percent of respondents in
While public sector staff earn more than private the agriculture and education sectors thought that
sector employees on average, the difference between wages were an advantage when working in the private
the salaries of government and full time private sector. Thus, while there may be an average premium
sector employees is limited and has been shrinking. of public over private sector wages, this may not be
Moreover, when controlling for job characteristics, true across wage scales and specific professional
private sector pay has surpassed that of the public profiles. For some, income opportunities may be
sector in recent years. It is also important to keep in substantially higher outside the public sector – while
mind that both public and private sector wages in job security is perceived to be lower in the private
Ethiopia are among the lowest globally. sector (see section 5).21

65. Worldwide data assembled by the World 67. From the perspective of MDAs and of individual
Bank’s Bureaucracy Lab indicates that while there civil servants, low wages are a key issue. In nearly
is a wage premium, both public and private sector all interviews with key informants who are heads
wages are low in Ethiopia.17 The data suggests an or directors in MDAs, the issue emerges that low
average public sector wage premium that is sizable wages hinder attracting and retaining capable and
in relative terms as measured (65 percent). At the motivated staff. Technically and managerially more
same time, private sector wages in Ethiopia are highly skilled staff appear to remain rare, as hiring
the fourth lowest of all countries for which data constraints of such staff are also reported in private
is available;18 while public sector wages are the sector surveys.22 In the 2016 ECSS, 75 percent of
fifth lowest globally, and the third lowest in SSA.19 staff responded that they are dissatisfied or very
Furthermore, it is noticeable that Rwanda, which dissatisfied with their salaries, with somewhat greater
pursues a development model similar to that of satisfaction at the regional level (about 30 percent
Ethiopia and which has similar private sector wage satisfied or very satisfied). Poor pay and benefits are
levels, has substantially higher public sector wages also raised as the biggest driver of staff turnover in
(and accordingly a particularly large relative wage the civil service survey across levels and sectors,
premium in favor of the public sector). Differences with the exception for ERCA employees at regional
between data sources and information about public levels;23 and it was by far the most frequent issue
sector wage adjustments suggests that the public raised in response to an open ended question about
sector wage premium in Ethiopia fluctuates over why staff might feel less motivated currently than
time, as private and public wages adjust to inflation when entering the public service (raised by just over
differently over time.20 20 percent of respondents).

16 Based on a 2016 World Bank study on urban labor markets.

17 The database assembles private and public sector wages as of 2011 in PPP terms for about 90 countries.
18 Slightly lower private sector wages are recorded for Rwanda, Liberia and Botswana.
19 Lower public sector wages are recorded for DR Congo, Georgia, Liberia, Tajikistan.
20 Also, different wages paid by institution/ministry make comparisons between public and private sector wages more complex;
see Annex 1 for a discussion of sector-by-sector issues.
21 In addition to the regular private sector there are at least two other relevant types of employment, in particular for federal
level staff: (i) employment with SOEs, and (ii) for some staff categories, employment with regional organizations.
22 World Bank Economic Update IV.
23 ERCA staff has received a special wage regime to address challenges of spreading corruption and high turnover in this function.
15

68. While wages and benefits stand out as a servants at the federal level in Addis Ababa, and
clear constraint, operations and maintenance (iii) some preferential treatment for the ‘housing
expenditures are also very low, and increasing lottery’ organized for Addis Ababa.
these selectively may be even more important.
As is discussed further in Chapter 4, staff 2.2.1 Pensions24
identify a lack of resources as an important
impediment on effectiveness, as well as a source 70. The usual retirement age for public servants
of dissatisfaction. This concern expressed by civil and private sector employees is 60 years. The
servants is consistent with findings from the 2016 minimum service requirement for an employee to
Public Expenditure Review (PER) that raised the be eligible for pension is 10 years and the pension
low level of Operations and Maintenance (O&M) will be effective when s/he attains retirement
expenditures as a key constraint. As discussed as age. Public sector pensions for civil servants are
part of the recommendations, the GoE will need managed by the Public Servants Social Security
to consider carefully where to allocate additional Agency (PSSSA).25 The pension is a defined benefit
fiscal resources, given that there are multiple plan. It is only partially funded from contributions,
competing demands. Focused O&M expenditures and given that it has been in existence for several
are potentially very important as they feed both decades, it is facing increasing financial pressures.
into the quality and timeliness of service delivery, The conditions of this public sector pension
as well as the motivation of public service staff. scheme are in line with those common in the
region in terms of contribution rates, retirement
69. Civil servants in Ethiopia enjoy some non- age, and so on. A private sector pension scheme
wage benefits, but these are rather small in was introduced much later (only in 2011), but
value and limited in nature compared to other has essentially the same conditions as the public
countries. The main benefits are (i) a public sector sector pension scheme. Thus, from an employee’s
pension scheme that has been in place since perspective, there is no greater benefit associated
1963, (ii) a transportation system created for civil with being in the public sector scheme.

Table 2.1: Public and private sector pension plans in Ethiopia

Pension Employer Employer Replacement Survivor Normal

fund rates after pension retiement age
30/40 years

Civil service 11 7 55/68 Widow/er 50%; 60 (55 with 25 years

orphan 20% each of service)

Private Employees 11 7 55/68 Widow/er 50%; orphan 60 (55 with 25 years

20% each of service)

Military and Police 25 7 63/80 Spouse 50%; orphan Determined by military

20% each and police decrees

Source World Bank (2017).

24
This section draws on the report by Palacio on the pension system in Ethiopia.
25
A separate pension scheme is in place for the military and the police.
 16

71. Despite a large (80 percent) increase in the completely portable between the public and private
minimum pension in 2014, the level of pensions sector schemes, it is possible that some senior
being paid remains rather low. Based on the government employees with long contribution
reported spending and estimated number of histories may leave the public service and join
pensioners, the average pension would be ETB a private firm at a much higher salary. Pensions
503 or roughly US$20 per month – less than the are calculated based on final wages so that the
absolute poverty line of $1.25 per day.26 calculation could yield a much higher pension.

72. The low pensions reflect the historically low 2.2.2 Health insurance
wages that have existed (most pensions are
based on the wages of employees who had retired 74. Plans have been prepared to introduce
during the last few decades), as well as the lack social health insurance since 2010 (Proclamation
of indexation against inflation – so that pensions 690/2010). The government sought to introduce
lose significant purchasing power over time.27 this first on a mandatory basis for civil servants
While the minimum pension is equivalent to 86.4 with a contribution rate of 3 percent of salaries,
percent of the minimum wage, the ratio of the combined with a government subsidy. However,
average pension to the average wage appears to civil servants protested that their wages are already
be much lower. While data are not available, this so low that they cannot make such an insurance
implies a very flat distribution of pensions, with contribution.29 The actual introduction of the
most of those who retired more than a few years scheme has been postponed. Civil servants can
ago receiving the minimum. This is largely due to the access free public health care as other citizens.
lack of automatic indexation, which results in older While such free health care remains of relatively
pensioners receiving the minimum.28 Aside from low quality, civil servants appear concerned that
this effect, the lack of the indexation of pensions making an insurance contribution will not give them
means that the relationship between contributions access to better quality care. As for the pension
and benefits is very weak. A public sector employee system, the social health insurance scheme is also
who has contributed twice as much as another open to formal private sector employees, but again
employee (because s/he had a higher gross salary) the incentives to use the scheme are limited.
will eventually receive the same minimum pension
as the second employee. It also means that many 2.2.3 Allowances
civil servants could have done better saving on their
own rather than contributing to the pension scheme. 75. The main allowances provided are those for
domestic travel, but for most staff, these are
73. Average pensions paid out from the still new below actual costs in many instances. The rates
private sector scheme are significantly larger than have been traditionally (too) low, i.e. insufficient to
those paid from the public scheme due to the lack meet the actual costs of accommodation and meals
of indexation and higher wages being paid in the (NORAD 2012). Per diem rates are differentiated by
(formal) private sector. Since pension accruals are salary levels, and are particularly inadequate for

26
The average is lower than the minimum pension which is probably due to a relatively high proportion of survivors’ pensions
which are equivalent to 50 percent of the original pension for spouses and 20 percent for orphans.
27
See for example, http://ethiopiaobservatory.com/2014/06/08/pensioning-in-ethiopia-no-country-for-old-men/.
28
In principle, pensions are adjusted every five years. This increment is the fifth since the EPRDF government came into power.
In 1992/93, an amendment was made to the lower salary of the civil servants, from 50 Br to 150 Br. There were also further
increments in 2000/01, 2006/07 and 2010/11. http://allafrica.com/stories/201408060369.html.
29
See e.g.: Getting Better: Is the New Health Reform the Cure? Ethiopian Business Review 2014, no. 12.
17

lower-level staff. As a consequence of low per diems,

management routinely allows staff to overstate the 78. The government began registering interested
number of days needed for a trip. Staff have become parties for the housing lottery in 2005, setting aside a
more reluctant to undertake any travel since per share of the housing for civil servants. Actual lotteries
diem rates are so inadequate. Furthermore, donor are held at irregular intervals whenever a batch of
and government per diems diverge substantially, housing nears completion. In 2015, the lottery was
and the per diem rates also vary from project to held in March and allocated 35,000 housing units
project (under ‘Channel 1’, i.e. project support). Sitting (from studios to 3-bedroom condos). In the general
allowances are not available to most regular civil lottery, 20 percent (i.e. about 7,000 condos at current
servants. Neither the government nor development rates of ‘production’) are set aside for civil servants,
partners pay sitting allowances for work or project- relative to the total number of federal civil servants
related meetings or reviews. (100,000, plus 65,000 civil servants working for the
Addis Ababa City Government).30 So the chances of
2.2.4 Transportation winning in the lottery are still quite limited for a civil
servant. When being allocated a condo through the
76. A transport benefit was phased in during 2014, lottery, winners must make a 20 percent contribution
based on a plan developed by the MPSHRD, for towards the price of the condo.
civil servants in Addis Ababa. It is managed through
a Public Service Employees Transport Service 79. Benefits and opportunities are significantly
Enterprise (PSETSE), which was created as a public better for senior government officials (i.e. political
enterprise for this purpose. Buses travel along 26 appointees). The benefit package for Ministers
routes all the way to outlying/suburban areas of and State Ministers includes government provided
Addis Ababa. The light blue colored buses became housing, as well as land to construct housing on.
widely visible throughout Addis Ababa in 2015-2016. In addition, it is common for ministers and state
To provide some funding for this scheme, buses ministers to be members of the Board for various
operate on a fee basis during the day for the general public enterprises, and significant allowances are
public. paid for this function.

2.2.5 Preferential participation in the 80. There are various plans and considerations to
housing lottery expand the benefit packages for civil servants; e.g.
to provide teachers moving to remote areas with a
77. Housing has become a significant challenge in housing option or with land on which housing can
Addis Ababa and other large cities in Ethiopia. In be constructed. Generally, the government considers
principle, housing constructed with government the provision of additional benefits as less prone to
funds are provided on a lottery basis, i.e. it is not stimulating inflation than increasing the monetary
an entitlement or a waiting list with a (more or less) wage of civil servants.
predictable date for receiving housing.

30 Most, but not all federal civil servants reside in Addis Ababa. Some may be
posted to border offices of ERCA etc.
 18

81. Overall, regular civil servants receive low salaries, wages or pensions. Despite these limited benefits, the
as well as rather limited benefits in Ethiopia. In a civil service by and large has still been able to attract
context of at least 8-10 percent consumer price staff, as the large expansion in recent years indicates,
inflation per year, they also face year-to-year although there are resulting challenges to retaining
uncertainty with regards to the purchasing power of and motivating staff as discussed further below.
19

3
Civil Service
Reform Efforts
 20

82. This chapter uses the Ethiopia Civil chapter focuses on the following reform tools
Servants Survey (ECSS) 2016, in combination and covers them individually: The Business
with other public service delivery data and Process Reengineering (BPR); the Balanced
information from interviews, to analyze civil Scorecard (BSC); the Change Army; and the
service reforms in Ethiopia. It focuses mainly Citizens Charter. It attempts to understand
on past and current reform tools initiated by the perceptions that civil servants have of the
the Government of Ethiopia to implement the reform tools, their achievements, how these
Civil Service Reform Program. In particular, the relate to each other and how successfully they
21

were implemented. It also includes a review practices and provides global and regional
of the impacts of the reforms on management perspectives on the reforms.

3.1 Brief overview of global and regional experiences with public service reforms
to situate Ethiopia’s reform approaches

83. Achieving progress with civil service reforms in Ethiopia has deployed a range of approaches
low income countries has posed major challenges that broadly fall into the category of ‘New Public
and is perceived as having yielded only mixed results Management’.31 The main successive reform tools
over the past two to three decades (Andrews 2013; include Business Process Re-engineering (BPR),
Therkildsen 2006). Low government effectiveness Balanced Scorecard (BSC), Citizens Charters,
and corruption continue to be widely diagnosed as well as the introduction of a Change Army.
in many countries. However, there are also partial While both BPR and BSC are classic ‘New Public
indicators of progress (World Bank 2008; DFID, Management’ style reforms, and also widely
Irish Aid, and Swedish International Development found in other countries seeking to strengthen
Agency 2013; Blum 2014). On average, government civil service performance, the Change Army is a
effectiveness in low income countries has slightly more Ethiopia-specific approach. These reform
declined rather than improved in recent years (see initiatives were usually centrally planned within
Figure 3.6). However, it has to be kept in mind the party and rolled-out in accordance to a
that a number of countries moved from low- to predetermined plan. They tend to lack strategic
lower-middle-income status, and hence there has communication tools that foster key stakeholder
been a slight trend towards improvements among collaboration and network building mechanisms
those countries. so that civil servants, the private sector, and civil
society can support the reform.
84. Against this background, Ethiopia stands out
as a country that has undertaken extensive civil 85. The reform tools used and their impacts as
service reforms during the last 20 years and with perceived by civil servants as well as key informant
some documented positive impacts. In particular, interviewees are discussed in detail below.

31 One of the frequently cited concerns in the literature is that NPM-inspired reforms were not successful or even had negative
consequences in low income countries because (i) they were undertaken in response to conditionality, and (ii) they induced
‘isomorphic mimicry’ (i.e. imitating a standard) rather than real reforms (Andrews 2013). However, while reforms in Ethiopia
have been externally supported, the selection and pursuit were very much internally decided.
 22

3.2 Initial reform period in Ethiopia

86. The structural adjustment program carried delivered services to the public. It also identified
out by the Transitional Government of Ethiopia the lack of a proper arrangement to ensure ethical
(TGE) included a review of the civil service. The behavior and tackle corruption, embezzlement, and
review identified and implemented retrenchment, fraud in government.
and supported the design of appropriate structures
to eventually devolve functions and personnel to 88. The task force findings and recommendations
regional administrations. With the enactment of the were reported to the Ethiopian government in
new Constitution in 1994, the federal government January 1996. The Civil Service Reform Program
initiated regionalization by devolving powers from (CSRP) was launched with the objective of creating
the federal government to the ethnically organized a fair, ethical, effective, efficient, transparent and
regional state governments. The civil service review accountable civil service. This was to be achieved
that took place in the early days of the TGE removed through strengthening institutional and human
the national civil service elite that had remained resource capacity; developing and installing new
from the past regime,32 which meant that most of and improved legislations and working systems;
the systems were either abandoned or couldn’t be introducing a culture of good governance; and
made functional. Thus, even as regionalization was enhancing performance in public service delivery.
being carried out, it was clear that the government’s It comprised of five major program areas: a) Top
capacity to collect revenue and deliver development Management System Reform Sub-Program; b)
initiatives across the country varied widely, and that Human Resources Management System Reform
all regions experienced a huge staffing problem.33 Sub-Program; c) Service Delivery Reform Sub-
The government’s strategy for addressing these Program; d) Expenditure Management and Control
issues and the rapid growth of civil service (as System Reform Sub-Program; and (e) Ethics Reform
discussed in Chapter 2) centered mainly on civil Sub-Program.35 The government started to enact
service reform. legislations, directives, and guidelines to improve
the internal processes of the civil service.
87. The Ethiopian government’s first major
attempt to reform the civil service started in 89. The internal dissent within the party and
November 1994 with the creation of a task subsequent first renewal process of the ruling
force in the Prime Minister’s Office (PMO).34 The party in 2001 was a key milestone in setting
task force undertook an in-depth review and the country’s overall direction. An important
identified a number of weaknesses as to how top ideological decision was made in 2001 in favor of
management set strategic priorities and monitored the ‘developmental state’. Adopting such a model
and evaluated performance in the implementation further emphasized the need for a professional and
of development policies, programs, and projects, reliable civil service, consistent with the experience
as well as the manner in which the civil service of earlier developmental states in East Asia (Evans
managed its human and financial resources and 1995, Vu 2007, World Bank 201336). The wider

32 July 2010, Assessment of the Implementation of the Civil Service Reform Program in Ethiopia, Ministry of Capacity Building.
33 May 27, 1993; IDA Credit to Support Structural Adjustment in the Ethiopian Economy, World Bank.
34 Reform efforts in the 1990s are elaborated in the annex.
35 May 2004, “Akem” “Capacity”, A Quarterly Bilingual Magazine Volume 1 No. 1, Ministry of Capacity Building.
36 International Bank for Reconstruction and Development and World Bank, 2013. Federal Democratic Republic of Ethiopia:
Ethiopia Public Sector Reform Approach Building the Developmental State – A Review and Assessment of the Ethiopian
Approach to Public Sector Reform. AFTP2, Report No: ACS3695. World Bank: Washington DC.
23

and more critical the role adopted by the state in the organizations’ mission, vision, and strategies.
pursuing social and economic transformation, the
more the state apparatus needs to be capable and 91. The detailed elements of the CSRP Action Plan
act with integrity, so as to avoid becoming a source include the following:
of distortions, unproductive rent-seeking, and
‘government failure’. With this in the background, • Staff awareness of CSRP
the government, using a “big push” approach, • Service delivery policy
moved to the second phase of decentralization by • Formulating a mission statement
devolving service delivery to woredas/districts. The • Promoting awareness among front-line staff
need to expand the number of those in civil service • Publicizing services to identified users
and build their capacity, as well as to develop • Facilitating access for users
institutions at a sub-national level, became an issue • Coordinating related services
of immense importance in discussions of Ethiopia’s • Establishing a complaints procedure
social and economic transformation during this • Consulting with users
time. A new Super Ministry of Capacity Building was • Setting service standards
established to coordinate all civil service reforms, • Developing cost effective services
including the CSRP. • Results-oriented appraisal system
• Ethics and anti-corruption actions
90. Although the CSRP’s five components and • Organizational assessment
vision have remained unchanged since its • Self-assessment of strengths and weaknesses
inception, the reform tools used to achieve • Stakeholder analysis (including a client survey)
its objectives and the emphasis given to each • Analysis of the gap between actual and
component have seen successive evolution. In the expected services
early years, following the initiation of the CSRP in • Environmental analysis
1996, efforts mainly focused on addressing internal • Service Improvement Action Plan
administrative issues (legislations, FM, Ethics) with
very limited impact on performance and service 92. The implementation of the CSRP Action
delivery to citizens. Starting from September 2002, Plan, through the above-indicated reform tools,
the Ethiopian government launched more outward brought a number of new ideas into the civil
looking initiatives that initially followed a Civil service. The Quick Wins approach brought about
Service Action Plan comprised of the following greater customer orientation in government offices
elements: i) the development of a Service Delivery through the introduction of name tags, information
Policy, in which service providers are encouraged counters, and suggestion boxes; the appropriate
to be more customer focused through the labeling of offices; and the posting of the office’s
‘Quick Wins’ initiative (the Quick Wins approach mission, vision, and values in a visible manner. The
incorporates short-term actions needed to deliver approach also facilitated the initial efforts to reduce
improved business processes and addresses work steps in providing services. The introduction of
identified organizational weaknesses with relatively the employee Result Oriented Performance Appraisal
few resources); ii) the implementation of a Result- System (ROPAS) introduced formal performance
Oriented Performance Appraisal, whereby supervisors targets to be agreed upon by staff and supervisors,
and staff are expected to agree on individual work while the strategic planning process established
plans and monitor their performance against higher-level objectives within public organizations,
these; and iii) the introduction of Strategic Planning albeit with a weak link between the two (i.e.,
and Management, according to which all public individual and organizational objectives). While a
organizations are required to develop a strategic number of these early improvements have been
plan that mainly encompasses the development of sustained (e.g., clear labelling of offices, wearing
 24

of name tags), others that did not initially take to make a quick impact, in 2005/06, the federal
root were further pursued through subsequent government deepened and accelerated CSRP
reform efforts, notably with regards to performance throughout the civil service with the introduction
management and complaints mechanisms. Anxious of the BPR tool.

Figure 3.1: Timeline of reform tools

Woreda Level
Decentralization

Change Army
Regional Strategic Planning Business Process Balanced
Level Ongoing efforts
& Management Re-engineering Scorecard
Decentralization (SPM) (BPR) (BSC)
Citizens Charter

Result-Oriented
Performance
Appriasal System
(ROPAS)

Service Delivery
Policy-Quick-
Wins

1994 2001 2006 2008 2011 2016

61. Figure 3.1 illustrates a timeline of the various civil service reform efforts undertaken by the Ethiopian
government.

3.3 Reform processes since 2005/2006

3.3.1 Business process reengineering (BPR) BPR is part of the wave of New Public Management
inspired reforms (for a comparative overview,
93. Given the Ethiopian civil service’s starting see Pollitt and Bouckaert 2000/2011). Typically,
point as an overly hierarchical and overly input BPR is seen as being at the more ‘radical’ end of
based institution, the federal government made NPM initiatives, pursued by countries especially
the sweeping decision to initiate the BPR across all interested in importing private sector practices into
government agencies, to achieve significant change the civil service and in reducing the distinctions
and establish the rapidly expanding civil service between the two, such as New Zealand, Australia,
on a new footing. From a comparative perspective, and the UK (Pollit and Bouckaert 2011, 116). Apart
25

from these core NPM countries, Ethiopia has been 94. The BPR specifically is based on the
one of the most prominent adopters of BPR as ideas of Michael Hammer, a US management
a key approach for civil service reforms. Given theorist coming from the field of computer
the diversity of applications globally, there is no science (see Assefa 2009). 37 It was intended
comprehensive assessment of how successful BPR to radically and fundamentally transform
reforms have fared in public or in private sectors. the business diamond of civil service
In one assessment, Jurisch et al. (2012) note that organizations, which includes process, jobs
60 to 80 percent of BPR efforts fail to achieve their and structure, management and measurement
goals, but also note that BPR can be a valuable and systems, and values and beliefs.
much-needed tool for public administrations.

Figure 3.2: The core functions of the BPR

PROCESS

JOBS & STRUCTURE ORGANIZATION MANAGEMENT & MEASUREMENT

VALUES & BELIEFS

37 Hammer was a professor of Computer Science at MIT who in 1990 published an influential article ‘Reengineering Work:
don’t automate, obliterate’ in the Harvard Business Review; which argued that efficient processes were key to a company’s
success in the 21st century.
 26

95. The BPR initiated by the Ethiopian federal and reduced processing times; 60-70 percent of
government identified four stages for the organization heads have also reported that the BPR
change process– i) planning; ii) understanding; increased the use of IT in organizations. Although
iii) redesigning; and iv) implementation. The organization heads agree to some extent that BPR
planning stage was used to create a shared vision improved workplace attitudes and teamwork, the
and system of beliefs, values, and attitudes within results in relation to the “soft” elements like shared
an organization to initiate the change process. Staff values, staffing/skills, incentives, and flexibility seem
were exhorted to believe that the true job of a public to be less tangible, as discussed in the management
servant was to provide service, and to demonstrate practices section (Section 3.6).
this new and improved attitude to customers. Staff
confessed to or accounted for less than acceptable 98. Around half of the organization heads
performance. This was followed by an effort to surveyed suggested that BPR implementation
understand the causes, effects, and magnitude of was constrained due to rushed planning; lack of
the problems associated with old business processes interlinkage with other reform tools; and weak
and systems using inputs from both internal and monitoring. Some 35 percent of organization
external stakeholders. During the redesigning stage, heads disagree to some extent that reforms
staff and management developed new processes are customized to the specific nature of the
and systems (manuals, new and improved legal organizations, which was also confirmed by the
frameworks, process charters, etc.) and structures qualitative interview. Organizations that have direct
(jobs, skills, competency framework, etc.), based interface with citizens seemed to benefit more from
on best practices, customers’ expectations, and the BPR. Survey results indicate that BPR reforms
the principles of process-based organizations. had a significant impact for the trade sector, which is
The implementation stage basically executed the consistent with its particular role of issuing licenses
framework agreed upon during the earlier stages. and similar tasks. Especially federal- and regional-
level staff agree that processes were dissolved or
96. Given the starting point in Ethiopia, where changed, and that processing times were reduced.
the civil service was known as overly hierarchical
and input based, aspiring to implement BPR was a 99. Similarly, a merger of the three entities
sweeping decision, seeking to achieve significant responsible for revenue collection as a result of
change and to establish the rapidly expanding BPR helped to reap efficiencies and streamline
public sector on a new footing. Overall, top leaders processes, including with regards to trade
across tiers of government and sectors have positive facilitation, i.e. by making the settling of customs
feedback on the implementation and the effects and tax obligations and refunds easier. ERCA
reform tools have had on civil service organizations sustained the structure suggested by the BPR and
across the board with only few concerns. Around there is a continuing need for ongoing Business
80 percent of the interviewed heads of ministries, Process improvements. On the other hand, BPR
bureaus and offices agree to some extent that the reform is currently seen as not having been a
reform efforts by the government focus on the right ‘good fit’ for the education sector, and having
things and real problems. resulted in some capacity losses that had to be
subsequently recouped. Similarly, the significant
97. Indeed, the BPR reform tool changed the “hard” cut in staff numbers was associated with a loss
elements of organizations such as the strategy, of capacity at the Ministry of Agriculture, as well
structures and system. Some 80-90 percent of as the creation of some wider resentment and
surveyed organization heads agree to some extent concern about civil service reforms, and the BPR
that the BPR transformed the organizational in particular. In health, the application of BPR in the
structure, improved service-delivery standards, sector mainly focused on analyzing institutional
27

structures and human resources in health. Survey performance using the Balanced Scorecard (BSC)
results indicate that compared to other sectors, tool, which aims to align individual and organizational
the view that BPR reforms achieved its objective performance. Performance accountability and
of reduced processing times is less common in the transparency are also being sought through the
health sector and the structures before the reform introduction of the Change Army, under which the
have now been reinstated in most of the cases. government seeks to unite the party, state, and society
to achieve the GTP, and Citizens Charters, a document
3.3.2 After the BPR - current reforms that represents a systematic effort to focus on the
commitment of public service providers towards their
69. Subsequently, efforts evolved towards a greater users with respect to standards of service, information,
emphasis on the measurement of public service and grievance redress mechanisms.

Balanced Scorecard (BSC)

100. Building on earlier efforts to establish performance is now being addressed through
strategic planning norms and an employee a system of individual performance agreements
results-oriented performance appraisal system, that are linked to team and organizational targets
the federal government introduced the BSC in derived from the four dimensions of the BSC.
2008. The BSC is being used as a planning/target These include: i) finance, ii) learning and growth;
setting, change management, and communication iii) client/customer satisfaction, and iv) processes
tool, and is also considered as an opportunity to (time volume, frequency). The first set of annual
embed strategic management principles into the performance agreements for individuals were
strategic plan. The BSC incorporates the higher- signed in July 2012. Currently 60 percent of the
level goals identified by the country’s Growth score is allocated towards achieving the agreed
and Transformation Plan (GTP) and tries to upon specific tasks and goals, while 40 percent is
operationalize these goals through a cascading dedicated to attitude or behavioral aspects (this
process of setting targets and disaggregating them can also be seen as a reaction to earlier criticism
by specific units and roles. As part of this process, of assessments introduced by the BPR that were
public organizations incorporated BSC measures seen as overemphasizing ‘attitudes’). Thus, the BSC
within strategic sector plans. seeks to create direct links between employees’
performance and the strategy of an organization
101. In 2012, the civil service’s “Individual towards its clients/citizens, budget/stewardship,
Employee Performance Management and internal processes, and learning and growth. Both
Measurement” directive was incorporated employee and organization performance reviews
into the BSC. The management of individual are carried out twice a year, in December and June.

The BSC incorporates the higher goals

identified by the country’s GTP.
 28

Change Army and the establishment of Public Wings

102. The Change Army aims to mobilize communities meets monthly with all staff members to discuss
and organize their contributions in support of the performance reports and resolve internal issues,
country’s development goals (GTP). Following its and with the Public Wing on a quarterly basis.
pilot in the rural areas, the Change Army has, since
2012, been streamlined into the civil service to bring 105. The Public Wing is a platform that constitutes
about national consensus by involving the party, of clients; associations for youth, women, and
state, and citizens in its three wings, namely the Party professionals; cooperatives; private sector
Wing, State Wing, and Public Wing. representatives; and others identified as relevant
by government ministries, agencies, and bureaus.
103. The Party Wing is concerned with building a On a quarterly basis, Public Wing members meet
national consensus on the country’s development with the State Wing (typically represented by a
strategy, and the civil service is one of the key minister or state minister at the federal level38) to
players in this discourse. The Party Wing is discuss strategic plans and their implementation,
responsible for providing orientation programs the objectives set, the service standards agreed
to enable civil service members to fully grasp upon by the Citizens Charter, as well as new
and effectively execute government policies and policy recommendations and issues related to
strategies. However, this study did not cover the rent seeking.
activities of the Party Wing extensively.
106. Focus group discussions with Public Wing
104. The State Wing (civil service) consists of representatives in four out of the five sectors
five levels: individual civil servants, ‘model’ civil covered indicate that while the consultations are
servants, lower-level leadership, middle-level overall appreciated, they also still have a number
leadership, and top leadership. At the lowest level, of weaknesses. Public Wing members feel that
the ‘1 to 5’ members are grouped together based their potential is not always fully harnessed and
on their work processes and led by a ‘model’ civil their capacity to be an effective partner to the
servant (i.e., someone identified as an effective government is constrained by the Charities and
performer). These model civil servants conduct Societies Proclamation of 2009 (otherwise known
daily/weekly peer reviews and evaluate the as the CSO law), which restricts professional
group’s performance. This is expected to create associations in terms of financing. Contributions
a mechanism to coach junior staff and improve from association members are minimal, and they
attitudes by putting peer pressure on civil servants. are not allowed to accept more than 10 percent
The outcomes of these weekly discussions are of their budget from foreign organizations. As a
presented to the lower-level leadership that result, they have had to return money from non-
constitutes of the lead civil servants from the ‘1 member sources due to this restriction. When
to 5’ groups, and/or the middle-level group that capacitated, a potential role exists for professional
constitutes of directors. The highest level, the associations in re-licensing and continuous
Council, constitutes of top managers and directors professional development issues, as experiences
and meets at least monthly. The Council also show from other countries.

38
Reform directorates from the respective Ministries and agencies hold a secretariat role.
29

Citizens’ Charters

107. The Citizens Charter enables government service standards in consultation with clients.
institutions to enter agreements with the Many organizations started to publicize these
general public on the roles and responsibilities standards prominently through noticeboards
of the providers and service users, the manner at the entrance of public offices. However,
in which public services have to be delivered, there was no systematic monitoring of the
and the modalities in which complaints will be organization’s actual performance against
resolved. The idea of accountability to citizens these standards. This led to significant variance
is a relatively new concept in the Ethiopian civil downward from the standards set under BPR in
service and there was no established system 2007/08. The government is now trying to re-
through which this could be strengthened. institute the service delivery standards under
Despite the government’s efforts to establish BPR through a formalized Citizens Charter.
structures to enhance citizens’ engagement, the The Citizens Charter, where it is implemented,
space for participation as well as the willingness is published and disseminated to users using
of citizens to be actively involved and demand for different mechanisms, including notice boards
better services is limited, although one can say posted where services are provided.
that both have seen some encouraging trends.
109. An example of taking steps beyond Citizens’
108. The basic principles of the Citizens Charters and creating a ‘right to public services’
Charter were first embraced by organizations is summarized in Box 3.1. The Indian experience
as part of the BPR process. All civil service highlights an effort to further strengthen the
institutions were encouraged to establish rights of citizens vis-à-vis service providers.
 30

Box 3.1: The Indian experience with a ‘Right to Public Services Act’

Growing public dissatisfaction with the various states includes granting of the ‘right
delivery of essential public services has been a to public services’, which are to be provided
worrying phenomenon in India in the last two to the public by designated officials within a
decades. Large numbers of citizens often face stipulated time frame. The public services,
ad-hoc service rules, denial or delayed action, which are to be granted as a right under the
while applying to government agencies for legislation, are generally notified separately
essential services and documentation. Media through a gazette notification. Some of the
reports highlight widespread corruption, common public services, which are to be
harassment of citizens, and lack of public provided within a fixed time frame as a right
accountability in service design and delivery. under the Act, include issuing caste, birth,
marriage and domicile certificates, electric
In the last few years, the use of legislative connections, voter cards, ration cards,
guarantees to ensure citizens receive services copies of land records, and so forth. Second,
in a fair, transparent and time-bound manner a notable feature of the legislation is that
has gained tremendous policy momentum if the designated officer fails to provide the
and been formalized into strategies, laws and service within the stipulated time or rejects
operational guidelines. the service request, the aggrieved person may
approach the appellate authorities specified
Central and state governments in India have under the Act. These authorities can order
sought to plug the gap in “Citizen’s Charters”, the public servant to provide the service to
which could not be enforced due to lack of the applicant, and impose a penalty on him/
incentives and legal legitimacy, by passing her for failure to provide the required service
the Right to Public Services Acts. These without any reasonable cause. The penalty
legal norms, adopted since 2010, codify may be financial, in the range of Rs. 500 to
statutory laws that guarantee time-bound Rs. 5000, and/or disciplinary proceedings.
and standardized delivery of various public The applicant may be compensated out of
services rendered by the government to the penalty imposed on the officer. These
citizens. These acts also provide a mechanism appellate authorities have been granted
for punishing errant public servants who are certain powers of a civil court. Thus, there
proved to be defficient in providing services is hope that such a codified process of
as stipulated under the statute. grievance redress imposing disciplinary and
pecuniary action for defficient or delayed
The RTPS law aims to correct the power responses to citizens could herald a more
imbalance between administrators and accountable and citizen oriented culture
applicants through two key features. within public administration.
First, by making timeliness of services
prominent, the bill allows a standardized and For rolling out the RTPS legislation, a number
verifiable measure of an administrator’s job of factors such as management capacity
performance based on citizen experience. of the state machinery at different levels,
The common framework of the legislation in resources, and behavioral factors have played
31

crucial roles. The discretion of governments activities (Mathur 2012). It is important to

in deciding the services covered is also a improve citizen awareness regarding this
decisive factor (Raha 2012). Challenges in Act along with proper planning of resources
the implementation of the RTPS Act arise and infrastructure for ensuring compliance
at the time of carrying out preparatory and transparency.

Source Implementation of the Right to Public Services Act in India: A Survey Assessment, Public Affairs Centre, Bengaluru and
the World Bank, 2016.

3.4 Civil servants’ perceptions of current reforms

110. The following section discusses the perceptions whether they had an “approripriate attitude” at
of civil servants with regard to the implementation work. At the federal level, the BSC registers as
of the BSC, Citizens Charter, and Change Army, the most successful reform tool. Among sectors,
as measured by the Ethiopia Civil Servant Survey the Change Army is particularly utilized by health
(ECSS). Understanding the perceptions of the civil sector staff, and relatively less so by staff working
servants will allow policy makers to better refine the with the Ministry or Bureaus of Trade. However,
various tools to address gaps in implementation. Citizens Charters are seen as relatively less
It’s important to note that the results represents successful thus far. Citizens Charters are more
a single snapshot in time, i.e. how civil servants frequently raised as a successful reform by staff
currently view these reforms, some of which were in the trade and revenue organizations, which
initiated sometime in the past, while others are interface with citizens and businesses around
more recent. specific ‘business processes’, such as paying taxes
or issuing a license. Conversely, appreciation for
111. Successful reforms: Civil servants highlight the BSC is especially pronounced at the federal
the Change Army as a reform tool that they level (48 percent relative to 25 percent at regional
feel has been rather successful (55 percent levels and 28 percent at woreda levels), where the
nationally).39 This is most pronounced at regional BSC is likely to be more entrenched. A similar
(59 percent) and woreda (58 percent) levels, picture is observed across sectors. In the trade
as opposed to the federal level (36 percent). sector, appreciation for the Change Army (46
During qualitative discussions and consultation percent) is slightly less pronounced than in other
workshops, however, participants also noted sectors, while the BSC (36 percent) is slightly
that they percieved the Change Army to be more. Staff from the revenue agency cite the BSC
politicized, as the mechanism tended to evaluate to be the least successful reform tool.40
civil servants using subjective measures, such as

39
Respondents were asked the question “Is there any one reform that you think has been especially successful, useful, or
beneficial? Which?” to which 55 percent responded ‘Change Army’, 30 percent ‘BSC’, 8 percent ‘Citizens Charter’, and 8
percent ‘Other’.
40
Respondents were asked: “Is there any one reform that you think has been especially unsuccessful, or detrimental? Which?”
 32

Figure 3.3: Perceptions of successful reforms by tier

Is there any one reform that you think has been especially successful, useful, or beneficial?
All civil servants by tier

Woreda 28.42% 57.82% 7.82% 5.94%

4.49%

Region 24.79% 58.55% 12.18%

Federal 48.41% 35.69% 11.66% 4.2%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

BSC Change Army Citizens Charter Others

Figure 3.4: Perceptions of successful reforms by sector

Is there any one reform that you think has been especially successful, useful, or beneficial?
All civil servants by tier

Trade 35.71% 46.43% 11.19% 6.67%

Revenue 23.56% 56.35% 11.32% 8.78%

3.70%

Health 30.79% 59.26% 6.25%

Education 31.70% 54.69% 6.03% 7.59%

3.56%

Agriculture 29.56% 58.44% 8.44%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

BSC Change Army Citizens Charter Others

112. Less successful reforms: The Citizens Charter in agriculture and trade express dissatisfaction
is mentioned the most when respondents are asked with the Citizens Charter more frequently, with
to highlight less successful reforms (33 percent 46 percent and 43 percent of respondents from
nationally). This is highest at the regional level (42 each sector, respectively, citing it as an especially
percent), followed by federal (36 percent), and unsuccessful reform.
woreda (34 percent) levels. Across sectors, staff
33

3.4.1 Perceptions of the BSC of civil servants to some extent agree that the BSC
reform improved the way that their organization
113. The ECSS confirms that the BSC is the major prioritizes activities.
planning and evaluation tool in the civil service.
Tasks tend to cascade from the organizational 115. BSC as a feedback/evaluation tool:
scorecard and evaluations are based on the BSC Nationally, some 78 percent of heads consult BSC
indicators. The BSC reform appears to be embedded targets and reports at least every six months. Its
across government tiers, but is most present at the highest use, recorded at 100 percent, is by heads at
federal level, with civil servants there more likely the federal level. However, it is concerning that 30
to say that a higher proportion of their activities percent of regional and 19 percent of woreda heads
are affected by the BSC reform. The distribution never use BSC targets and reports. Civil servants
of agreement lies relatively constant across both consider the BSC to be an important feedback tool
sectors and tiers; however, two outliers stand out— (21 percent consider it to be very important, while
regional-level civil servants have the most positive 49 percent say it is quite important), suggesting
attitude towards the BSC, while civil servants in the that individual civil servants do take account of
agriculture sector have the least. the BSC process. Also, 59 percent of civil servants
agree that the BSC increased the efficiency of their
114. BSC as a planning tool: In all, 62 percent of directorate and 53 percent of civil servants to some
organization heads reported that they use BSC as extent agree that the BSC reform led to them being
a strategic planning tool. However, the fact that 24 better trained. However, around 60 percent of
percent of organization heads report limited or no civil servants agree to some extent that the BSC’s
use of the BSC as a strategic planning tool shows subjective component is non-meritocratic and
there is still a substantial gap in its use. From the unfairly biased by personal preferences (Figure
perspective of individual civil servants, there is a 3.5). This perception is particularly strong among
positive view of the BSC reform. Fifty-nine percent staff in the revenue sector.

Figure 3.5: Staff perception of unfair bias inherent in the subjective component of the BSC

Agriculture

Education

Health

Revenue

Trade

0 .2 .4 .6
Proportion

The proportion of all staff that responded Agree or Strongly Agree to the following:
The subjective component of the BSC is unfairly biased by the personal preferences of the supervisor.
 34

116. Furthermore, while staff are being 119. In terms of the Public Wing, 87 percent of
assessed and informed about their individual organization heads reported that the public
performance, there is little direct consequence wing members are involved in planning and
to good or poor performance. Training (38 evaluation processes. Heads are also positive
percent), followed by nothing (33 percent) are the about the implementation success of Public
most common responses to a very poor BSC score; Wing reform: meeting minutes are recorded
while non-financial rewards (37 percent), followed effectively, the reform is harmonized with other
by nothing (35 percent) are the most common activities, and is associated with an improvement
responses to a very good BSC score. A majority of in workplace practices.
directors or heads surveyed pointed out that they
do not have sufficiently clear guidance on how to 120. Yet, key informant interviews found that
reward staff based on BSC scores. understanding of the Public Wing members is
not complete, especially in the agriculture sector.
3.4.2 Perceptions of the Change Army Although the Public Wing has helped identify key
constraints in the health, education, and revenue
117. According to the ECSS, almost all organization sectors, it operates in an ad-hoc manner.
heads consider the Change Army’s State Wing to
have improved peer-to-peer learning, workplace 3.4.3 Perception on the Citizens Charter
practices and efficiency, client interactions, service
delivery, and the spread of best practices. The 121. According to almost all organization heads, the
survey results also suggest that the State Wing Citizens Charter has improved information flows in
has led to improvements in the flow of information the majority of cases, and clearly communicates
within organizations, likely as the result of the service delivery standards. Almost all heads
increased number of meetings and peer-to-peer agree that a clear complaint handling system and
interactions. Although regional- and woreda-level sufficient accountability exists because of the
organization heads report more frequent State Citizens Charter.
Wing meetings, the impact of the Change Army
on daily activities from the point of view of all civil 122. But the consolidated result for organization
servants is greater at the federal level. heads and civil servants draws a different picture.
The Citizens Charter is mentioned most when
118. Furthermore, 60 percent of organization respondents are asked to identify especially
heads considered the State Wing to be effective unsuccessful reforms (33 percent nationally).
in discussing and identifying solutions to rent- This is highest at the regional level (42 percent),
seeking in the civil service. Across all civil followed by the federal level (36 percent), and
servants, the majority selected the Change Army then the woreda level (34 percent). Among
(specifically, the ‘1 to 5’) as one of the most effective sectors, civil servants in agriculture and trade
tools for limiting rent-seeking, suggesting the view the Citizens Charter most negatively, with
reform’s perceived potential at dealing with this 46 percent and 43 percent of respondents from
issue. The only organization in which this issue is each sector, respectively, suggesting it as an
raised in a very limited way is the Federal Ministry especially unsuccessful reform. Similarly, while
of Agriculture, implying that more work needs to be around 60 percent of heads agree that the Citizens
done in this tier-sector to strengthen the Change Charter has improved workplace practices, the
Army reform. The Public Wing is much less cited response from the civil servants shows that the
by the civil servants as a form of dealing with rent- Citizens Charter has a limited influence on the
seeking, although it does appear more important daily activities of civil servants. This is true for all
in the trade sector at the regional level. three tiers of government.
35

3.5 Regional comparison of government effectiveness

123. Overall, as indicated in Figure 3.7 below, and even somewhat declined, in comparison to
Ethiopia, along with Rwanda, stand out as countries Rwanda, which has remained relatively steady.
that have made greater progress on strengthening Rwanda has deployed a particularly strong focus
Government Effectiveness over the past two on performance management (Versailles 2012),
decades, reaching the average level of a lower and stands out as a country that has made the
middle income country. However, in recent years, greatest degree of progress in the region.
Ethiopia’s average performance has stagnated

Figure 3.6: Government effectiveness by income blocks

80
70
60
50
40
30
20
10
0
High income Upper middle income Lower middle income Low income
non OECD countries countries countries
1998 2004 2009 2014

Source Worldwide Governance Indicators 2016.

Figure 3.7: Changes in government effectiveness in selected African countries – 1998-2015

80
70
60
50
40
30
20
10
0
Ethiopia Rwanda Liberia Cote d’Ivoire Tanzania Botswana

1998 2004 2009 2014 2015

Source Worldwide Governance Indicators 2017.

124. While the World Governance Indicators Ethiopian government set out in this report’s
(WGIs) are based on the aggregation of multiple introduction, that there is a need to reinvigorate
indicators, most of which are perception based, reform efforts, so as to be able to meet rising
the most recent trend gives an indication about expectations of citizens as well as manage
the need for further reform efforts in Ethiopia. and guide increasingly complex economic
This view complements the perspective of the development efforts effectively.
 36

3.6 Management practices within Ethiopia’s civil service

125. The reform tools discussed in the previous measured on a scale of 1 to 5, where 1 indicates
section aim to improve the performance of civil worst practice and 5 indicates best practice.
service organizations to achieve their objectives.
In order to have a better understanding of the 126. According to the ECSS, monitoring and targeting
current state of civil service organizations, the study are the relatively stronger management aspects
implemented a standard management measurement in Ethiopia, while flexibility, incentives, and staff
tool—the World Management Survey—across all involvement are the weakest. The best managed area
offices surveyed by the ECSS to assess the quality at the national level is monitoring (with an average
of their management practices. Key components score of 3.48) and the least is staff involvement
of management practices to be discussed in this (score of 2.0). Federal organizations exhibit the highest
section include targeting (setting targets for units management scores overall (average score 3.1), with
and individual staff), monitoring (based on targets education appearing as the best-managed sector
set), roles (autonomy afforded to staff), flexibility (average score 2.9). Addis Ababa, Amhara, and Tigray
(the ability to confront new demands and/or ways are the regions with the best-managed organizations
of working), incentives, staff contributions, and (average of score 3). Gambella shows the lowest scores
staffing. The quality of management practices is across all dimensions, with an average score of 2.1.

Figure 3.8: Average management scores across management dimensions

4
3
2
1
0

Monitoring Targeting Staffing Flexibility Roles

Incentives Staff involvement

Figure 3.9: Average quality of management (composite measure of all management dimensions)

15

10
Percent

0
1 2 3 4
Management
37

127. With the strong emphasis on the BSC and sense of direction and purpose. Employees benefit
Change Army, the performance of targeting and from clearly understanding what is expected from
monitoring should be expected above a level them. The ECSS confirms that through the BSC,
of 3, i.e. above ‘average’, for a larger number of tasks tend to cascade from the organizational
organizations. Similarly, the autonomy of roles scorecard and evaluations are based on the BSC
and staffing might be expected to perform more indicators. For more information, please refer to
strongly as the result of the BPR reform, as the BPR the section on perceptions of BSC in this chapter.
aimed to empower front line workers and deploy
the right staff for the right jobs. It is relatively 129. Figure 3.10 below demonstrates a positive
less surprising that flexibility, incentives, and staff correlation between those organizations that
involvement are weaker areas, as these have not implemented the BSC reform with greater
been a focus of recent reforms and these areas intensity (that a greater proportion of activities
might benefit from greater attention in the future. are stated to be influenced by the reform) and
the increased use of performance targeting by
3.6.1 Targeting the management. This figure provides suggestive
evidence that the BSC reform is associated with
128. Setting targets and breaking them down for a higher score in the World Management Survey
units and individual staff members provides a targeting indicator.

Figure 3.10: Relationship between the intensity of BSC reform

implementation and the use of performance targeting by management

2
World Management Targeting Z-Score

1
[Organization average]

-1

-2 Raw data
Filled values
0 20 40 60 80 100

What percentage of your activities in the civil service has been substantially
affected both positively and negatively, by the following reforms?
[Organization average]

R-squared=0.2328
 38

130. Overall performance on targeting is it is not always clear how the targets contribute to
‘middling’ on average, with a score of 3. With their organization’s goals. Importantly, the tasks
the strong emphasis on BSC, one might have assigned to staff on a day-to-day basis are not
expected a performance on targeting above always related to those targets. Targeting is best
‘average’, for a larger number of organizations. managed at the federal-level (average score of
This score indicates that overall, organizations in 3.8) followed by the regional (3.1) and woreda (2.9)
the civil service assign targets to their directorates, levels. The distribution of targeting quality across
which are then broken down to managerial and sectors shows Afar to be the worst-managed
employee-level targets. These are generally well region in terms of targeting (average score of 2.3),
understood by mid-level staff, although they are and Addis Ababa as the best (3.8). The distribution
not necessarily communicated in a clear and is similar across sectors (ranging from an average
concise manner to lower-level staff, which suggests score of 2.8-3.4).

Figure 3.11: Targeting

40

30
Percent

20

10

1 2 3 4 5

Targeting

3.6.2 Monitoring below. The BSC reform introduces targets, which

feed into planning, but also function heavily as a
131. The main purpose of monitoring is to track monitoring and evaluation tool. This is suggested
implementation and outputs systematically to in the data, as organizations who applied the BSC
determine exactly when targets are on track with a greater intensity (that a greater proportion
and when changes may be needed. The BSC of activities are stated to be influenced by the
reform is positively associated with the use of reform) display greater scores in the monitoring
performance monitoring, as shown in Figure 3.12 indicator of the World Management Survey.
39

Figure 3.12: Relationship between the intensity of BSC reform

implementation and the use of performance monitoring by management

2
World Management Monitoring Z-Score

1
[Organization average]

-1

-2 Raw data
Filled values
0 20 40 60 80 100

What percentage of your activities in the civil service has been substantially
affected both positively and negatively, by the following reforms?
[Organization average]

R-squared=0.3199

132. The mean score for monitoring management may be shared with relevant staff members, there
is 3.5. This indicates that directorates generally is limited communication about performance and
track a limited number of performance indicators how is it is monitored and reviewed. This suggests
somewhat regularly, which are seen and reviewed by that although some monitoring is taking place at the
senior management only. Although some of this data directorate-level, it is done inconsistently.

Figure 3.13: Monitoring

40

30
Percent

20

10

0
1 2 3 4 5
Monitoring
 40

133. Again, monitoring is best managed at the BSC report score are limited.
federal level, for which we find an average score
of 3.7, followed by 3.5 at the woreda level and 135. The mean score for performance
3.4 at the regional level. The education sector monitoring and incentive system is 2.4. This
appears to have higher quality monitoring, indicates that performance is evaluated through
with an average score of 3.7. Addis Ababa and a formal system and good performance could be
Tigray are the best managed regions in terms of rewarded (financially or non-financially), however
monitoring (average score 4), followed closely there is no system or clear criteria for doing
by Amhara (3.9). Gambella lags behind with an so. Similarly, poor performance is addressed
average score of 2.6. inconsistently and on an ad-hoc basis. In such
cases concrete action may be taken to rectify
134. Performance monitoring and links to the problem, but it is not consistently followed-
incentives. The Individual Performance Appraisal through. Again, lack of discipline or bad behavior
System was added to the BSC in 2012. Around by employees is similarly managed. According
60 percent of civil servants agree to some to the respondents, the most common response
extent that the behavioral component of the to a very poor BSC score by management is
individual employee performance management assigning training (38 percent) or doing nothing
and measurement is non-meritocratic and (33 percent). Similarly, the most common
unfairly biased by personal preferences. Even managerial response to a very good BSC score
so, when the BSC is used as an evaluation tool, is granting non-financial rewards (37 percent)
the consequences of a very good or very poor or doing nothing (35 percent).

Figure 3.14: Incentives

50

40

30
Percent

20

10

0
1 2 3 4 5
Incentives

136. The federal level has the best employee of 3.2, followed by Tigray with a score 3, and
performance management system, with a score Gambella scoring lowest, with 1.4. Thus, while
of 2.7. Again, scores across sectors are consistent, the monitoring of targets is done fairly well, it
but less so across regions, with Addis Ababa once does not yet translate into effective employee
again leading the ranking with an average score performance management.
41

3.6.3 Roles Decisions on staffing and task assignments are

subject to managers’ personal knowledge of staff’s
137. The mean score for roles (measuring skills and competencies, rather than on accurate.
the level of autonomy staff are given in up-to-date documentation on each staff member.
undertaking tasks) is 2.5. This indicates that very This means that not all staff are engaged equally
little discretion is afforded to officers to carry out and that they may be assigned to tasks to which
their assignments on a day-to-day basis. Staff they are not suited. Some staff members or staff
contributions towards decisions on how to carry groups are more burdened than others depending
out tasks, as well as in the implementation of on a) their relationship with the manager, and b)
policy is infrequent and done in an ad-hoc manner. the type of work they are responsible for.

Figure 3.15: Roles management

50

40

30
Percent

20

10

0
1 2 3 4 5
Roles

138. The management of role autonomy is once working) is 2.5. This indicates that when
again highest at the federal-level (average score directorates are confronted by new demands
2.9). The distribution is very similar across sectors, or the specific requirements of communities,
but less so across regions, where there is almost a clients, or other stakeholders, efforts are
one-point difference between the best-managed occasionally made to tailor procedures to
region, Addis Ababa (score 3), and the worst, Afar, those specific needs, although this is not
Gambella and SNNPR (score 2.1). done frequently and can present significant
difficulties when those needs or demands are
3.6.4 Flexibility complex. New ideas or practices are adopted,
but in an informal and/or isolated manner.
139. The mean score for flexibility (the ability Generally, directorates are slow to integrate
to confront new demands and/or ways of new practices into their operations.
 42

Figure 3.16: Flexibility

50

40

30
Percent

20

10

0
1 2 3 4 5
Flexibility

140. The federal sector shows the highest 3.6.5 Staffing

Staffing score in flexibility, with an average
of 3 (2.6 and 2.3 at the regional and woreda 141. The mean score for staffing is 2.7. This indicates
levels, respectively). Again, the distribution that when it comes to talent in the organization,
is very similar across sectors but much less senior management believes that attracting and
across regions: Addis Ababa and Tigray are at developing talent is important, but there is no clear
the top with scores of 3 and 2.9 respectively, system for identifying, attracting, or retaining such
and Gambella (1.9) and SNNPR (2) are at the talent. The promotion is based on performance
bottom of the ranking. and tenure, although this is more heavily weighted
towards performance. There are limited opportunities
for promotion and very little room for growth.

Figure 3.17: Staffing

50

40

30
Percent

20

10

0
1 2 3 4 5
Staffing
43

142. The federal level scores a 3.1, regional level that because there is no system for identifying
2.8, and woreda level 2.6. The education sector and solving problems, staff are not involved
shows higher management practices than the in providing solutions and making decisions.
other sectors. Addis Ababa, Tigray and Amhara Suggestions may be taken from staff, but these
score highest by a relatively large margin, and occur on a rare and ad-hoc basis. Similarly, non-
once again Afar and Gambella score lowest managerial staff do not actively contribute in staff
(almost one point below average). meetings and rarely provide any kind of feedback.

3.6.6 Staff involvement 144. There is almost no variation in the quality

of staff involvement management across tiers or
143. The mean score for staff involvement (the sectors, and only slight variation across regions:
ability of staff to become involved in operations once again Addis Ababa and Tigray are the
and contribute towards decisions) is 2, by far the highest scoring (with 2.5 and 2.4 respectively),
lowest scoring management area. This indicates and Gambella is the lowest, with a score of 1.7.

Figure 3.18: Staff involvement

60

40
Percent

20

0
1 2 3 4 5
Staff involvement
44
45

4
The State of the
Service Today
 46

4.1 Recruitment and management of staff

145. The recruitment and management of staff educational qualifications can pose a challenge.
is a critical aspect of a public sector. Merit- Once staff have been brought into the public
based recruitment (i.e., not based on personal sector, merit-based treatment and promotions
relationships) has long been recognized as central are important to set incentives for performance, as
to an effective government (Fukuyama 2011). In well as identify and promote capable and effective
particular in low-income contexts, the availability leaders.
of a potential pool of applicants with sufficient
47

4.1.1 Performance management recruitment processes had not been followed).

practices in the civil service Sector-by-sector, formal recruitment processes
appear to be mostly followed in the education
146. Hiring and other staffing decisions are made by sector across levels of government, but somewhat
individual MABs, in line with procedures set out in less so in agriculture, health, and revenue sectors
the law, and as prescribed by the Ministry of Public at the regional and woreda level, and for trade at
Service, respectively the Bureaus of Public Service the regional level.
at sub-national levels. A merit-based selection
is prescribed in principle, and available evidence 149. Probation periods of six months are common.
suggests that patronage-based hiring is not as Seventy-eight percent of heads of agencies
much a concern as in a number of other countries reported this length of probation periods, while
in the region. Rather, key problems with recruitment only 2.3 percent reported no probation period.
appear to be that on the one hand, MAB managers
believe that they cannot consistently attract qualified 4.1.2 Treatment of staff
applicants, and on the other hand, some anecdotal
evidence suggests that some selections are not fully 150. Respondents were asked to rank the most
based on meritocratic criteria, but also on political important determinants of how staff are treated
considerations and other factors. including who gets hired, receives the best
rewards and training opportunities and so on, in
147. Educational opportunities have been greatly the organization. Figure 4.1 below shows the most
expanded over the past two decades, and most important perceived determinants of how staff are
professional-level public servants have completed treated. Similar to responses for recruitment as such,
post-secondary training. Formal qualifications of merit and performance are the top factors by a large
civil service employees are adequate (diploma or margin (59 percent nationally), followed by the quality
first degree of tertiary education). In the survey of the relationship with supervisors or managers (11
targeting professional staff as well as heads of percent), and political connections (6 percent).
agencies, over 99 percent of respondents held a
post-high-school qualification, and 82 percent had 151. Merit and performance is perceived to be
at least an undergraduate qualification. However, equally important across all levels of government
there are issues with the quality of training, as on average, but with variation across sectors, and
discussed further below. especially regions. Sixty-five percent of revenue
staff perceive merit and performance to be the
148. Based on the survey, the main channels for key determinant of treatment, compared to 53
becoming aware of a position are job boards percent of health staff. The difference across
(especially for regional and woreda levels), adverts regions is even more pronounced: While 91
in newspapers (federal level), followed by transfers percent of respondents in Tigray consider merit
and informal invites. Online advertisements so far and performance the key factor determining
play a very minor role. Formal selection processes treatment, only 23 percent of Gambella-based
are very common at the federal level, and less respondents do. Interestingly, Addis Ababa,
so at regional and woreda levels (28 percent which ranks top in most management metrics
responded that there was no formal process at (see section 3.4), comes second to last in the
the regional level, and around 21 percent for the importance of merit and performance as a
woreda level). Formal recruitment processes are determinant of treatment, with only 37 percent
less frequently followed in emerging regions, in (it is still the most important criteria, but by a
particular in Gambella, Harari, Somali, and Afar significantly smaller margin). The second most
(45 to 55 percent of respondents saying that formal important determinant of staff treatment in Addis
 48

Ababa is political connections (26 percent). managers is considered the primary determinant
The importance of political connections is of treatment by some federal staff (14 percent),
significantly lower in all other regions, a distant and regional staff (8 percent). Across sectors, this
second being Harari with 11 percent. Political is perceived as most important by agriculture staff
connections are especially unimportant in Somali (13 percent) and least by revenue staff (9 percent).
(1 percent) and Tigray (2 percent), and entirely The region where this is mentioned most frequently
insignificant in Dire Dawa. is Afar, with 22 percent of respondents identifying
it as the primary determinant. It is considered least
152. The quality of the relationships with important in Tigray, with 2 percent.

Figure 4.1: Top determinants of how staff are treated

Merit/performance

Quality of relationship with manager

Political connections

Education connections

Ethnicity

Length of service in the public sector

Length of service in this organization

Family connections

Office conduct/behaviour

Other connections

Providing gifts or unofficial payments

0 20 40 60
Percent

4.1.3 Is staff promotion merit based? 154. There is some noticeable variation across regions
and sectors. Regionally, merit is commonly among
153. Generally, there is some confidence that the top three most important criteria for promotion,
staff promotions are merit based, followed by except in Afar, Benishangul-Gumuz, and Gambella
considerations about the length of service both (Figure 4.2). On average, merit appears to be a
in the public service in general and in a specific slightly less important criterion at the regional level
organization. Those surveyed percieve the top (74 percent) than at federal (78 percent) and woreda
five most important criteria for promotion to (76 percent) levels. Across the five sectors, merit is
be merit/ performance, quality of relationship seen as most important in the education sector (81
with manager, length of service (in the public percent) and least in the health sector (69 percent).
sector and in the organization), and political The lowest views about merit-based promotions are
connections. expressed by regional health sector staff.
49

Figure 4.2: Importance of merit as a criterion for promotion. By tier and sector. By region

1
Merit / performance / education

.8

.6

.4

.2

Regional Woreda

Government Tier

Afar Amhara Benishangul Gambella Oromiya SNNPR Somali

Tigray Harar Dire Dawa Addis Ababa

Rank the 3 most important criteria for promotions in your organization.

(Indicator if chosen among top 3.)

Figure 4.3: Importance of merit as a criterion for promotion. By tier and sector

.9
Merit / performance / education

.8

.7

.6

.5
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

Rank the 3 most important criteria for promotions in your organization.

(Inidcator if chosen among top 3.)
 50

155. Relationship with supervisor. The importance service is considered a slightly more important
of one’s relationship with the supervisor is the criterion for promotion than length of service
second most important perceived criteria for in the organization (8 percent and 7 percent
promotion, and is most strongly felt at the respectively). Tenure in both the service and the
federal level (13 percent). The importance of this organization is perceived as a more important
relationship on promotion is more significantly factor at the federal level, followed closely by
cited among federal level trade staff and least the woreda and regional levels. Length in the
among revenue staff at the regional level (2 organization is more likely to be perceived as
percent). This criterion is by far most important a key factor among revenue staff (11 percent).
in the Afar region, being cited by 26 percent of While length of service in the organization is
respondents, followed by Amhara (15 percent). relatively most frequently cited as important in
It is not considered important in Gambella or Addis Ababa (21 percent) and length of service
Benishangul-Gumuz, where only 3 and 4 percent in the civil service in Benishangul-Gumuz (20
of respondents respectively cited it as a criterion percent), neither is perceived as a criterion at all
for promotion. in Harari (not a single respondent identified this
is as criteria).
156. Length of service in the public sector and
in the organization. Length of service in the civil

Figure 4.4: Importance of length of service in the civil service. By tier and sector

.8
Length of service in the public sector

.6

.4

.2

0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

157. Political connections. Political connections considered at all important (with no respondents
are considered the fifth most important criteria citing this as a criteria) in the health and trade
for promotion. This is highest at the regional sectors at federal level. Across sectors, political
level (8 percent) and lowest at the federal level connections are of high importance in Addis
(5 percent). Agriculture and education staff cite Ababa, where such connections are perceived as
such connections slightly more frequently (9 and 8 the principal criteria for promotion according to
percent respectively). Political connections are not 32 percent of respondents. Political connections
51

are significantly less important in all other regions, tiers. Civil servants in the Somali region are by
the next highest being Oromia with 12 percent. far the most confident that if they perform well
Political connections are least important in SNNPR they will get promoted, with 82 percent, compared
(1 percent) and Tigray (2 percent). to 26 percent in Gambella. Trade staff are most
confident (with 55 percent) and revenue staff least
158. While civil servants perceive that promotions confident (with 40 percent). Moreover, what stands
are overall merit based, confidence in actually out is that a substantial share of staff (around
receiving a promotion is more limited and 30 percent across levels) expresses the view that
divided. Around 45 percent of civil servants are they are very unconfident about being promoted,
confident that they will get promoted if they even if performing well; with an additional 12-18
perform well. The distributions are similar across percent of staff being ‘somewhat unconfident’.

Figure 4.5: How confident are you that you will get promoted if you perform well?

Federal

50
40
Percent

30
20
10
0
Very Somewhat Neither Somewhat Very
unconfident unconfident confident confident

Regional

50
40
Percent

30
20
10
0
Very Somewhat Neither Somewhat Very
unconfident unconfident confident confident

Woreda

50
40
Percent

30
20
10
0
Very Somewhat Neither Somewhat Very
unconfident unconfident confident confident
 52

159. One aspect to keep in mind when considering to those who work hard to achieve the goals of
promotion in Ethiopia is that the level at which the FBR.” Twenty-nine percent of officials in the
political appointees start is lower than in most Philippines stated that their department was one
other countries. Political appointees start at the in which “Most promotions went to people who did
Director level, rather than being formally limited not meet the formal qualifications for promotion,”
to the level of Minister and State Minister. This can and 28 percent agreed that, “Promotions are mainly
pose some limits for professional but non-political given to those that have friends and family at higher
staff in terms of promotion opportunities. levels in your department.” Where we find the least
evidence for merit-based advancement, in Pakistan,
4.1.4 Merit-based advancement: officials state that ‘managerial favoritism’ is what
a cross-country comparison is driving career paths. Even in the Philippines,
over 50 percent of staff agree that, “Favoritism
160. Table 4.1 summarizes officials’ beliefs around among managers and employees often makes
whether advancement in the civil service is based it difficult for public officials to perform in their
on merit.41 Looking at promotion specifically, jobs.” Overall, Ethiopia is in the middle among the
50 percent of civil servants in Pakistan disagree various countries in terms of perceptions among
with the statement that “Promotions/bonuses go civil servants that promotions are merit-based.

Table 4.1: Perceived criteria for advancement in the civil service across countries

All (1) (2) (3) (4) (5)

Ethiopia Ghana Indonesia Pakistan Philippines

Promotions are based on merit 0.65 0.56 0.87 0.89 0.22 0.75
[agree=1]

Other criteria for advancement

Merit 0.71 0.56 0.75 0.9 0.61 0.76

Tenure 0.47 0.14 0.67 — — —

Connections 0.25 0.12 0.25 0.39 0.17 0.35

Bribes 0.02 0 0.12 0.1 0.01 0.02

41 Note that there is some variation in the question and answer choices among the surveys in different countries, so the
results should be considered indicative. In Indonesia, Pakistan and the Philippines, the relevant question on promotion
was “Rewards/Promotions go to those who work hard to further the goals of [the department/organization]”. In Ghana,
the relevant question was “In the past three years, have elected officials, their appointees, or political party officials tried to
influenced any hiring decisions and or promotions in your organization?”. The Ghanaian question was the best available,
but includes hiring considerations. They are also phrased positively and negatively respectively, which may change the
responses of officials. In Ethiopia, the relevant question on promotion was more direct: “Rank the three most important
criteria for promotions in your organization”.
53

161. Tenure and connections are the other two an insignificant role in the countries we studied
factors that are frequently selected across and is completely absent in Ethiopia, but direct
countries. Tenure is frequently highlighted as questioning may not be the most appropriate
important in the civil service (it is high for Ghana, way to get at this topic and advances in survey
the only other country for which we have this methodology may be required to gain more
data, with 0.67); and is comparatively much less credible estimates of the effects of within-
important in Ethiopia (14 percent). Bribes play service bribery.

4.2 Drivers of staff turnover

162. There is a general concern about substantial 164. The survey also indicates that staff turnover
staff turnover in Ethiopia, although actual turnover is particularly acute at the federal level in the
is relatively limited, with variation by sector and health and trade sectors. Respondents recall that
level of staff. As a previous survey and turn-over six to seven staff left their respective directorates
study from 2014 indicate, while overall voluntary over the past 12 months. The type of staff who
turn-over is lower than perceived, it is concentrated leave are perceived to be mostly professionals,
in certain regions and sectors, notably Addis Ababa a finding that is consistent with the 2014 study
and Benishangul-Gumuz, and the revenue and on turnover.
justice sectors.42 Departing staff were mostly in
the professional category, and it often took a long 165. Figures 4.7 and 4.8 reflect the views of heads
time to replace staff, contributing to high vacancy of organizations and staff, respectively, with
rates. The turnover survey also indicated a high regard to motivations for turnover, and both
prevalence of low motivation, an issue that is point to a lack of opportunities for promotion
explored in greater detail in the 2016 ECSS. as the main reason. Directors cite the following
reasons for why previous staff have departed: 1)
163. The 2016 ECSS indicates that just under half lack of opportunities for promotion, 2) working
of respondents are dissatisfied or very dissatisfied, climate, 3) poor pay and benefits, 4) workload,
and many intend to change jobs. Satisfaction is and 5) problems with one’s manager. As is
lowest at the federal level, followed by the woreda discussed further in Chapter 5, dissatisfaction
level, and is somewhat higher at regional levels. with salaries and benefits appears to be the
On the question of whether staff plan to stay or most important factor impacting motivation;
leave their current positions, over sixty percent while with regards to turnover, promotion
of staff say that they intend to change their jobs opportunities play a significantly greater role.
within the next two years (71 percent at the federal For staff considering to change jobs, the top five
level). While actual voluntary departures tend to issues are 1) limited promotion opportunities,
be significantly lower than the intention to leave, 2) conditions of service apart from wages, 3)
such a high level of ‘intentions to leave’ poses risks, poor training and development opportunities,
both in terms of actual turnover materializing, as 4) wages, and 5) limited opportunity to have
well as demotivated staff remaining in post. an impact.

42
The study was undertaken in 2013-14 and based on two surveys of: i) 474 HR Officers responsible for 9 percent of the
total civil servants across the three levels of government; and ii) 2,307 civil servants randomly selected from professional
groups that leave organizations the most.
 54

Figure 4.6: Top 5 drivers of turnover (directors/heads)43

Lack of opportunities for promotion

The working climate is not appealing

Poor pay and benefits

The workload is too much

Problem with the manager

0 10 20 30 40
Percent

Figure 4.7: Reasons for wanting to change jobs (staff)44

Limited promotion opportunities

Conditions of service apart from wage

Poor training and development opportunities

Wage

Limited opportunity to have impact

Other

Limited responsibility

Office space/working environment

Role does not match skill set

Work is not interesting

Culture

0 10 20 30 40
Percent

166. Lack of opportunities for promotion. federal-level agriculture staff perceive it to be the
Directors cited the lack of opportunities for main driver, it is not an issue at all among federal-
promotion most frequently at the regional and level trade staff (not a single respondent cited the
woreda levels. Across sectors, while 50 percent of lack of promotion opportunities as a driver). There

43
This question was asked to Directors only: “What do you think are the greatest drivers of turnover? List all relevant responses.”
44
“Which of the following issues are reasons you want to change jobs? Is it because of the…” for those who answered yes to
“In the next two years, would you want to change your job?” to employees only.
55

is significantly less variation at the woreda level. frequently at the federal level with 28 percent of
Dire Dawa stands out as the region where a lack of directors seeing it as the main driver, compared
promotion opportunity is especially widely cited to only 11 percent at woreda level. At the federal
(57 percent), whereas it is not a significant issue level, revenue and trade directors perceive wage
in Tigray (8 percent). and non-wage benefits to be the primary driver
of turnover (both of them with 42 percent). The
167. Among staff, limited opportunities for sector-pattern holds at the woreda level (although
promotion were especially pointed out by staff much less significantly), but not at the regional
in the agriculture and health sectors across levels. level where limited wage and non-wage rewards
Career structure schemes, which allow predictable are perceived as the main drivers of turnover
promotion upon satisfactory performance without especially in the education sector. Perceived
requiring a vacancy, have been implemented poor pay and benefits were overwhelmingly the
in a few professional groups (teachers, health most cited reason for a decrease in motivation, as
professionals, agriculture extension workers). captured by the open-ended questions (see also
Similar career structures are not in place for section 5.1 below).
administrative staff working in the same ministries
and offices, which could explain the particular 169. The issue of poor training and development
dissatisfaction in these sectors. The government was cited by staff as the third most important
is now working to put in place an analytical point reason for considering job changes, and
factors job evaluation scheme but until the law is significantly more frequently at the regional
amended accordingly, the Position Classification level. As is discussed further in section 5.2 below,
System remains the only legally binding method. for about 30 percent of staff, opportunities for
learning are an important point of attraction to the
168. Poor pay and benefits is the third most public sector. Thus, expanding such opportunities
important perceived driver of turnover by could be considered as one of the ways through
directors, while among staff, other conditions which to strengthen continuous staff development
of service (such as holiday allowance or leave, and motivation.
health insurance provision, or transportation
allowance) are cited more frequently than wages 170. A problematic relationship with a manager
as such. Poor pay and benefits are cited most and too high a workload are the other two of the

Lack of opportunities for promotion is

reported as one of the main reasons for
wanting to change jobs.
 56

top five drivers of turnover perceived by directors, workers between directorates. Moreover, 133 out
but are rarely mentioned by staff. Problematic of 312 woreda heads of organization (43 percent)
relationships appear less important at the regional exercise a regular reallocation of one or more
level (7 percent) than at the federal and woreda employees between directorates. The distribution
levels (15 percent and 13 percent respectively). of responses across sectors shows that around
Across sectors, it is more likely to be perceived half of organization heads in each sector state that
as a driver of turnover in the health (18 percent) they conduct a regular reallocation of one or more
and trade (17 percent) sectors. Too much work workers across directorates, except for revenue
is a slightly greater issue at the federal level, where this is much less common – only 22 percent
with 25 percent than at the regional (17 percent) of organization heads stated that they do. Rotations
and woreda (20 percent) level. There is slightly across sectors are much less common. Two out of
more variation across sectors, with 32 percent three federal ministers stated that they do not do
of revenue staff perceiving this as a driver of this; 31 out of 37 regional bureau heads; and 227
turnover, compared to only 14 percent in trade. out of 312 woreda-level heads.

171. At least a part of the observed turnover of 172. In terms of impact, turnover was mentioned
staff is directly within the control of respective as limiting the productivity of remaining staff
Ministries and Bureaus. The survey asked especially for three federal-level sectors: trade,
organization heads the extent to which they revenue, and health. For all other sectors and
regularly moved one or more staff around in their levels, other factors were selected more frequently
organization. One out of three Federal Ministers in terms of negatively impacting productivity.
stated that their organization employs a regular These responses are somewhat consistent with
reallocation of one or more workers between the issues raised by members of Public Wings,
directorates. At the regional level, 19 out of 37 (51 which pointed out that there are technical gaps
percent) bureau heads stated that their organization and capacity limits due to significant staff turnover
exercises a regular reallocation of one or more in the Ministry of Health.45

4.3 Staff capacities - diverging perceptions

173. A vital issue raised in key informant interviews 174. Several reasons are seen as contributing to the
is that as a consequence of the very rapid concerns about quality. One is the fact that many
expansion in training and hiring staff into the tertiary institutions were recently established and
public service, quality has declined. This concern have young and not sufficiently trained faculty.46
is shared across different technical specialties, Second, given the rapid expansion of students
including education, health, and agricultural staff, to be trained, funding as well as management
as well as other professions. A significant share attention at tertiary-level institutions has become
of staff whose qualifications have been tested in absorbed by administrative issues, such as
service through competency assessments have running dormitories and hiring additional teaching
failed to exhibit sufficient basic knowledge. staff as opposed to focusing on strengthening

45 While the Public Wing structure of the Ministry of Trade seems to be rather weak overall and did not address the issue.
46 In recent years, 33 new Universities were created, and addition, many colleges and other tertiary training institutions
were established.
57

curricula and teaching methods. Third, there are at woreda levels, only 3 percent of directors have
insufficient funds to ensure that students receive full access to computers/IT systems).
practical training to complement the theoretical
knowledge being acquired. This negatively affects 178. There are several possible reasons for the
the actual skills that those graduating and then discrepancy between widespread concerns about
being recruited into the public service can bring the lack of training and skills on the one hand,
to their assignments – despite the fact that formal and staff self-perceptions on the other hand. One
qualifications have improved as noted above. is that the 2016 survey targets staff in ministries,
bureaus, and offices rather than actual front line
175. A lack of technical knowledge among staff service delivery staff such as teachers, nurses, or
in federal-level ministries was also raised in agricultural extension workers, and so practical
several discussions with Public Wing members. experience is less critical. It may also be that
In particular, members of the health sector’s bureau and ministry positions attract and recruit
Public Wing noted significant capacity gaps in better trained staff. Another possibility is that
the responsible ministry, which they see as being staff overestimate their own skill levels. Including
due to a lack of technical specialists as well as bureau-level staff in competency tests may be one
staff turnover. Members of the agriculture sector’s way to assess whether the latter is a significant
Public Wing also pointed out that sector ministry issue. ICT skills clearly are particularly relevant for
staff often lack in-depth knowledge about the bureau staff and should receive attention.
issues they work on.
179. Plans are being made to make the testing
176. In contrast to the concerns about quality and of civil service staff more widespread, followed
adequacy of skills expressed in interviews and by mandatory additional or remedial training,
focus group discussions, civil servants covered in e.g. during the summer. This process currently
the ECSS perceive themselves as having adequate remains at the initial stages, and the potential
skills for their jobs. Only 2.1 percent of staff cited costs of such efforts are substantial. Moreover,
their lack of the right skill set for their tasks as some effort would need to focus on strengthening
a key constraint to their effectiveness. Only 22 tertiary training institutions in the first place so
percent of staff were aware of competency tests as to reduce the continuation of the problem of
for their area of work; while among those aware insufficient initial training.
of such tests, most agreed that they evaluate
appropriate skills for the job. 180. For federal-level staff, it appears to be
critical that staff are truly well prepared in the
177. While technical skills and knowledge are a subject matter they are responsible for, and
primary concern, there is also a specific issue remain in their post long enough to acquire
with the ability to use the IT applications that are strong knowledge. Federal-level staff fulfill
being rolled out. In 64 percent of the organizations critical policy roles. In successful developmental
surveyed, fewer than half of the staff in a given states such as South Korea, there was a strong
unit are able to create an Excel spreadsheet or emphasis on having competent and well respected
a Power Point presentation, and in 40 percent staff in place in key roles so as to ensure that
of organizations, fewer than half of all staff are effective guidance could be provided. Moving
able to use a computer to write a memo. Still, 74 in this direction would also imply that rotations
percent of all staff surveyed agreed that they have need to be appropriately staggered to avoid a
the necessary skills to make use of the technology loss of knowledge.
that is available to them (while as noted below,
 58

4.4 Use of databases and client feedback

4.4.1 Do staff make use of available of organizations at the federal level, 70 percent at
databases? regional level and 40 percent at woreda-level. The
health sector has the highest prevalence of MIS with
181. Just over half of organizations surveyed (52 84 percent, significantly higher than the next highest
percent) have a functioning MIS in place: 80 percent sectors, trade (48 percent) and education (47 percent).

Figure 4.8: Do you have a Management Information System (MIS) programme in place and functioning?

Federal

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Regional

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Woreda

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable
59

182. However, even among organizations that (51 percent), and most notably in the health sector
officially have an MIS in place, accessing the (50 percent). The lack of usefulness of the data
data can be difficult. Two challenges to accessing in the MIS is the second most cited challenge (23
that data stand out: network problems and the percent). Agriculture and education sector staff
lack of usefulness of the data available. Network are more likely to suggest that the MIS does not
problems (39 percent) is the most cited challenge have any useful data, particularly at the federal and
to accessing data, especially for regional bureaus regional level (40 and 50 percent. respectively).

Figure 4.9: “What would be the main challenges to getting the data?”

.8 .7
Directorate does not have any useful data
MIS (Management Information System)

.6
.6
Network problems

.5

.4
.4

.2 .3

.2
0
re

or

ue

re

or

ue

e
ad

ad
tio

tio
tu

tu
ct

ct
n

n
Tr

Tr
ve

ve
ca

ca
Se

Se
ul

ul
Re

Re
ic

ic
u

u
th

th
Ed

Ed
r

r
Ag

Ag
al

al
He

He

Federal Regional Woreda

183. Relative to other sources of information, MISs (18 percent), and informal interactions such as those
still rank relatively low. The three main sources of with co-workers (16 percent). The MIS is only used
information used by staff to find out about the state as the primary source of information by 9 percent
of service delivery in their jurisdiction are formal of civil servants nationally: 15 percent at the federal,
field visits (22 percent), reports from the frontline 14 percent at the regional, and 6 percent at woreda

Only 9% of civil servants, nationally,

use the MIS as the primary source of
information
 60

levels. Health sector staff reported using the MIS disagreeing strongly. Eighty percent of directors
as a main source of data the most frequently (17 also agree that the MIS allowed them to learn new
percent) compared to agriculture sector staff, who, things about their jurisdiction, and that it feeds
at 5 percent, are the least frequent users. into the performance review (similar distributions).

184. In terms of impact, at the organization level, 185. However, directors do not believe that
around 85 percent of directors agree to some employees worked harder because they had
extent that the introduction of an MIS increased more information on what works, with 64
the information employees of their organization percent of directors disagreeing to some extent
had access to regarding the state of service (34 percent of them strongly). Disagreement is
delivery in the jurisdiction. There is a similar highest at the woreda level (66 percent) and
level of agreement at all tiers of government lowest at the federal (46 percent). Of similar
(84-86 percent). The health and revenue sectors importance is the high degree of disagreement
shows a greater degree of agreement, with 94 among directors as to whether employees worked
percent and 87 percent agreement respectively. harder because they were being monitored more
The agriculture sector shows a significantly lower (64 percent disagree). This is highest at the
degree of agreement, with 18 percent of directors regional level (66 percent).

Figure 4.10: To what extent do you agree that the introduction of the MIS increased the information
employees of their organization had access to regarding the state of service delivery in the jurisdiction?

Federal

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Regional

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Woreda

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable
61

Figure 4.11: To what extent do you agree that the introduction of the MIS has made
individuals work harder because they had more information about what works?

Federal

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Regional

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

Woreda

100
80
Percent

60
40
20
0
Strongly Slightly Neutral Slightly Strongly Not
disagree disagree agree agree applicable

4.4.2 Attention to clients’ feedback 187. Feedback from citizens is indicated as very
and complaints important across all tiers. It is considered most
important in the revenue sector (96 percent), and
186. Respondents were asked how important in the Dire Dawa, Harari and Somali regions, where
they felt different types of feedback were in all respondents consider citizen feedback either
determining how well they were doing as a ‘quite important’ or ‘very important’. Rates are
civil servant. The three most important types of slightly lower at the regional level (89 percent),
feedback nationally are feedback from colleagues but still high.
(97 percent), feedback from citizens/clients (94
percent), and feedback from supervisors (94 188. Further, 80 percent of respondents believe
percent). Respondents were given the option to there is a clear and effective complaints
specify how important each type of feedback is management system in place, with 46 percent
(‘very’ or ‘quite’), and the option ‘very important’ agreeing ‘strongly’ that this is the case. A high
was higher for citizen feedback than it was for percentage of respondents (77 percent) also
supervisor feedback. believe that there is sufficient accountability and a
 62

clear course of action to be taken when complaints some caveats. The highly expressed importance
are received. As noted above, civil servants also accorded to citizen feedback is somewhat
consider citizen complaint mechanisms to be an inconsistent with the fact that, as discussed
important tool for addressing corruption. in Chapter 3, Citizens Charters are seen as a
relatively less successful reform, given that
189. Feedback from supervisors and managers complaints would in part be based on the service
is similarly important at all tiers of government terms outlined in such charters. An insufficient
and sectors. It is considered most important ‘service attitude’ was also repeatedly raised by the
in the Harari and Dire Dawa regions, where all key informant interviews as one of the continuing
respondents identify it as the principal form of obstacles to better service delivery.
feedback, and least in Tigray, where 90 percent
considered it so. 191. Going forward, it would be highly desirable
to survey service users as a way to identify
190. Generally, this suggests that the civil service whether citizens similarly feel that their feedback
reform efforts that were made over the past two is sought and effectively taken into account, and
decades have had a positive impact in terms to ensure that civil servants are aware of the
of the attitudes of civil servants, albeit with results of such surveys.
63

5
Constraints to
Public Service
Delivery Identified
by Civil Servants
 64

192. Civil service reform tools, as discussed and other available sources.
in previous chapters, can bolster government
performance; however, a range of constraints not 193. The quality of the service delivered by any
directly targeted by such reforms can ultimately organization is highly dependent on the quality
hinder their effectiveness. The 2016 Ethiopia Civil of its staff and their ability to fulfil their roles.
Servants Survey (ECSS) generates a picture of Civil servants face a myriad of challenges in their
such potential constraints as perceived by public work. Given their day-to-day experience, the views
servants. As in the previous chapters, the information of staff on which challenges are most important
derived from the survey is also triangulated with can offer some idea and guidance on what may
information and views that emerge from interviews be important to address going forward.
65

5.1 Five highest ranked challenges

194. A key survey question that asked respondents 195. The five most frequently voiced constraints
to identify constraints was “What are the biggest are: (i) inadequate resources (42 percent), (ii)
challenges to you being able to complete your lack of motivation (21 percent), (iii) inadequate
most important tasks effectively?” Figure 5.1 leadership (16 percent), (iv) disruptions by ad hoc
highlights the top five challenges identified by work requests (12 percent), and (v) delays getting
civil servants across the three tiers of government. inputs from others (8 percent). Each of these is
These issues were emphasised throughout the discussed in turn below.
surveys as key bottlenecks to officials effectively
undertaking their daily tasks.47

Figure 5.1: Top 5 most common challenges to completing tasks effectively

Inadequate resources / IT / equipment

Lack of motivation

Inadequate leadership

Last-minute requests

Delays getting required inputs

0 10 20 30 40
Percent

5.1.1 Inadequate resources servants feel that they are challenged to complete
tasks effectively.
196. Inadequate resources, transportation, and
IT present civil servants with a challenge at all 197. Resources are generally less of a constraint
levels, increasing in importance from federal (23 at the federal level, except for the agricultural
percent) to woreda level (48 percent), where it sector. This is somewhat surprising given the
represents a significant challenge in all five sectors. critical role that this sector plays for Ethiopia’s
Without sufficient resources to work with, civil growth strategy. In contrast, health sector staff at

47
Respondents were asked to rank their top three choices from a pre-determined list of options. For the purpose of this
analysis the single key constraint chosen by each respondent was taken, as it felt this was the most interesting and
potentially useful statistic. However, it should be noted that another way of interpreting this data would be to look at
how many times the constraints listed featured in each respondent’s ‘top three’ challenges.
 66

the federal level perceive a resource constraint operating resources have remained very small – 3
much less frequently.48 At the woreda level, percent of GDP by 2011/12, and 3.8 percent by
concerns about resource constraints are very high 2013/14 (most recent year available). A number
across all five sectors. of interlocutors emphasized in the qualitative
interviews that operating and maintenance
198. The importance given to this issue by spending is very limited, and particularly so
civil servants is consistent with some of the at sub-national levels. 49 This was particularly
information from recent expenditure analyses, raised for the agricultural sector, in terms of
and from interviews. Due to the same constraints their ability to reach farmers, which requires
of a low revenue to GDP ratio, low per capita transport expenses and potentially per diems
GDP, and an overriding commitment to dedicate that, as discussed in Section 2.2, have also been
scarce fiscal resources to capital expenditures, inadequate for regular staff.

Particular challenges around IT use and networks

199. IT use is being promoted by many governments 200. However, as the 2016 World Development
around the world in their public sectors. The aim Report (WDR) sets out, deploying IT systems
is to (i) capture data for accountability as well as effectively is not an easy task and many
management purposes, (ii) make processes faster, governments experience various types of
more reliable, and trackable as files are migrated difficulties. Many public sector digital technology
from manual to automatized processing, and projects fail, and even if e-government projects
(iii) potentially facilitate better access in a cost- are successfully implemented, they may actually
effective way in remote areas (e.g. for licensing worsen outcomes as, without proper regulatory
requests), (iv) to facilitate user feedback. After safeguards in place, automation makes it easier
a period of catching up with the private sector, to perpetrate fraud and corrupt practices, erase
governments have become more intensive users records, or avoid capturing them altogether,
of information technology than firms. thereby eroding transparency mechanisms.

Figure 5.2: Success rate of large ICT public sector projects

58% 13%
Partially failed Succeeded

29%
Partially failed

Source WDR 2016 team, based on Standish Group 2014 and World Bank (2015a) Digital Governance Projects Database.
Data at: http://bit.do/WDR2016-FigB3_5_L.

48
As noted in section 1.2, 80 percent of public funding for the health sector is provided by development partners, which
may entail ampler resources at the federal level for non-wage recurrent expenditures.
49
Some sectors receive significant external support at sub-national levels for recurrent costs, notably through the PBS
project (water, health, education, agriculture). However, with a total volume of 400-500USD for 6 years, the additional
resources per year are still quite limited.
67

201. Ethiopia is in the process of pursuing greater 202. As would be expected, in Ethiopia IT-use
IT use in its public sector. The GoE e-government by public servants currently remains clearly
strategy (2005-2015) has facilitated the adoption stratified by level of government, and is very
of mobile technologies and expansion of networks limited at the woreda level. While among federal
and local contents. A number of informational level staff, 79 percent of respondents mention that
and transactional services were developed and all professional staff have access to computers.
became operational. The Ethiopian government At the regional level this drops to 48 percent, and
envisions having a capacity of 103 million mobile at the woreda level computers are available to all
phones in the country in the next five years. There staff in only 3 percent of organizations. Woreda-
are various potential opportunities and challenges level staff also have limited access to the internet:
that are emanating from the expansion of these 57 percent of organizations have no access to the
technologies. The strategy has been revised. internet, and only 5 percent have regular access.

Figure 5.3: Out of every 10 experts, how many have access to a computer? By tier and sector

12
experts, how many have
access to a computer
Out of every ten [10]

(desktop or laptop)

10
8
6
4
2
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

Figure 5.4: Out of every 10 experts, how many have access to a computer? By region50

10
Out of every ten [10] experts, how
many have access to a computer

8
(desktop or laptop)

0
Regional Woreda
Government Tier

Afar Amhara Benishangul Gambella Oromiya SNNPR Somali

Tigray Harar Dire Dawa Addis Ababa

50
As noted in section 1.1, no woreda level results are reported for Harari, Dire Dawa, and Addis Ababa.
 68

Figure 5.5: Out of the five working days,

how many days is the internet access good enough to check email?

6
internet access good?
working days, how
many days is their
Out of the five [5]

0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

203. Network interruptions pose a significant challenge sector, with 62 percent of respondents disagreeing that
across all levels of government. For 59 percent of disruptions are not a challenge. Network disruptions
federal respondents, all network interruptions are a are the principal impediment to acquiring data from
challenge. This is similar at the regional level, and is the MIS. Again, this is an especially important issue
even higher at the woreda level, at 65 percent. Network in the health sector, at 51 percent, given that MIS is
disruptions present a significant challenge in the health generally more widely used in this sector.

Figure 5.6: Please indicate the extent to which you agree with the following:
Network interruptions are not a challenge towards me being able to complete my tasks

Federal
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree disagree agree agree applicable

Regional
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree disagree agree agree applicable

Woreda
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree disagree agree agree applicable
69

204. Moreover, phones also do not work continuously. the civil service cannot use IT effectively to better
When staff cannot use the internet, they rely on phone manage and track processes and provide services
calls to exchange information. However, only 31.6 within the time limits foreseen. Software is a large
percent of organizations have a phone that works for constraint at the woreda level (54 percent disagreed
at least 50 percent of calls, all five days of the week. that there was enough software to complete tasks),
but much less so at the federal level (where 30 percent
205. While the survey does not delve deeper into disagreed) and regional level (40 percent disagreed).
this issue, the concern about network disruptions at Software constraints appear to be a significant
woreda levels imply that in principle, some processes problem in the agriculture sector, with 61 percent of
and tasks require at least some computer and respondents raising them as a challenge. Furthermore,
internet use, rather than being completely manual. maintenance, updating, and management of IT
Considering the implications from this analysis, it equipment was seen as a challenge by a substantial
would be useful to clarify further to what extent IT number of staff. Overall, 47 percent of respondents
applications are expected to be used across the five disagree that maintenance and management of IT
sectors, and how effectively IT systems are de facto equipment was sufficient. The level of disagreement
deployed at local levels in particular. is especially high at the woreda level, at 56 percent,
but is consistent across sectors, with the exception of
206. Software and IT management are important revenue, where disagreement is lower than in other
constraints. Without the right software packages, sectors (35 percent).

Figure 5.7: Please indicate the extent to which you agree with the following: software
and applications are not a challenge towards me being able to complete my tasks

Federal
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree agree agree agree applicable

Regional
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree agree agree agree applicable

Woreda
50
40
Percent

30
20
10
0
Strongly Slightly Neither Slightly Strongly Not
disagree agree agree agree applicable
 70

207. WoredaNet. The ECSS shows that less than of organizations have power for at least four
40 percent of woredas have WoredaNet. In those hours of every working day. That said, for two
woredas where there is access, on average staff sectors (agriculture and revenue), continuous
used it on a weekly (35 percent of woredas), power seems – rather surprisingly – limited at the
monthly (25 percent), and quarterly (10 percent) federal level. Given that staff at the federal level
basis. WoredaNet is most used in the trade sector are meant to work with computers, uninterrupted
and least prevalent in the agriculture sector. power appears particularly important to ensure
Woredas in Benishangul-Gumuz and Gambella effective use of staff time. Regionally, Gambella
have no access at all. and Afar are the two regions with relatively least
full-time access to electricity. As the power supply
208. Access to electricity – another key input for is expected to improve further for many regions,
a modern civil service to function – is relatively access to electricity should be good enough so as
good, given Ethiopia’s income level and still not to stall the deployment and usage of improved
very limited power generation. In all, 78 percent IT systems.

Figure 5.8: During a typical working day, how many hours is there electricity? By tier and sector
from 9am to 5pm). How

8
working day (8 hours

many hours is there

During a typical

6
electricity?

0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

Figure 5.9: During a typical working day, how many hours is there electricity? By region
from 9am to 5pm). How many hours is
During a typical working day (8 hours

8
there electricity?

Regional Woreda
Government Tier

Afar Amhara Benishangul Gambella Oromiya SNNPR Somali

Tigray Harar Dire Dawa Addis Ababa

71

5.1.2 Motivation (22 percent) regions, and less so in Dire Dawa (9

percent) and Gambella (8 percent).
209. The high frequency with which low
motivation is cited as a direct hindrance to 210. Changes in motivation largely decrease
working effectively is a significant concern. over time, while staying high for some. One of
A lack of motivation to undertake their job is the questions in the survey asks respondents
the second biggest challenge according to civil to imagine that when they entered the service
servants at all three tiers and across all sectors, their motivation was 100, and to state what their
with the exception of federal staff in the health motivation is now relative to the initial level. On
sector (for whom it is the joint sixth challenge). average civil servants feel 18 percent less motivated
A lack of motivation was identified as the main now than they did when they first entered the
challenge by 20 percent of federal staff, 14 service, 31 percent feel just as motivated, and
percent of regional staff, and 18 percent of 8 percent feel more motivated than when they
woreda staff. A lack of motivation is most keenly entered the service. Woreda-level staff show the
felt in the education sector (23 percent), and least change in motivation (13 percent) and federal
least in the health sector (13 percent). In terms of staff the most (21 percent). Staff in the agriculture
regions, low motivation is a greater challenge in sector show the largest decrease in motivation
the Harari (24 percent) and Benishangul-Gumuz (21 percent).

Figure 5.10: ‘Imagine when you started your motivation was 100.
What number would you say it is now relative to that?’ By tier of government

40

30
Percent

20

10

0 50 100 150 200

Federal Regional Woreda

 72

211. Section 5.3 below explores a number of pace of work, reduce the time spent at work, or are
motivation issues further. The survey did not less friendly to customers – this may be something
cover questions of how lack of motivation exactly to explore in greater detail in repeat surveys that
affects what staff do – e.g. whether they slow their may be undertaken in future.

Motivation: a cross-country comparison

212. Many civil servant surveys ask questions their job, and their specific satisfaction with their
related to the satisfaction of officials along multiple wage and other benefits. Unfortunately, the precise
margins, typically their overall satisfaction with wording of the questions varied across countries.51

Table 5.1: Level of satisfaction with job in the civil service: a cross-country comparison

(1) (2) (3) (4) (5) (6)

All Ethiopia Ghana Indonesia Nigeria Pakistan

Satisfied with job [satisfied=1] 0.75 0.56 0.53 0.85 0.89 0.71

213. With overall job satisfaction, we can fairly are better than those in the public sector, but
straightforwardly compare amongst surveys overall it is better to have a job than not at
in Ghana, Indonesia, and Pakistan, as these all. Such a story would be consistent with the
all ask about the experience of working in the fact that Nigeria has the highest overall level of
public sector relative to the private sector. satisfaction (with 89 percent of civil servants
Looking at the country averages in Table 5.1, being neutral or positively satisfied with their
we see relatively substantial differences, with 53 job overall).
percent of Ghanaian civil servants being neutral
or positively satisfied with their jobs relative to 214. The question asked in the Nigerian survey
the private sector and 85 percent of Indonesian is closely related to the question asked in the
civil servants. The Nigeria survey, however, asks Ethiopian questionnaire (stated above). Fifty-
about their overall satisfaction with their job. six percent of the Ethiopian civil servants we
Officials may feel that jobs in the private sector surveyed stated that they were satisfied or very

51 In Ethiopia, officials were asked, “To what extent would you say you are satisfied with your experience of the civil service?”
with options ‘Very satisfied’, ‘Satisfied’, ‘Very dissatisfied’ and ‘Dissatisfied’. In Ghana, officials were asked to what extent
they agreed with, “Working in the public sector is generally better than working in the private sector”, “My salary is very
satisfactory” and “My other benefits (pension, health, etc.) are very satisfactory”. In Indonesia and Pakistan, they were
asked, “How do you compare [your organization] as a place to work with private sector firms that are in a similar area
as [your organization]?” and the extent to which they agreed that, “Your pay is fair compared to staff doing similar jobs
in other [ministries]”. In Nigeria, officials were asked directly whether they were satisfied with their current job, current
income and working conditions. In the Philippines, the question was the extent to which they agreed that, “You are
satisfied with the pay you receive for your work”.
73

satisfied with their ‘experience of the civil service’. at the median organization in the Ghanaian data
This is substantially lower than the Nigeria are roughly 40 percent less satisfied with their
average. We can make comparisons to the other jobs than those in the Indonesian data. A similar
countries, but as discussed in footnote 33, this is claim could be made about the comparison
conditional on the comparability of the concepts between Ghana and Nigeria, but this would be
assessed. Assuming they are comparable, conditional on the wording issues outlined above.
Ethiopia’s service has a similar average to Ghana,
but lower than Indonesia or Pakistan. 216. We see that the profile of Ethiopia is
very similar to that of Ghana, with both being
215. We can also look at the distribution of relative outliers in the extent of variation their
satisfaction within and across government organizations exhibit in average satisfaction.
organizations. Figure 5.11 plots, for each The other surveys fluctuate between 60 percent
organization in different samples of civil servants and 100 percent of staff satisfied with their
across these five countries, the proportion of civil jobs overall. However, Ethiopia has one of the
servants in that organization that are neutral highest proportions of organizations in which
or positively satisfied with their job. These all staff are satisfied or very satisfied. A crude
proportions are plotted against the percentile of characterization is that Ethiopia has a relatively
average satisfaction at an organization within the low level of motivation as well as a relatively
country. Thus, Figure 5.11 shows us that officials high degree of inequality across organizations.

Figure 5.11: Proportion of civil servants in that organization

that are neutral or positively satisfied with their job

.8
Proportion satisfied

.6

.4

Ethiopia
.2 Ghana
Indonesia
Nigeria
0
Pakistan

0 .2 .4 .6 .8 1
Ranking of organization
 74

5.1.3 Inadequate leadership a thorough understanding of the services that are

to be delivered matters.
217. Third on the list of main challenges identified
by civil servants is the absence of adequate 220. Political appointees may at times be selected
leadership in the organization, with 15 percent on criteria such as regional or group balances,
of federal, 9 percent of regional and 19 percent or political loyalty. These criteria may come at
of woreda level staff identifying this as the most the cost of having relevant expertise, or having
important obstacle to completing tasks effectively. a track record as a good manager. Moreover,
Inadequate leadership is particularly important seeking a political career can distract attention
in the health sector (24 percent), and within the away from the good management of a ministry
health sector at the woreda level, but much less or agency, which requires focusing on the good
so for the federal or regional level. At the federal use of internal resources and maximizing service
level, inadequate leadership was especially flagged delivery performance.
in the trade and education sectors; and at the
woreda level also for the revenue sector. It is rarely 221. One particular issue which emerges from the
flagged among regional-level staff. ECSS is that leaders within the public service do
not use performance information consistently.
218. In the open-ended section of the survey, 5.6 Thus, 46 percent of heads of departments believe
percent of respondents raised poor leadership as a that bad performance is addressed inconsistently
particular reason for demotivation.52 While this is far or not at all; and a full 63 percent believe that
below the frequency of salaries and poor conditions poor performers either stay in their positions or
of work, it is still something that the government their performance may be addressed on an ad hoc
should pay attention to, as it can weaken particular basis only. Conversely, most heads of agencies
organizations and areas of the public sector. believe that there are no clear systems or criteria
for rewarding well performing staff.
219. The quality of leadership had emerged quite
strongly in the qualitative discussions, with the 5.1.4 Last minute/ad hoc work requests
main question raised about whether consistently
good leaders are selected in the public sector. In 222. Around 30 percent of civil servants’ time
Ethiopia, political appointments start relatively low is used for ad hoc tasks that are not part of
in the hierarchy of the public administration. On their main job. This is similar across tiers and
the one hand, this is an honest approach: in many sectors, (highest in agriculture with 33 percent
countries, permanent secretaries, undersecretaries, and lowest in trade with 27 percent). Again, there
and director generals are classified as ‘civil is significantly more variation across regions with
servants’, but nonetheless de facto are political civil servants in Afar spending 57 percent of their
appointees. Formally designating them as political time on ad-hoc tasks, compared to only 19 percent
appointees acknowledges this more clearly. On of the civil servants’ time in the Dire Dawa, Harari,
the other hand, selecting good managers who have and Somali regions.

52 Inappropriate management, leadership practices, abuse of power – raised by 3.5 percent, and lack of qualified and
competent leadership – raised by 2.1 percent.
75

Figure 5.12: What percentage of your time is used on tasks which are not part of your main job?

15

10
Percent

0
0 20 40 60 80 100

5.1.5 Delays receiving inputs from others as the main constraint, compared to regions such
as Gambella or Harari where this problem was not
223. Delays in receiving required inputs from selected by respondents.
others is another important challenge encountered
prominently at the federal level (15 percent), but less 224. In principle, the BPR has addressed workflows,
so at the woreda level (5 percent). The distribution including a focus on reducing idle time due to
is similar across all sectors (8-9 percent). There is waiting for inputs from others. For those offices
significantly more variation in the distribution across where this is a greater concern, it might be relevant
regions, presenting a significant problem in Addis to revisit whether workflows are still sufficiently
Ababa, where 23 percent of respondents identify this well organized.

Around 30% of civil servants’ time is

used for ad hoc tasks that are not part
of their main job

5.2 Other constraints and challenges

5.2.1 Rent seeking and corruption region. Notably Rwanda appears to control
corruption more effectively relative to Ethiopia
225. Rent seeking and corruption are concerns (see Graph 6.1). On the positive side, control
in Ethiopia even if they are perceived to be of corruption is seen as continuing to improve
less prevalent than in other countries in the in recent years.
 76

Figure 5.13: Comparative progress in control of corruption, Ethiopia and Rwanda

Ethiopia 2005
2009
2011
Control of Corruption

2013
2015

Rwanda 2005
2009
2011
2013
2015

0 20 40 60 80 100

Source Worldwide Governance Indicators 2017.

226. The areas that the government itself, as well 228. The revenue and trade sectors, in particular,
as other sources, has identified as most prone to raised rent seeking as constraints. For the
corruption include land management, revenue Ethiopian Revenue and Customs Authority,
authorities, commercial licensing, procurement, reducing corruption is a major challenge. Taxpayers
and construction/public infrastructure. In these have a significant incentive to offer bribes to tax
areas, civil servants on comparatively low salaries officials so as to reduce assessed tax obligations.
are interacting with significant economic interests, Depending on the size of the taxpayer, the
such as contracts worth the equivalent of millions bribes being offered can be substantial. A Good
of US-Dollars, or tax obligations worth thousands Governance Plan has been produced annually
of US-Dollars. Corruption in service delivery since 2013. For the Ministry of Trade (MoT),
sectors such as health and education seems to corruption is seen as a significant challenge, as the
remain relatively uncommon. MoT and Trade Office have a controlling function,
and this creates incentives to offer bribes to evade
227. Rent seeking was raised as one of the major controls. Areas considered as particularly prone
challenges in the key informant interviews. According to bribery include the speeding up licensing
to several interlocutors, there is a widespread processes, and seeking approval for the import
sense that requests for bribes are becoming more of goods – for example, when traders try to import
common. The areas that the government itself, as goods that do not meet the technical specifications
well as other sources, has identified as most prone and standards set by trade institutions. Certificate
to corruption include land management, revenue of Competency (CoC) is another type of license
authorities, commercial licensing, procurement, that is seen to be prone to corruption. At the same
and construction/public infrastructure. The Change time, non-government stakeholders interviewed
Army reform and the Public Wing are considered as for this argued that licensing requirements can
tools to limit rent seeking, which was explored in the be overly narrow and re-licensing requirements
survey and discussed below. (overly) frequent.
77

229. The survey sheds some further light on this 6.43 percent), ethics training (5.97 percent),
issue, mainly with regards to mechanisms for eliminating red tape (4.92 percent), transparency
containing or combating corruption. Corruption of service requirements (4.60 percent), and Public
was only named by 2.7 percent of civil servants Wings (3.45 percent). Only 1.3 percent thought
as hindering work effectiveness. It was mentioned that investigations by the Federal and Regional
somewhat more frequently at the woreda level and Ethics and Anti-Corruption Commissions were
in the agriculture, health, and education sectors. effective to limit rent seeking.
Corruption also appears to present a greater
challenge in Oromia, but was not mentioned by 231. The Change Army is considered as effective
respondents in Harari, Dire Dawa, or Addis Ababa. in identifying solutions to rent seeking in the
Generally, due to a limited number of targeted civil service by 65 percent of organization
questions, the survey does not shed much light on heads (Figure 5.15). Across all civil servants, the
how widespread civil servants believe corruption majority select the Change Army (specifically
and rent seeking is. the ‘1 to 5’) as one of the most effective tools
at limiting rent seeking, suggesting the reforms
230. The views regarding the effectiveness of perceived potential at dealing with this issue in
different anti-corruption tools show a different the civil service (Figure 5.15). The organization
distribution than might be expected. As tools to in which this is raised in a very limited way is the
limit rent-seeking, civil servants saw the following Federal Ministry of Agriculture, implying that more
as most effective: (i) the Change Army/’1 to 5’ work needs to be done in this tier and sector to
groups (26.6 percent 53), and (ii) complaints strengthen the Change Army reform. The Public
from users (23.8 percent) as the two by far most Wing is much less cited by the civil servants as a
effective mechanisms. These are followed by: form of dealing with rent seeking, though it does
monitoring by managers (14.43 percent), internal appear to be more important in the trade sector
audits (6.53 percent), monitoring by peers at the regional level.

Figure 5.14: How does the Change Army deal with rent seeking? Heads of organization only.

Federal Regional Woreda

100
80
Percent

60
40
20
0
t

en t

d- nd

en t

d- nd

en t

d- nd
no

no

no
pp bu

pp bu

pp bu
we s a

we s a

we s a
s

up

up

up
es

es

es
ha t

ha t

ha t
g no

g no

g no
llo e

llo e

llo e
do

do

do
fo do

fo do

fo do
in s

in s

in s
th oe

th oe

th oe
It

It

It
It

It

It
no It d

no d

no It d
It

How does the daily change army meeting deal with rent-seeking?

Source Graphs by Tier of government

53 Percentages are for those referencing a particular tool as the most important. Frequencies are higher if including all
those mentioning a tool/mechanism as among the top three.
 78

Figure 5.15: Which of these tools do you think are most effective at limiting rent seeking?

.4
.8
Change army (1:5 specifically)

.3
.6

Public wing
.2

.4

.1

.2
0
re

or

ue

re

or

ue

e
ad

ad
tio

tio
tu

tu
ct

ct
n

n
Tr

Tr
ve

ve
ca

ca
Se

Se
ul

ul
Re

Re
ic

ic
u

u
th

th
Ed

Ed
r

r
Ag

Ag
al

al
He

He
Federal Regional Woreda

Which of these tools do you think are most effective at Which of these tools do you think are most effective at
limiting rent-seeking? limiting rent-seeking?
Indicator if selected Indicator if selected

232. Concerning the (limited) role of Public regard than civil servants may perceive.
Wings in addressing rent seeking and corruption,
perceptions of civil servants broadly match those 233. In addition to the limited role of Public Wings,
of the Public Wing members interviewed. Most civil servants also expressed some concerns
Public Wing members feel that corruption and about how other employees and leaders in the
rent-seeking problems are sensitive and are not organization would react if they sought to address
discussed much at Public Wing meetings. For the rent seeking. Answers to the open-ended question
health sector, Public Wing members mentioned “What is stopping you from further reducing rent-
that they had offered to contribute to a better seeking in your organization or directorate?” are
monitoring of corruption in the sector, as they shown in 5.2. This suggests that organizational
are able to see and hear many things. However, leadership needs to signal and demonstrate more
according to Public Wing members’ accounts, this clearly that it is committed to following up on
offer was not very welcome. Thus, if anything, the rent-seeking allegations when they are raised by
Public Wings’ roles are even more limited in this internal whistleblowers.

The Public Wings’ roles are even more

limited than civil servants may perceive
79

Table 5.2: Factors that stop civil servants to fight rent seeking. All civil servants.

No. Response Frequency %

1. Nothing 476 16.02

2. Fear of retaliation from higher bodies/those involved in rent seeking 297 9.99

3. Lack of commitment/negligence/reluctance among employees 203 6.83

4. Strong network among rent seekers and higher government bodies 179 6.02

5. Lack of awareness about rent seeking and its consequences 156 5.25

6. Lack of transparency and accountability 134 4.51

7. Lack of evidence/secret nature of rent seeking 125 4.21

8. Leadership is not committed/leniency in leadership about rent seeking 121 4.07

9. Wrong attitude towards rent seeking 117 3.94

10. Rent seekers are not accountable for their deeds/no legal action taken 99 3.33
against them/no real consequence

11. Lack of strong check and balance system to fight rent seeking 87 2.93

12. I don’t know 55 1.85

13. Favoritism/nepotism 53 1.78

14. The legal system is weak on rent seeking/no rule of law 48 1.62

15. The involvement of political leaders/higher officials in rent seeking/corruption 47 1.58

16. Lack of commitment by officials to take serious measures on rent seeking/ 47 1.58
reluctance to catch the big fish

17. Pressure/interference by political leaders 42 1.41

18. Lack of readiness to take corrective action 41 1.38

19. Lack of conducive environment to fight rent seeking 38 1.28

20. Ethnic affiliation 38 1.28

5.2.2 Effective use of staff time 39 hours a week. When asked about the actual
number of hours they work a week, only 8 percent
234. Hours Worked: According to the Civil Service of staff worked the standard 39 hours: 33 percent
Proclamation, the regular working hours of a of staff work less than the 39 hours and 58 percent
civil servant should be determined based on the work more. The majority of staff (41 percent) work
condition of the work, and should not exceed 40 hours a week during a typical workweek.
 80

235. Staff work an average of 39.5 hours at the hours), whereas those employed in agriculture
federal level, 40 at the regional-level, and 40.5, work the most (43 hours). Where we find the
at the woreda level; this is higher for managers, most variation in the number of hours worked is
who on average work 46.6, 45 and 46 hours at the across regions. Civil servants in Gambella work
federal, regional, and woreda levels respectively. an average of 32 hours a week, compared to 49
Civil servants employed in the trade sector work on and 50 hours among civil servants in Amhara and
average the least number of hours per week (39.5 Tigray respectively.

Managers’ use of employee time

Figure 5.16: Hours worked by civil servants per
week (staff and directors only)

236. Time use and coordination on tasks 25

assigned: Civil servants in Ethiopia have an
average of two superiors who assign them 20
tasks regularly, and spend, on average, 57
percent of their time on tasks given to them Percent 15
by their direct superiors. This does not differ
significantly across tiers, sectors, or regions. 10
The majority of tasks in the civil service are
assigned to staff by either director (47 percent) 5
or team leaders (40 percent) from within the
organization (see Figure 5.17). This is similar 0
across tiers and sectors, with the exception of 0 20 40 60 80
health and trade where deputy heads of the
organization assign tasks in 8 percent and 9 lack of coordination in the allocation of tasks and
percent of cases. Addis Ababa is where there is use of time, leading them to be overworked. Staff
most variation in who allocates tasks to staff, in the Somali region indicate the highest level of
with directors and team leaders from outside coordination, with 47 percent.
the organization accounting for 11 percent of
task allocations.

237. A majority of respondents across all tiers

believe their managers do not actively compete for
their time and do not coordinate, leading staff to
be overworked and overburdened. This perception
of a lack of coordination that leads managers to
assign an excessive workload is highest at the
woreda level (61 percent), followed by the regional
(58 percent) and federal (53 percent) levels.
The level of coordination is similarly low across
sectors, the trade sector showing the highest level
of coordination, with 44 percent of respondents
saying their managers coordinate, and ensure their
time is used effectively and reasonably. The lack of
coordination is especially important in Dire Dawa
where 84 percent of respondents say there is a
81

Figure 5.17: “Which of these groups regularly give you tasks as part of your formal work duties?”

Directors from the organization

Team leaders in the organization
Deputy Head of organization
Head of the organization
Directors from outside of the organization
Others
Team leaders outside of the organization
Clients
Heads or Deputy Heads from outside the organization

0 10 20 30 40 50
Percent

Meetings

238. To measure the perception on the effectiveness levels. Pre-determined agendas appear to be more
of meetings, respondents were asked about the commonly used at the federal levels with 51 percent
proportion of meetings that i) stick to a fixed of staff believing that meetings stick to a fixed
agenda, ii) achieve meeting goals in the minimum agenda more than 75 percent of the time. This is
amount of time, and iii) are of substantive use to significantly higher than at woreda-level, where only
daily tasks. 30 percent of meetings are perceived to stick to an
agenda more than 75 percent of the time (Figure
239. Pre-determined agendas. Federal staff seem 5.18). The distribution is similar across sectors, with
significantly more positive about the effectiveness 30-38 percent of respondents believing meetings
of meetings than staff at the regional and woreda- stick to an agenda in more than 75 percent of cases.

Figure 5.18: Perception of proportion of meetings that stick to a fixed agenda

Federal Regional Woreda

50
40
Percent

30
20
10
0
%

0%

5%

0%

5%

0%

5%

%
25

00

25

00

25

00
-5

-7

-5

-7

-5

-7
0-

-1

0-

-1

0-

-1
%

%
%

%
25

50

25

50

25

50
75

75

75

Source Graphs by Tier of government

 82

240. Goals of meeting are achieved in a 12 percent of respondents believing meetings

minimum amount of time. Perceptions on the achieve their goals in a minimum amount of
proportion of meetings which achieve their time more than 75 percent of the time, and 53
goals in a minimum amount of time are overall percent believing this happens less than 25
lower across the tiers, but especially so at the percent of the time. Conversely, the agriculture
woreda level where 52 percent of respondents sector does best, with 25 percent believing
think meetings achieve this less than 25 percent goals are achieved in a timely manner in 25
of the time (Figure 5.19). The education sector percent of cases, and 44 percent believing it
appears to struggle with this most, with only happens in less than 25 percent.

Figure 5.19: Perception of proportion of meetings that achieve their goals in minimum time

Federal Regional Woreda

50
40
Percent

30
20
10
0
%

0%

5%

%
%

0%

5%

0%
%

5%

%
25

25

25
00

00

00
-5

-7

-5

-7

-5

-7
0-

0-

0-
-1

-1

-1
%

%
%

%
25

50

25

50

25

50
75

75

75
241. Usefulness of meetings. When it comes to of cases (Figure 5.20). Meetings appear less
the usefulness of meetings towards the daily useful in Gambella, where only 4 percent of
tasks of respondents, federal staff are more respondents think meetings are useful more than
positive than regional and woreda-level staff. 75 percent of the time, and where 57 percent
While 37 percent of federal respondents believe think they are useful less than 25 percent of
meetings are useful in more than 75 percent the time. This contrasts with the Tigray region,
of cases, 45 percent of woreda staff believe where 42 percent think meetings are useful in
meetings are useful in less than 25 percent more than 75 percent of cases.

Figure 5.20: Perception of proportion of meetings that are of substantive use to daily tasks

Federal Regional Woreda

50
40
Percent

30
20
10
0
%

0%

5%

%
%

0%

5%

0%
%

5%

%
25

25

25
00

00

00
-5

-7

-5

-7

-5

-7
0-

0-

0-
-1

-1

-1
%

%
%

%
25

50

25

50

25

50
75

75

75
83

242. The survey thus suggests some areas of “To what extent would you say you are satisfied
improvement with regards to time use. Overall, with your experience of the civil service?” The
working hours appear to be broadly reasonable, data shows that around half of civil servants are
but the balance across regions may need some dissatisfied or very dissatisfied at all tiers, but are
attention, to avoid both excessive working hours particularly so at the federal and woreda level
and overly short hours. There appears to be some and less so at the regional level (see Figure 5.21).
scope to further improve management practices Agriculture staff are the least satisfied, with a 49
with regards to time use of staff, coordination when percent satisfaction rate, followed by health staff
assigning tasks, and ensuring an effective use of with 56 percent. Female staff appear to be more
meetings, in particular at regional and woreda levels. satisfied with their experience with the service,
with 67 percent satisfaction compared to 52
5.2.3 Satisfaction and perceived percent among men. The youngest staff (25 and
advantages and disadvantages/ under) in the service are more satisfied than any
satisfaction/experience other age groups (60 percent); another group with
high satisfaction rates are those who have been in
243. As noted earlier in this report, staff the service between 30 and 35 years (74 percent
satisfaction in Ethiopia is quite low, but it is also satisfaction). The following sub-sections further
quite dispersed across organizations. In the survey explore specific aspects of satisfaction.

Figure 5.21: “To what extent would you say you are satisfied with your experience of the civil service?”

Federal Regional Woreda

50
40
Percent

30
20
10
0
d

d
d

d
ie

ie

ie
ie

f ie

fie

ie

f ie

fie

fie

fie

fie
isf

isf

isf
isf

isf
tis

tis

tis

tis

is

t is

tis
at

at

at
at

at

at
Sa

sa

Sa

sa

Sa

sa
ss

ss

ss
iss

iss

iss
ry

ry

ry
Di

Di

Di
yd

yd

yd
Ve

Ve

Ve
r

r
Ve

Ve

Ve

Advantages of working in the civil service

244. Job security. Job security is the principal percent) regions, less so for staff in Amhara and
attraction to the civil service, particularly at the SNNPR (both 33 percent). As noted above, this
regional and woreda levels (42 percent and 41 suggests that civil service positions are particularly
percent respectively). This is the main attraction attractive in regions with limited private sector
for trade staff in particular (46 percent), and for formal employment opportunities.
staff in the Somali (70 percent) and Gambella (64
 84

Figure 5.22: What would you say are the advantages of working in the
civil service relative to the private sector? Job security. By tier and sector

.6
Job security

.4

.2

0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

245. Opportunities for learning. The attraction are the main attraction for 22 percent of federal
offered by opportunities for learning is considered employees, 15 percent regional employees, and 21
the main attraction of the civil service by federal percent woreda employees. Revenue staff are the
staff with 30 percent, compared to 27 percent at least likely to have been attracted into the civil
the regional and 28 percent at the woreda-level. service by career opportunities with 16 percent,
The distribution is similar across sectors and most and health staff the most, with 23 percent. Career
regions, with the exception of Tigray, Somali, and opportunities are a main attraction in the Somali
Dire Dawa, where staff give higher importance region (25 percent) and least in Harari (10 percent).
to the learning opportunities offered by the civil Similarly, career opportunities as an attraction
service (52 percent, 37 percent, and 35 percent to join the service are less important the older
respectively). This is not considered as important the respondent is (ranging from 25 percent to 10
a factor by both the eldest civil servants (between percent). It is also less of an attraction for those
50 and 60 years) and those that have been in the respondents who have been in the civil service
civil service for more than 30 years. longest. The attraction of career opportunities is
more important for those respondents who are
246. Career Opportunities. Career opportunities confident they will be promoted if they perform well.

Figure 5.23: What would you say are the advantages of working in the civil service
relative to the private sector? Opportunities for learning. By tier and sector
learning / development

.6
Opportunities for

.4

.2

0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

85

Figure 5.24: What would you say are the advantages of working in the civil service
relative to the private sector? Career opportunities. By tier and sector

.5
.4
opportunities
Better career

.3
.2
.1
0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

247. Prestige. Prestige is particularly low at the which prestige represents a main attraction (31
federal level (8 percent), and increases somewhat percent), compared to the lower-ranking regions
at sub-national levels, to 11 percent at the regional of Amhara and Benishangul-Gumuz, both with 6
and 14 percent at the woreda levels. Prestige is percent. Typically, the importance given to prestige
most important as an attraction in the health as an attraction to the civil service decreases with
sector with 14 percent. Somali is the region in the number of years served in the civil service.

Figure 5.25: What would you say are the advantages of working in the civil service
relative to the private sector? Prestige/ social status. By tier and sector

.5
Prestige / social status

.4
.3
.2
.1
0
Agriculture Education Health Sector Revenue Trade

Federal Regional Woreda

248. Overall, these preferences as expressed by As noted above, the very low prestige that civil
civil servants suggest more of a ‘security’ and servants perceive particularly at the federal level
less of a growth mind-set. This is reflected in the is a cause of concern in terms of having a high
relatively lower values attached to opportunities caliber, effective civil service that can play a critical
for learning and expected career development. role in social and economic transformation.
 86

Most favored and painful aspects of the civil service

249. In addition to the closed questions concerning are discernible. There are three main perceived
working in the civil service, respondents were asked, benefits to working in the Ethiopian civil service: (i)
“what is the best thing about working in the civil the opportunity to serve the public and/or Ethiopia
service?” and “what pains you most about working (38 percent), (ii) the job security and pension (26
in the civil service?” as open-ended questions percent), and (iii) the opportunity for learning and
without providing pre-determined options. career development (13 percent). This indicates a
significant public service orientation and patriotism
250. The responses can be categorized into broad as a motivation.
themes, under which certain patterns of preference

Figure 5.26: Favored aspects of working in the civil service

45
40
35
30
Percent

25
20
15
10
5
0

Serve the public and Job security Opportunity for learning Gaining valuable
Ethiopia and pension and career development experience

251. In terms of greatest pain, the most frequent percent). The observation that a share of civil servants
responses referred to low salaries, with 41 percent find training opportunities to be lacking, indicates
of respondents indicating this as the most painful that there is an unfulfilled potential, i.e. at least some
aspect. It is followed by a lack of training and learning civil servants see more learning opportunities as even
opportunities (8 percent), and the lack of benefits (7 more critical than improving salaries.
87

6
Options for
the Way Forward
 88

252. Based on the discussion in the preceding 253. The proposed implications draw on the survey
chapters, this section outlines what might be as well as on the qualitative analysis undertaken.
done going forward. The suggestions made As described in Section 1.2, the survey covers staff
are mindful of resource constraints and seek working in ministries, bureaus, and offices down to the
to minimize suggestions that would have a woreda level, but for the agriculture, education, and
substantial fiscal cost. In moving forward, the health sectors, it does not include front-line service
Government of Ethiopia, and the Ministry of delivery staff such as extension workers, teachers,
Public Service specifically and in collaboration or doctors and nurses. The qualitative analysis has
with other ministries, can build on the substantial included discussions on front-line service delivery
reform efforts and progress that have already been issues, such as training needs, needs for additional
made, and which indicate that Ethiopia has a good resources, and others. The team has also reviewed
capacity to pursue reforms once the direction of existing key information on specific service delivery
these has been agreed upon and decided. challenges in the five sectors (see Annex 1).
89

254. The Ministry of Public Service and Human and sharing such lessons with other parts of the
Resource Development, jointly with the sector government.
ministries and the Ministry of Finance and Economic
Cooperation, can play a critical role in deciding and 255. An overall insight, setting this analysis in the
supporting the implementation of further public context of Ethiopia’s overall development, is that it
sector reforms, with the Reform Directorates of has been possible to achieve progress, and to have
sector ministries being a key interface. Most of the a relatively close alignment of policies and actions
recommendations provided below could first be (i.e. a relatively low implementation gap), despite a
piloted, e.g. for a set of organizations such as the relatively high level of dissatisfaction among civil
regional bureaus of a particular sector, if that is servants. However, civil service improvements are
an approach the government would like to take. a continuous challenge as the country develops,
The Ministry of Public Service would also have a and as the complexity of development tasks grows;
critical role to play with regards to monitoring ‘what the suggestions made here are intended to provide
works’ and what improvements are being achieved, a set of options for the way forward.

6.1 Recognize strengths and preserve what works

256. Ethiopia’s public sector stands out for two main are very important strengths, and many other low
strengths: (i) a lot has been achieved with very limited income countries have struggled with these aspects.
resources and (ii) there is a relatively good alignment Thus, in seeking to address challenges such as
between taking decisions and actual implementation, staff motivation and reducing delivery bottlenecks,
so that practices such as targeting and monitoring attention should also be given to maintaining and
are deployed relatively effectively. These two aspects further developing these strengths.

6.2 Options for addressing the challenge of low motivation and low prestige of
the civil service

257. Low motivation and low prestige appear is particularly risky, since the state is meant to play
as important aspects of the findings presented an important role in the economy and in society.
across several chapters of this report. Pursuing a The following are potential options for seeking to
‘developmental state’ model, such low motivation improve motivation:

i Establishing a system based on competitive with the experience of previous developmental

entry exams. Such processes have been used states such as South Korea or Japan.
in Asian developmental states, as well as
in parts of continental Europe. They create ii Allocating greater funds for key operational
a greater sense that only the best and the tasks and systems. The findings presented
brightest can join key positions in the civil in Chapter 5 indicate that having sufficient
service, and hence contribute to greater funds and systems (such as functioning IT
prestige. Specialized competitive exams can be systems) are the most important constraint
used for recruitment into sectors. Selecting the to performance and to effectively delivering
best and the brightest is particularly critical for public services, and are the top level
key economic policy functions, again in line constraint mentioned by staff from all
 90

five sectors. International research on the leadership appears as an important constraint

importance of this (e.g. relative to public especially in the health and revenue sectors
sector pay and non-wage incentives) is where concerns about leadership are voiced
limited; however, the specific constellation more frequently.
in Ethiopia suggests O&M as a key area with
the greatest potential initial pay-off in terms iv Seeking to understand further why
of allocating additional resources. motivation varies so greatly between
organizations within Ethiopia. As addressed
iii Leadership selection and leadership in Chapter 5.1, the dispersion of motivation
training focused on motivating staff. The among staff in different organizational
importance of good leadership for motivating units is very noticeable. This merits further
performance is well recognized. Inadequate exploration.

6.3 Options for further improving management practices

258. Over the past 10 to 15 years, the reforms in sectors and organizational units would appear
pursued have been relatively standardized, and as useful, and contribute towards empowering/
rolled out across all sectors (e.g. BPR, BSC, Change motivating staff – without losing the benefits of
Army). In terms of current management practices, target setting and monitoring. The Change Army
surveyed civil servants rate target setting and appears as a particularly well-regarded reform
monitoring appear higher, and flexibility and staff approach, and it is at least in part a more flexible
involvement lower. Going forward, a more specific and adaptive reform tool.
approach that addresses particular constraints

i Moving from general reform models to contributions to improved service delivery.

identifying and enabling sector and task In addition, this may also contribute to an
specific improvements: As is described further increased perception of staff involvement in
in Annex 1, the specific challenges faced and terms of good management practices (see
the role and functions of staff differ across also Box 6.1 on problem-driven approaches
sectors. Making public service reform efforts to reforms). Within sector ministries, Reform
more specific can entail at least two avenues: Directorates can play a critical role in shaping
(a) The Ministry of Public Service may want public sector changes that support wider
to develop its further reform efforts in close sector reform efforts and coordinating these
collaboration with specific sector ministries, with the Ministry of Public Service.
or clusters of ministries such as health and
education, or trade and revenue; while at ii Consider rebranding the Change Army as
the same time keeping an eye on overall Quality Circles: The case team workers seem
coherence, e.g. in terms of staff grading, pay comparable to “quality circles” in that they
scales, and recruitment standards; (b) within are a formal, institutionalized mechanism for
individual sectors, eliciting ideas from staff participatory problem solving, peer to peer
of potential organizational improvements learning and continuous improvement of
(within their units and with regards to the performance. However, the qualitative interviews
sectors managed) could make valuable and dissemination discussions indicate that
91

the Change Army system is perceived to be as well as greater involvement in policy

politicized and there is too much emphasis on and monitoring functions is expected
attitudes and not enough on knowledge and to contribute to staff satisfaction, and
constructive problem solving. Rebranding the to facilitate attracting, retaining, and
Change Army will encourage team members to motivating well-performing staff. The
cooperate with one another to focus on finding latter aspect is particularly critical, as low
solutions for improving service delivery and to motivation has been identified as a central
steer away from fault finding. driver that limits performance.

iii Facilitating learning across organizational v Better use of staff time should receive
units in the same sector/level of government: attention to improve overall public sector
The Change Army addresses peer learning efficiency. The wage bill essentially pays for
within organizational units. However, given that a set of staff time, skills, and efforts. As the
there are significant differences between units survey indicates, ad hoc tasks, and meetings
within the same sector in terms of motivation that are not sufficiently pertinent and which
and management scores, it would seem useful are not efficient, in particular at woreda levels,
to facilitate learning across such units. reduce public sector efficiency. Investments
in leadership training can include greater
iv Consider enhanced staff involvement guidance on how to use time and how to
in policy formulation and monitoring conduct meetings efficiently and in a way that
functions. Greater autonomy and flexibility, supports organizational performance.

Box 6.1: Problem Driven Iterative Adaptation (PDIA)

as a way to generate specific reform options

The Problem Driven Iterative Adaptation political, and administrative agenda. In

approach suggests that problems provide order to achieve these impacts, the focal
common windows through which agents can problem needs to reflect on a performance
examine their contexts, identify necessary deficiency that cannot be denied or ignored
changes, and explore alternatives to find and that matters to key change agents. The
appropriate solutions. It argues that construction process involves gathering key
problems can be used to drive processes change agents to answer four questions:
of state capability building in practice; “What is the problem?” “Why does it matter?”
given the need to construct problems, “To whom does it matter?” “Who needs to
deconstruct problems, and then promote care more?” and “How do we get them to
problem-driven sequencing. give it more attention?” It is important—in
principle and practice—to think about who
The first step in doing PDIA is constructing answers these questions (and frames the
problems out of conditions, drawing problem). It provokes reflection, mobilizes
attention to the need for change and attention, and promotes targeted and
bringing such change onto the social, context-sensitive engagement.
 92

space for change or readiness. The change

The second step is deconstructing problems space is contingent on contextual factors
into manageable problems. The problems commonly found to influence policy and
identified in the first step are usually difficult reform success, shaping what and how much
to answer because the problems are complex one can do in any policy or reform initiative
and the “right” solutions are hard to identify. at any time. Authority, Acceptance, Ability
Therefore, one needs break the problems are found to be critical in influencing space
down into smaller components that are for change /readiness.
more open to localized solution building.
This involves deconstructing the problem to • “Authority” refers to the support needed to
reveal its causes (where one can ask what is effect reform or policy change or build state
going wrong and why), and look for workable capability (political, legal, organizational,
solutions to these problems. Deconstructing and personal). Some change needs more
problems in this manner also helps one authority than other change, and it is always
identify multiple points at which to pursue important to assess the extent of authority
short- and medium-term successes (or quick one already has—and the authority gaps
wins), which are vital when dealing with a big that need to be closed.
problem that will likely only be solved in the
long run (and which is therefore not likely to • “Acceptance” relates to the extent to
attract the needed short- and medium-term which those who will be affected by
political support). reform or policy change accept the
need for change and the implications of
Deconstruction provides the basis for the last change. Different types of change require
step, which is problem-driven sequencing in different levels of acceptance (from
the change process, where sequencing refers narrow or broad groups and at different
to the timing and staging of interventions and depths) and the key is to recognize what
engagement. It begins with recognizing that acceptance exists and what gaps need to
most deconstructed problems take the form be closed to foster change.
of meta-problems (with many dimensions
and indeed many problems making up the • “Ability” focuses on the practical side of
larger problem). Solving these problems reform or policy change, and the need
requires multiple interventions, which allows for time, money, skills and the like to
multiple entry points for change. Each cause even start any kind of intervention. It is
and sub-cause is essentially a separate— important to ask what abilities exist and
albeit connected—point of engagement, what gaps need to be closed.
and each causal dimension offers different

Source Andrews, Pritchett, and Woolcock (2017).

93

6.4 Options for incentivizing performance and professional growth

259. Findings discussed in section 5.1 and 5.2 even opportunities for training and learning are
(perceived advantages/disadvantages of the limited. Also, while a performance system is in
public sector), and section 4.1 on the treatment place, it does not offer much in terms of rewards
of staff suggest that there is a perception of for good performance, and poor performance is
limited opportunities for growth. Staff believe not dealt with systematically.
that opportunities for promotions are slim, and

i A first consideration should be to create good performers. While regular performance

more training and learning opportunities for assessments have been put in place through
staff. Staff should be more widely encouraged the BSC reforms, survey responses suggest
to scale up their skills, knowledge, and that the follow-up on assessments remains
competencies. This can take various forms, limited (see section 5.1.3). A lack of follow
e.g. class-room based learning, but also up may make staff reluctant to make strong
temporary assignments to different tasks, as investments in better performance.
well as on the job mentoring. As the report has
noted, staff seem to be reluctant to admit to iii Consider to somewhat expand promotion
skills gaps, despite these being widely noted opportunities. As reflected in Chapter 4 of
by key informants, as well as by Public Wing this report, while staff believe that promotions
representatives. An approach that combines are largely merit-based, there is also a
stronger assessments of actual skills, perception that promotion opportunities are
competitive entry, and greater opportunities limited. A lack of promotion opportunities is
for learning can contribute to an upgrading of also cited as the most important reason for
competencies of staff, as well as contributing turnover. While we have to be careful to make
to job satisfaction and motivation. causal assumptions, it seems that limited
opportunities for promotion also contribute
ii It is important that staff action is taken with to declining motivation among staff. Some
regards to staff who are identified as poor broadening of promotion opportunities should
performers, and to provide recognition to be considered in pay and grading exercises.

6.5 Ensuring that tools and systems work as intended

i Pay attention to deployment and usability of and so on, to ensure continuous usability.
ICT systems. Staff clearly struggle with systems
being available and useable; and the latter ii Simplify reform tools. Interventions that are
includes the ability to reliably make phone simple are more likely to be implemented
calls. Investing in functioning ICT systems sustainably. The BSC is a good tool in principle,
may have the greatest pay-off for revenue but remains difficult to use effectively, and
collection, followed by agriculture as a key in practice, the process of target setting,
productive sector. The required investments monitoring and performance assessments
are not just in hardware and software, but remain challenging. There appears to be some
also in the availability of requisite staff to scope for simplification on the one hand, and
provide trouble-shooting, software updates, for automatization on the other.
 94

6.6 Demand-side accountability for reforms and further service

delivery improvements

260. Citizens’ feedback mechanisms that work vary. The understanding of the role of Public
are important for strengthening the service Wings is not complete and the Public Wings are
delivery orientation of the public sector. As has not consistently well organized for an effective
been discussed in Chapter 3 of this report, the discussion within each sector. At the same time,
Citizens Charters, the tool that was expected to civil servants themselves perceive that they are
provide citizens’ power to know and ask their paying substantial attention to feedback from
rights is considered to be a relatively unimportant citizens and clients.
and less successful reform tool by implementers.
Similarly, the key informant interviews suggest 231. The following avenues could be pursued to
that the level of activity, seriousness of the further strengthen accountability and effective
engagement, and usefulness of Public Wings feedback mechanisms:

i Surveys of civil servants such as those carried to monitor patterns of complaints and how
out for this report should be complemented complaints are addressed.
by surveys of service users. This would
be extremely valuable for understanding iii Opportunities for more effective public-private
whether there are matches or mismatches in partnerships for effective service delivery.
perceptions. Findings could then be fed back There is a need for increased formalization
to public sector managers to take action and of the Public Wing consultation mechanisms,
monitor future improvements. Such surveys also in terms of agreed actions and follow up
could also serve as an incentive for identifying on such actions. It may be relevant to consider
and rewarding particularly well performing arrangements that enable more issue specific
teams. Such efforts could initially be piloted discussions, either through the formation
in selected urban areas to limit costs. of sub-committees (e.g. on cancer care in
the health sector), or through the advance
ii Continuing efforts should be made to develop announcement of meeting agendas for Public
Citizens Charters and effective complaints Wing meetings. Moreover, there are areas in
mechanisms. Citizens Charters appear to which the government can consider to ‘do less’
be less effective than desirable, while some or crowd in private sector contributions more,
progress appears to have been achieved e.g. through involving the private sector in the
already on which further efforts can build. continuing training of specialists.
Charters should be clear and realistic, e.g. in
terms of processing times. The government iv Improved communication: The government
can also consider making Citizens Charters may consider to communicate more
legally binding, as has been done in India explicitly and widely about what public
through the adoption of a ‘Rights to Public sector reforms are being pursued, why, how,
Services Act’.54 It also appears as important and with what expected results. This will

54 This would require ensuring that resources are available to service delivery units to stay reliably within targets set in
Citizens Charters, or else such an approach could trigger substantial litigation.
95

help stakeholders to be then attentive to and then follow up by staff and managers
what is actually being done and to provide as appropriate, as well as Public Wing
feedback and proposals accordingly. meetings. It may be helpful to broaden the
feedback being solicited from ‘complaints’
v Broaden opportunities for feedback: The to ‘suggestions, ideas, and complaints’. This
main feedback channel from citizens/ way, service users would feel encouraged to
service users to specific organizational units also contribute ideas of what could be done
currently are complaints mechanisms, e.g. better, and to provide additional channels for
in the form of boxes to collect complaints, ‘bottom-up’ communication.
 96

References

Andrews, M. 2013. The Limits of Institutional Reform in Development: Changing Rules for Realistic Solutions.
Cambridge: Harvard University Press.

Assefa, B. 2009. “Business Process Re-engineering in Ethiopia.” Note for xxx.

Debela, T. n.d. “Civil Service in Ethiopia: The Marriage between Profession and Politics.” UNPAN.

DFID (U.K. Department for International Development), Irish Aid, and Swedish International Development
Agency. 2013. Summary Report of the Public Sector Governance Reform Evaluation. https://www.gov.uk/
government/publications/summary-report-of-the-public-sector-governance-reform-evaluation.

Evans, P. 1995. Embedded Autonomy: States and Industrial Transformation. Princeton: Princeton University Press.

Fukuyama, F. 2011. The Origins of Political Order. New York: Farrar, Straus, and Giroux.

Jurisch, M., C. Ikas, W. Palka, R. Wolf, and H. Krcmar. 2012. “A Review of Success Factors and Challenges of
Public Sector BPR Implementations.” 2012 Hawaii International Conference on Systems Sciences.

———. 2006. “New Public Management as a Template for Reforms in Low-Income Countries: Issues and Lessons
from Ghana.” International Journal of Organization Theory and Behavior 9 (3): 378–407.
97

Lindelow, M., and P. Serneels. 2006. “The Performance of Health Workers in Ethiopia: Results from Qualitative
Research.” Social Science & Medicine 62 (2006): 2225–2235.

Ministry of Capacity Building. 2004. “Capacity/Akem” Quarterly Bilingual Magazine. 1 (1), 26-33.

Ministry of Capacity Building and AH Consulting. 2010. “Assessment of the Implementation of the Civil Service
Reform Program in Ethiopia.”

Norad. 2012. Hunting for Per Diems. The Uses and Abuses of Travel Compensation in Three Developing Countries.
Norad Evaluation Department Report 02/2012. Tina Søreide, Arne Tostensen and Ingvild Aagedal Skage.

Pollit, C., and G. Bouckaert. 2000/2011. Public Management Reform – A Comparative Analysis. New Public
Management, Governance, and the Neo-Weberian State. Oxford: Oxford University Press.

Therkildsen, O. 2006. “Elusive Public Sector Reforms in East and Southern Africa.” In Public Sector Reform
in Developing Countries – Capacity Challenges to Improve Services, edited by Bangura and Larbi. London:
Palgrave Macmillan.

Versailles, B. 2012. Rwanda Performance Contracts. London: ODI/Budget Strengthening Initiative.

World Bank. 1993. “IDA Credit to Support Structural Adjustment in the Ethiopian Economy.”

———. 2008. Public Sector Reforms – What Works and Why? Washington, DC: IEG Evaluation Report.

———. 2013. “Implementation Completion and Results Report – Public Sector Capacity Building Support Project.”

———. 2016. “Public Expenditure Review.” http://documents.worldbank.org/curated/en/176471468178145744/

pdf/ACS14541-WP-OUO-9-Ethiopia-PER-final-May-12.pdf.

———. 2017. “Ethiopia’s Pension System: Background, Current Status and Prospects for Future Expansion.”
Revised Draft Report.
 98

Annex 1
Specific Issues in Focus Sectors

1. This section reviews the particular challenges education, health, revenue administration, and
in the way that civil service reforms have taken trade. This section briefly highlights the particular
place in the five sectors selected as a focus for issues that were raised by representatives from
this analysis. The selected sectors are agriculture, the five sectors.

A1.1 Agriculture in agricultural productivity (Figure A1.1).

Available data suggests that taken together,
Key sector dynamics important increases in productivity have been
achieved, while further progress remains highly
2. Agriculture is one of the critical economic important for food security as well as to enable
sectors for Ethiopia, and the government has further growth in per capita incomes, especially
made significant efforts at supporting increases in rural areas.55

Figure A.1.1: Increase in cereal yields, Ethiopia compared to East Africa, 1992-2014

25k

20k
Hg/Ha

15k

10k

5k
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20

M = Million, K = Thousand Eastern Africa Cereals, Total Yield Ethiopia Cereals, Total Yield

Source FAO statistics (2016).

55 This will also require a further shift in the allocation of labor from agriculture to industry and services as highlighted
by the recent SCD. These issues are not explored further here.
99

3. Two key challenges for effective government of the Federal Minister and State Minister. As for
action on agriculture have been how to ensure other sectors, the idea is to form ‘1 to 5’ clusters of
strong competency both at the Ministry of ‘change teams’. In principle, model farmers should
Agriculture and in its outreach, via agricultural have greater credibility than development agents,
extension services. With regards to the former, given that they are selected based on their greater
the Ministry of Agriculture was one of the first productivity and successful use of inputs.
federal ministries that underwent restructuring
as part of the BPR reform process in 2006-2008. 6. Within the Federal Ministry, ‘Change Army’
As part of this reform process, the number of efforts are cascaded from the level of the
staff in the ministry was reduced from about 270 Minister, through the mid-level management, to
to around 70. This very significant cut in staff specific teams. At each of these levels, weekly or
numbers was associated with a loss of capacity, bi-weekly meetings are held to discuss potential
as well as the creation of some wider resentment best practices and improvements. Consistent with
and concern about civil service reforms, and the the emphasis on ‘internal’ as well as the ‘Public
BPR in particular. Wing’, issues that are discussed concern both
internal processes as well as consultations with
4. In order to regain some of the capacity that was external stakeholders. The latter also includes a
lost as part of the reform process, the Agricultural review of complaints received. ‘Good governance’
Transformation Agency (ATA) was established in issues are a cross-cutting concern of these
late 2010.55 As an agency headed by a CEO, the discussions; for example, to ensure that scarce
ATA is an institution sitting ‘outside’ the regular resources such as per diems are appropriately
public service structure, similar to a number of other allocated, or that discussions on procurement
agencies that have been established in recent years. processes take place. Land issues are also a major
concern, and a frequent point of discussion with
5. The main current reform efforts in the external stakeholders.
agricultural sector include a focus on the
Change Army, as well as the application of the 7. Staff reviews based on the BSC are held twice
Balanced Scorecard (BSC) for setting targets and a year. In principle, efforts are being made to
assessing performance. At the grassroots level, identify good performers, and provide at least
the Change Army effort that has been initiated some rewards for such performance (e.g. monetary
in recent years is centered on ‘model farmers’ rewards, opportunities for further study). As in
that are identified for using good practices. It is other ministries, performance is assessed in terms
also applied to upstream levels of government of achievement of results (60 percent) as well as
agencies involved in the sector, up to the level behavior (40 percent).

Discussions with public wing representatives at the federal level

8. Representatives of organizations that have hand, a number of representatives stressed that

been asked to join Public Wing meetings for the their specific interactions with specific directorates,
agricultural sector have mixed views: on the one agencies, and institutes under the Ministry are

56 See also: http://www.ata.gov.et/about-ata/origin-history-2/.

 100

very useful and productive; on the other hand, the limits their overall effectiveness in policy
general public wing discussions as such are seen development and support to the sector. Public
as useful, but also rather broad, as well as ad hoc Wing meetings as such remain ad hoc – the idea to
in terms of the issues being discussed. create a permanent secretariat was discussed, but
has not been pursued further. Public Wing members
9. Public wing representatives find that are asked to provide evaluation feedback on the
professional staff in the Ministry of Agriculture directorates of the ministry; but this is not seen as
and other related agencies are not consistently very deep or effective, since representatives cannot
and sufficiently knowledgeable, and that this fully assess the performance of many directorates.

Staff survey results for agriculture

10. On average, management scores across the 11. Staff in the agricultural sector particularly rarely
five categories set out in section 3.6 of this report cite learning or professional development as an
are similar to those of all sectors; however, the advantage of working in the public compared to
spread among different organizational units in the private sector. Inadequate resources are most
the agricultural sector is wide, and many units commonly cited as a constraint by agriculture staff
are below the mid-point. Nearly 60 percent of compared to the other sectors (by 46 percent), while
all organizational units are rated below the mid- other factors are relatively much less important
point. The wide spread of perceived management (lack of motivation and inadequate leadership
practices may need further attention, in particular are mentioned by 12 and 13 percent of staff). The
to understand further where and why low scored higher share of citing resources as a key constraint
units are seen negatively by staff, and whether is especially driven by federal level staff expressing
this matches with service delivery performance. such a view far more frequently than in other sectors.

Information available on front-line staff delivery challenges from other sources

12. With regards to front-line service delivery, development agents is very limited. This reduces
the number of agricultural ‘Development Agents’ their credibility vis-à-vis farmers, and their ability
was rapidly expanded, but with many quality to provide advice and assistance that is effective
and associated effectiveness challenges. Most in increasing productivity (e.g. practical knowledge
development agents recruited into the public about the use of improved seeds, application of
service have received two years of training at pesticides and fertilizers, etc.).
Agricultural Technical and Vocational Training
(ATVET) colleges. Similar to tertiary training 13. Development agents themselves have a more
institutes for other professions, the number of positive perception of their knowledge and the
such colleges was rapidly expanded over the quality of training they’ve received. According
past two decades, from an initial nucleus of a to the 2016 Development Agent Tracking Survey,
few established Colleges of Agriculture. This development agents assess their own knowledge
led to a deterioration of the training content. as sufficient, but raise the issue of insufficient
Moreover, practical training for prospective in-service training and lack of opportunities for
101

further training.57 A relatively low share (21 percent) management. There is less satisfaction with
agreed that they had methodological knowledge regards to animal husbandry, and information
and skills gaps, and only 3.6 percent assessed that provision regarding markets and water (see
they had gaps in technical knowledge. Over 90 Table A1.1). Thus, development agents are seen
percent of surveyed development agents thought as helpful in some areas, but less so in others.
that the theoretical and practical training at The survey also suggests that the level of general
ATVETs was adequate. effort by development agents is substantial, with
many visiting assigned households/farmers at
14. The same Tracking Survey shows that least once a month, for a typical assignment
households have some satisfaction with the under 600 households, despite most development
services provided by development agents, in agents not having access to motorized transport.58
particular with regards to advice and services The Tracking Survey does not report in detail the
related to crop farming and natural resource number of households assigned to each DA.

Table A1.1: Percentage of farmers’ qualitative rating on the extent of DA-FES helping farmers to
address technical problems

Success/result (%)

S. Package Adequate to very Average Below

N adequate average

1 Crop farming 48.3 29.0 22.7

2 Animal husbandry 32.4 29.9 37.7

3 Natural resources management 49.2 23.3 27.5

4 Land tenure security and certificate provision 44.2 14.9 40.9

5 Market and water information provision 23.5 21.1 55.4

6 Establishment and strengthening of self-help group 37.4 26.7 35.9

and cooperatives

7 Others(WASH) related advice, good governance and other 47.3 23.0 29.7
community interests

Source DA Performance Tracking Study, January 2015, MoA.

57 A ‘Tracking Survey’ of Development Agent Services was prepared on behalf of the Ministry of Agriculture in 2015 (published
in 2016). It covers six regions: Tigray, Gambella and Afar, Amhara, Oromia, and SNNP.
58 Over 67% of the interviewed farmers reported 1-4 meetings a month. The outreach figures seem to be on the high side,
as for an assigned number of 500, a development agent would need to visit 25 households per day on every working
day. Development agents do complain that they are not regularly receiving free weekends or leave.
 102

15. In addition to very significant limits in majority of development agents would prefer not
the quality of training, concerns have been to be involved in such activities.
voiced that development agents typically have
very limited funding for operating costs (see A1.2 Education sector
also Negera 2014; Gebru 2012). One crucial
limitation that this creates is that development Key sector dynamics
agents can rarely travel to visit farmers further
away from kebele centers. In surveys undertaken 18. For the education sector, the sheer scale of
for groups of development agents in specific the recent rapid expansion of service delivery
regions, motivation is seen as quite low and poses major challenges. As in other low-income
turnover as substantial. countries, the Ethiopian education sector is
grappling with the challenges of providing
16. Turnover among development agents in the universal free access to primary and secondary
six regions covered by the Tracking Survey is education, while at the same time ensuring at
quite high at 38 percent, and thus notably higher least a basic quality. The numbers of schools and
than for the civil service on average (10 percent). enrollment levels have rapidly increased over the
Based on the survey, development agents voice a past two decades;59 however, quality issues have
number of key concerns: (i) heavy workload with been noted with regard to infrastructure, as well
limited resources; (ii) insufficient pay and benefits as curricula and the capabilities of teachers, and
relative to health extension workers and teachers; ultimately learning outcomes. Recent ambitions to
(iii) few opportunities for further education universally have three languages of instruction –
and career progression; (iv) being burdened by the local language, English, and Amharic – further
additional responsibilities that are not part of the increase the challenges. Learning assessments
core job description. that have been carried out since 2000 for grade
4 show a slight worsening in terms of performance
17. The additional responsibilities given to over time.60 Apart from increasing to a level of
development agents include ‘facilitating tax 100 percent enrollment at the primary level,
collection’, organizing and mobilizing farmers for key government goals for the sector under GTP
communal development activities, and organizing II include having a primary school within a 2.5
and mobilizing farmers for administrative and kilometer radius, and a secondary school within
political purposes and activities. Sixty percent of a 10 kilometer radius. Furthermore, the provision
surveyed development agents reported that these of free school meals is intended to be expanded
‘auxiliary’ activities take up a substantial share as a way to improve retention rates.
of their working time (up to 50 percent); and a

Civil service issues and reforms in the sector

19. Regional and woreda education bureaus of the current 36 Teacher Training Colleges (TTCs).
recruit teachers. Most teachers are trained at one While teachers’ basic knowledge is generally seen

59 Net Enrollment Ratio (NER) for primary education (grades 1-8) expanded from 24.9 percent in 1996/97 to 87.9 percent in
2009/10. The government aims at a NER of 100%. The student-teacher ratio is still very high; a Service Delivery Indicator
survey carried out in 2014 measured an average ratio of 54.5, with a higher ratio (69:1) in urban compared to rural areas
(46:1). http://www.majersite.org/issue16/16_3.pdf./
60 See PAD 2013.
103

as sufficient, with some gaps related to English,

pedagogical skills are very limited. More broadly, 22. In contrast, the implementation of the BSC
teaching is not seen as an aspirational career approach is seen as suitable for the sector and
choice, and consequently, TTCs have been unable as contributing to the pursuit of better results.
to enforce a strong enough selection of candidates Targets are being set for the three state ministers
to be trained. responsible for different parts of the education
sector, and are then further broken down for
20. The Federal Ministry of Education has sought particular directors and teams. The education
to utilize civil service reform tools as well as other sector has been the first sector to automate its
means to improve performance in the sector. The BSC processes. Targets and the measurement of
main reform tools that are currently being applied results achieved are based on the data collected
are the BSC and the ‘Change Army’. In addition, the in the EMIS, and also include information on
education sector has invested in establishing an required resources.61 Individual assessments
Education Management Information System (EMIS) are carried out (with a 60/40 weighting for the
that captures service delivery data and is intended to achievement of results and behavior/attitudes),
make it easier for the staff in education bureaus to using the automated system. The average rating
assess progress with regards to delivering education is 88 percent, seen as a ‘B’ rating (‘A’ ratings are
services. However, as discussed in Chapter 4 of this 94 percent and above). Out of over 500 staff in
report, education sector staff, especially at regional the Federal Ministry of Education, 24 employees
and federal levels, widely criticize that the system were identified as high performers last year. The
does not contain sufficiently useful data (about ministry seeks to provide some performance
31.7 percent at the regional and 55.6 percent at incentives to such staff, including paid vacations
the federal level). and visits abroad.

21. BPR-focused reforms were implemented in 23. A ‘Change Army’ approach has been used in
the sector in the 2000s, from the federal level the education sector since 2011-12. The tool is
to the level of district education offices. Overall, highly valued by the leadership of the ministry,
BPR reforms are currently seen as not having been with the Minister calling it ‘the best reform tool’
a ‘good fit’ for the sector, and having resulted in given its more problem-solving approach. At
some capacity losses that had to be subsequently the same time, acceptance of this approach in
recouped. The emphasis on streamlining processes different parts of the sector has been mixed, with
in general has greater relevance for institutions strong uptake in some parts and resistance in
that are involved in particular approvals, such others. For example, ‘1 to 5’ teams are being used
as issuing licenses, than for MDAs involved in by teachers to help improve teaching methods,
service provision. One of the departments that was and have been proposed for students. There has
eliminated as part of the BPR effort was that for been some resistance, e.g. some potential ‘model
teachers’ licensing and re-licensing. This has been students’ have complained that they need to spend
re-created more recently, given that ensuring good time completing their own assignments rather
enough quality of teachers poses a key challenge than dedicating most of their study time to helping
in the sector. others; and similar concerns have been raised by

61
BSC related performance targets and assessments are only done for policy and administrative staff, not for front-line
staff, i.e. teachers.
 104

some faculty staff at tertiary level institutions. out of in-service testing, licensing, and re-
Considerations are under way to re-brand the licensing. As a start, 20,000 teachers have taken
approach as ‘cooperative learning’ to move tests on a voluntary basis. The intention for GTP
away from the military campaign connotations II is to rollout testing to all 500k + public sector
associated with an ‘army’. teaching staff, and to devise an approach that not
only involves paper-based tests but also includes
24. In parallel, the Education Sector Management observation of teachers’ performances in the
Information System (EMIS) was introduced classroom. Failure to pass these tests is expected
at the woreda level from 2009-2010 onwards. to result in mandatory additional training. This
Prior to the introduction of this system, it was would initially be offered free of charge by the
difficult to generate aggregate data on education government through summer training programs,
sector trends in a given woreda from various but might subsequently also require teachers
manually maintained files. Some recent research to pay.
suggests that the EMIS program led to significant
improvements in education service delivery 27. On the positive side, while management scores
through increased information availability for are above those of other sectors, motivation is
decision makers. However, as discussed above, a frequently cited as a challenge. Management
substantial share of survey respondents were still scores in the sector are somewhat above those
quite skeptical about the quality of data contained for all sectors on average, while the ordering of
in the EMIS, indicating that such a reform is a best to less well performing aspects is the same.
continuing effort. Low motivation appears to be a significant issue
in the sector among the types of staff surveyed.
25. As other spending ministries, the Federal A high share of respondents see promotions and
Ministry of Education is keen to dedicate treatment as merit based.
additional resources so as to be able to improve
service delivery. Specific changes that are currently 28. Overall, the education sector reflects the
being sought include an increase in the starting dynamism, ambition, and challenges affecting the
salary of teachers, and introducing housing options civil service in Ethiopia – the latter including the
and transport packages for teachers. For students, significant needs for additional resources. Given
the intention is to introduce free school meals to the scale and the planned further expansion of the
increase retention rates. This fiscal year, school education sector, investing in higher salaries and
meals have already been introduced for children working conditions for teachers and rolling out a
in drought-affected areas, with a significant effect stronger supervision system will involve significant
on dropout rates in those areas. costs. As a basic service, the government has been
able to access grants or concessional financing to
26. In parallel, the Ministry of Education is some extent for the sector.
seeking to tackle quality constraints the rolling

The EMIS program has lead to

significant improvements in
education service delivery
105

A1.3 Health sector tertiary institutions providing training for health

care workers has rapidly expanded, as has the
Key sector dynamics number of students graduating annually; but
as for the other sectors, this fast expansion has
29. Similar to the education sector, the health been associated with a dilution in quality.63 As
sector has seen a rapid expansion in recent the Health Sector Transformation Strategy notes:
years,62 but quality continues to pose major “There are concerns about validity and consistent
challenges. Thus, Ethiopia made good progress implementation of certificate of competence
towards the MDG targets of reducing the mortality assessment for low and mid-level healthcare
of infants and of children under 5 years of age. providers. There is a need to put structures
However, the MDG on reducing maternal mortality and processes in place to review ethics and
was not met, even though an initial significant competence of health workers at facility, district,
reduction was achieved – from 1,400 maternal and regional levels.” (MOH 2015: 47).
deaths per 100,000 live births in 1990 to 420
deaths per 100,000 births by 2013. For all health 31. Assessments of clinical knowledge conducted
indicators, there are still significant differences in recent years show a mixed picture: 71 percent
between more and less developed regions of of all health care providers were able to correctly
Ethiopia (Ethiopia MDG report 2015). diagnose key medical conditions such as malaria,
TBC, and post-partum hemorrhage (SPA 2014:
30. As for other fields, the training that 242). Adherence to medical guidelines for
medical students receive tends to be focused managing the conditions was lower, ranging from
on theoretical learning, and does not involve around 30 to 50 percent for different diseases and
sufficient practical training. The number of aspects of guidelines (EPHI et al. 2014).

Civil service issues and reforms in the sector

32. BPR reforms were seen as a more problematic 33. The introduction of the BSC reform evolved
reform. The application of BPR in the sector over almost a decade. BSC as a management
mainly focused on analyzing institutional tool was introduced into the sector starting
structures and human resources in health. Survey in 2006, with considerable interest from the
results indicate that compared to other sectors, Federal Ministry (BSCI 2013). However, interest
the view that BPR reforms achieved its objective flagged somewhat after a change in leadership
of reduced processing times is less common in at the Federal Ministry of Health. One challenge
the health sector. was an initially very high number of strategic

62
Between 2005 and 2013, the number of small health posts or clinics nearly quadrupled from 4,211 to 14,416, the number
of health centers increased from 519 to 3,245, and the number of public hospitals grew from 79 to 127. According to
Ministry of Health data, primary health service coverage reached 93.4 % of the population in 2012/13 and 94.0 % in 2013/14.
63
The number of public higher educational institutions have increased from eight to 57. Of these, 34 are universities and
hospital-based colleges offering degree programs while 23 are regional health science colleges offering technical and
vocational qualifications (level 1 to 5). Private health science colleges have also flourished, with 24 institutions offering
accredited programs as of 2012/2013. […] Graduation output from higher educational institutions has increased close
to 16-fold from 1,041 in 1999/2000 to 16,017 by 2012/2013. (MOE, 2014; MoH 2013; Jhpiego, 2014). http://www.moh.gov.
et/documents/26765/0/Health+Sector+Transformation+Plan/5542a23a-9bc7-46a2-8c1f-8b32c2603208?version=1.0.
 106

objectives (over 80). A simplified approach was out, and training and awareness as having been
developed in 2009/2010, with support from insufficient. While BSC reporting and use of
the Balanced Scorecard Institute, a US-based BSC targets appears now fairly well established
management consultancy. This was followed in especially at federal and regional levels, the use
2011-13 by a process of ‘cascading’ the overall for rewarding good performance or taking action
goals, derived from the MDGs to specific targets on poor performance is still limited in this as in
for health teams. However, challenges remain other sectors. The Change Army reform is assessed
with actually monitoring progress, as well as positively by a majority of respondents, in terms
with providing incentives for delivering good or of improved workplace attitudes, spread of best
improving performance. practices, and improved efficiency and service
delivery; views on a number of these dimensions
34. Views on the BSC are somewhat mixed, while are more strongly positive among regional
they are more positive on the Change Army. than among woreda level health sector staff (in
For the BSC, a substantial share of respondents particular with regards to workplace attitudes and
(over 50 percent) see it as having been drawn improved efficiency).

Information available on front-line staff delivery challenges from existing sources

35. Absenteeism appears to present a significant 36. Recent information does not indicate a high
problem, in part associated with incentives to level of dissatisfaction with health care services,
practice privately on the side. Given the balance and suggests some possible improvements over
of opportunities, medical professionals either time, but available information remains limited.
leave the civil service or start taking up private Interviews were conducted with those using health
practice on the side. According to a facility survey, care services for the SPA (a sample of 1,908
the average absenteeism rate from a facility is 33 individual health care users countrywide). As a
percent (see also Lewis 2006).64 The 2014 SPA+ countrywide average, 10 percent of caretakers
found a similar absenteeism rate of 33 percent, of children complained about long waiting
of which facility management did not authorize times to see a provider (with a range of 2 to 24
9 percent (EPHI et al. 2014).65 Absenteeism rates percent across regions). On average, 9 percent
were highest in health posts and for health complained about the absence of medicines at
extension workers (40 percent), for whom 17 a facility (ranging from 3 to 19 percent), and 8
percent of absences were unapproved. As the percent about insufficient explanations about the
Health Sector Transformation Plan (2015) notes, illness, ranging from 1 to 25 percent. Given that
there is an urgent need to generally improve health care services are still very basic, the level
Human Resource Management practices in the of complaints seems limited, with the caveat that
sector (MOH 2015: 47). the sample size is rather small (EPHI 2014).

64 Focus groups in Ethiopia among health workers revealed common understaffing due to late arrivals, long breaks and a
general disregard for the necessity of staffing clinics. Absences are frequently motivated by responsibilities at second
jobs. Lack of management and manager’s reluctance to confront physicians inspires lower level workers to behave
accordingly, leading to high absenteeism and low productivity at all levels. (Lindelow, Serneels and Lemma, 2003). http://
citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.542.2159&rep=rep1&type=pdf
65 Based on a sample of 3,383 individual health service staff.
107

37. Earlier research based on focus group 40. As in the education sector, reports on the health
discussions undertaken in 2003 reflected a greater sector highlight the need for substantial additional
level of dissatisfaction. “Both health workers and resources in order to facilitate further progress
users provide a sobering account that indicates a in expanding and improving service delivery. This
high level of distrust and frustration with the public includes a need to further scale up staffing (see
health system” (Lindelow and Serneels 2006: 2228). Health Sector Transformation Plan 2015),67 as
Complaints included slow service and long waiting well as further investments in the quality of initial
times, no service orientation among providers, poor training, certification, and continuous learning.
diagnostic knowledge, as well as growing corruption According to the 2014 SPA+, only 50 percent of
due to increased opportunities to ‘leak’ medicines most types of health facilities have the level of
and medical supplies as well as staff time to private staff that is defined as the norm for this facility
practice. The difference in results between 2003 (EPHI et al. 2014: 37).
and 2014 suggests a ‘direction of travel’ towards
improvements; however, the research designs are A1.4 Revenue sector
very different and both have significant limitations.
Key sector dynamics
38. Complaints mechanisms at the facility level
appear to be still very scarce. Among health 41. The current Ethiopian Revenue and Customs
facilities other than health posts, only 19 percent Authority (ERCA) is a relatively recently created
had a feedback mechanism for clients, and among agency. It was established in mid-2008, when
health posts, the level was even lower, at only 6 the Ministry of Revenue, the Ethiopian Customs
percent (among the 802 facilities surveyed). The Authority, and the Federal Inland Revenue
ratio is significantly higher for referral and primary Authority were merged. The agency is currently
hospitals (59 and 52 percent respectively) (EPHI headed by a Director General and five Deputy
et al. 2014: 35).65 Directors Generals, and has 34 offices across
Ethiopia, as well as external posts in the port of
39. An important initiative in the sector is the Djibouti and in Burbera, Somalia.
Health Development Army, which, similar to
the other sectors discussed here, has been 42. An important challenge is that revenue
established as a way of social mobilization and collection relative to GDP remains limited,
the spreading of good practices. Under the related albeit gradually increasing. For FY15/16, tax
‘Health Extension Program; the government has revenue reached 13.5 percent, up from 11.7
invested in rapidly training thousands of health percent five years earlier. Identified causes
extension workers, doubling their numbers to include a mix of policy (tax exemptions) as well
38,000 from 2011-2014.66 as administrative issues.

66 Ethiopia’s Fifth National Health Accounts, 2010/2011 (2014). Government of Ethiopia: Addis Ababa. https://www.
hfgproject.org/wp-content/uploads/2014/04/Ethiopia-Main-NHA-Report.pdf.
67 “The doctor, health officer, nurse and midwife to population ratio is 0.7 per 1000 population, far behind the minimum
threshold of 2.3 doctor, nurse and midwife to 1000 population ratio required to ensure high coverage with essential
health interventions (revised HRH strategy-MoH, 2014)”.
See:http://www.moh.gov.et/documents/26765/0/Health+Sector+Transformation+Plan/5542a23a-9bc7-46a2-8c1f-
8b32c2603208?version=1.0
 108

Civil service issues and reforms in the sector

43. ERCA exemplifies the challenges related Consequently, continuous monitoring and
to staffing and recruitment: staffing has been improving of business processes has significant
significantly scaled up in recent years - from relevance for an institution such as ERCA.
6,095 in 2011 to 11,500 in 2015. In addition to the
expansion in staff numbers, turnover is particularly 46. As an agency with a mandate that is critical
high for the agency. As a consequence, hiring has for the government, and in line with practices
been in the order of 2,000-3,000 staff per year. in several other countries for revenue agencies,
The agency has established a relatively intensive ERCA has been allowed to offer higher salary
3-month long induction training for new recruits. levels than those used in other MDAs. ERCAs
salaries range from 4,000 Birr at the lowest to
44. The creation of ERCA was in part linked 14,000 Birr at the highest level. While this was
to the BPR reform process. The BPR review of significantly higher than salaries offered by other
the previously existing MDAs responsible for agencies at the time of the agency’s establishment,
revenue collection was started in late 2007, there is now a sense that pay levels for civil servants
and concluded that a merger would help to in other areas is ‘catching up’, and that ERCA’s
reap efficiencies and to streamline processes, salaries are therefore no longer as attractive. As for
including with regards to trade facilitation, other sectors for which interviews were conducted,
i.e. making the settling of customs and tax ERCA’s management is also seeking permission to
obligations and refunds easier. In the view further increase remuneration levels.
of ERCA’s current leadership, while the BPR
reforms are outdated, there is a continuing need 47. Despite offering higher salary levels, turnover
for ongoing Business Process improvements. has been extremely high – reaching levels of up
to 50 percent of staff – and is one of the crucial
45. In principle, business process efficiency concerns of ERCA’s management. Several reasons
and reliability are crucial aspects for a revenue are seen as important: the challenges of being
agency, given that it has a large number of clients a tax auditor, remote postings for some of the
who must be served, while at the same time, it customs staff in particular, as well as the fact
must be credible that compliance will be enforced that businesses like to hire staff that have been
if necessary. Moreover, how the revenue agency trained by ERCA so as to assist with their tax
functions has a direct bearing on the state’s compliance. Also, most staff are hired to ERCA
ability to raise revenue, taxpayer compliance, as directly after completing their BA or higher-level
well as on the business sector – which is in turn degrees at universities and colleges, and such
critical to promoting growth and to job-creation. young employees tend to be more mobile.

Businesses like to hire staff that have

been trained by ERCA so as to assist
with their tax compliance.
109

48. ERCA management is seeking to address high the BSC system is used for target setting and
turn-over, and as part of this is paying particular monitoring, staff see very little translation of
attention to the hiring and promotion of female individual BSC results into rewards or actions
staff – who have been identified as less likely to taken in the case of poor performance.
leave the agency, and who are also seen as more
resistant to corruption. Because of this strategy, 51. The Change Army system is being deployed
currently 51 percent of ERCA’s staff is female across three levels: (i) employees, (ii) management,
(compared to 34 percent for the civil service and (iii) taxpayers. Among employees, ‘1 to 5’
overall), and in ‘decision-making’ roles (team meetings are held daily for 20-30 minutes, while
leader and above), female staff account for 30 other larger meetings are organized weekly at the
percent. Among staff dismissed for corruption in division level, bi-weekly at the management level,
recent years, 91 percent have been male, and 9 and monthly at a larger scale.
percent have been female.
52. The Change Army reform is seen positively
49. Reducing corruption is a major challenge also among revenue staff, albeit slightly less so
for ERCA. Taxpayers have a significant incentive than among other sectors. In particular, peer-to-
to offer bribe payments to tax officials so as to peer learning is seen as a benefit, and efficiency
reduce assessed tax obligations. Depending on and service delivery are considered to have
the size of the taxpayer, the bribes being offered improved. The spread of best practices is less
can be substantial. In 2013, the then head of ERCA frequently cited as a positive effect than among
and a deputy director general were arrested on staff from other sectors. One reason for this could
corruption charges. A Good Governance Plan has be that staffturnover holds back the establishment
been produced annually since 2013. of good practices.

50. The range of current reform tools is also being 53. Citizens Charter efforts were introduced only
pursued in ERCA. The BSC is being implemented in early 2016, and are still in the process of being
since the start of GTP I; divided into six strategic rolled out. The aim is to clearly publicize the rights
areas and three plans (operational plan, revenue and obligations of taxpayers; and this has been
collection plan, and good governance plan). Thus rolled out to all 34 offices of the agency. ERCA
far, the system is not automated and the manual has also formed of a partnership with private
compilation for thousands of employees is seen sector associations to jointly supervise tax and
as burdensome, while the implementation of an customs branch offices, aiming to identify issues
automated system is being pursued jointly with and options for improvement.
the Ministry of Public Service. Furthermore, there
are challenges with regards to ensuring that the 54. A set of complaints mechanisms are in place.
performance assessments help to assess strengths This includes a telephone hotline (which also
and weaknesses. According to the survey, while receives anonymous complaints), an internal ethics

Taxpayers have a significant incentive

to offer bribe payments to tax officials
 110

office, and a tax appeals system for taxpayers. The that reform initiatives and particular efforts to
number of customer service complaints has been implement them, e.g. within a given directorate,
increasing over time – which ERCA’s management are not followed through. The wider initiative
sees as related to having become more pro-active to reduce turnover to more moderate levels has
in terms of publicizing taxpayers’ rights. therefore direct implications also for ERCA’s
ability to make good use of the current set of
55. Discussions of ERCA management with civil service reform tools.
public wing representatives were started in
2015, while the level of activity and the impact 57. Apart from concerns about being able to
is seen as constrained thus far. In the focus group pay wages that are sufficiently attractive, there
discussion, held in late 2016, the following issues are also other important resource constraints.
were raised: while a policy exists on Citizens’ One important constraint is space: both to
Charters, these are not consistently available/ accommodate the growing number of staff, and
displayed at regional and woreda levels. There to ensure that offices serving taxpayers are well
is also a concern that tax and customs officers organized, and function smoothly even during
are not consistently knowledgeable, and as a ‘peak periods’ such as the end of the month or
consequence rules are applied unevenly, and it the quarter. IT functionality is another important
can be difficult for entrepreneurs or traders to resource constraint. Both the current IT systems
receive correct information on payments due. for customs (Automated System for Customs
The interaction through the Public Wing helps to Data, ASYCUDA) and tax (Standard Integrated
spread knowledge among the business community Government Tax Administration System, SIGITAS)
about existing tax and customs rules. Public Wing that ERCA operates are considered to be in need
discussions have also raised the need for some of updating. Network interruptions render it
legal amendments or clarifications. After an initial difficult to use systems continuously. This is also
effort of carrying out assessments of key issues for confirmed by survey results. These indicate a very
improvements was completed, follow-up action low availability of computers especially at woreda
was limited, and there is a perception that the levels, and the fact that network access remains
Public Wing meetings are a formality. One area difficult – according to the survey, internet access
for which follow-up was observed were initial is ‘good enough’ on 1 out of 5 days for revenue
improvements in working environments and some sector staff at woreda levels.
improvements of customs services.
58. Certain functionalities, such as e-filing, do
56. Given high staff turnover in recent years, there not really function reliably. The agency currently
has been a constant need to induct new staff in employs 70 IT specialists who support the running
the various performance and good governance of systems throughout the country; however, there
tools being used in the agency. The agency is is also a constant need to ensure that this team
seeking to ensure that new management staff in has up to date skills and can assist all offices
particular receives training on the BSC. Weak skills in a timely way. Furthermore, as ERCA officials
with regards to change management, motivating point out, due to resource constraints, it is often
staff, and agency leadership towards realizing necessary to make trade-offs between investing in
improvements are seen as ongoing challenges. facilitating voluntary tax compliance and investing
Very high turnover inevitably increases a risk in effective controls.
111

Available information on the impact of reform efforts

59. Revenue collection trends and limited export coordination. Thus far, export promotion
existing research suggests that at least some has focused on agricultural commodities, while
improvements have been achieved in recent there is an intention to also strengthen the
years. Revenue collection has increased since the export promotion of industrial production –
establishment of ERCA, significantly in absolute resource constraints permitting. In terms of staff
terms, and gradually in terms of the ratio of numbers, the Federal MoT has 966 approved
revenue to GDP. In absolute terms, tax revenue positions, out of which 491 positions were filled
(including customs duties) increased from as of March 2016.
23.8bn Birr in 2007/08, the year prior to ERCA’s
establishment, to 196.2bn Birr in 2015/16, i.e. 61. One fundamental challenge that the sector
almost 8.5 time.68 In relative terms, tax revenue faces is its capacity to implement and, as
(including customs duties) increased from 9.7 needed, reform the existing set of rules and
percent to 13.5 percent of GDP over the same regulations. There are some inconsistencies
period. At the same time, Debela and Hagos among existing rules and regulations that have
(2011) report for a small sample of tax offices been identified, but still need to be addressed.
and a customs office that a majority of taxpayers Given staff constraints, it has been challenging
believe that there have been improvements in to make progress on this agenda. Some observers
terms of waiting times at the tax administration also point out that it would be useful to consider
(as of mid-2010). whether there are sensible opportunities for
reducing the regulatory burden. Any regulation
A1.5 Trade sector that exists requires staff time and other costs
associated with enforcement, while at the same
Key sector characteristics and trends time also typically increasing the cost of doing
business. While many regulations serve important
60. The institutional structures responsible purposes – such as protecting consumers or
for trade play an important role with regards ensuring that pollution is being minimized – some
to Ethiopia’s economic development. The regulations can be duplicative, unnecessary, or
institutional structure consists of the Federal not designed in ways that minimize the burden
Ministry of Trade (MoT) and regional Offices of compliance.
of Trade and Industry. The Federal MoT also
oversees the Ethiopian Commodities Exchange, 62. Corruption is seen as a significant
the Consumer Protection Agency, and the Grain challenge. The MoT and Trade Office have a
Trade Enterprise. The ministry and offices play controlling function, and this creates incentives
an important role with regards to business to offer bribes for evading controls. Areas that
licensing and regulation as well as with regards are seen as particularly prone to bribery is to
to trade facilitation. Business licensing is a speed up licensing processes and seeking
decentralized function, while overall regulation approval for the import of goods – for example,
is a federal prerogative. The Federal Ministry has when traders are trying to import something
13 branch offices, primarily related to import and that does not meet the technical specifications

68 Source: IMF, all data for ‘General Government’.

 112

and inspection set by the trade institutions. 63. A success that has been achieved in recent
Certificates of competency (CoC) are another years is the integration of trade administration
type of licenses that are seen as prone to with customs for clearing imports and exports.
corruption. At the same time, non-government An on-line system has been established for this. A
stakeholders interviewed for this report argued key purpose is to be able to certify the standards
that licensing requirements in this regard can of imports and exports, and to make sure that
be overly narrow, and re-licensing requirements trade license requirements are checked when
are frequent. clearing customs.

Civil service issues and reforms

64. The Federal MoT is engaged on the current reallocations; while nothing is done in response
civil service reform tools, while still seeking to good performance.
to define how best to use them. Similar to
other federal ministries, the Federal MoT has 67. While key informants from federal sector
a Directorate of System Improvement and Human management had some reservations about the
Resource Management, which is meant to lead the relevance of the ‘Change Army’ for the sector,
implementation of civil service reform efforts for survey results suggest positive perceptions. The
the ministry, in collaboration with the Planning spread of best practices, as well as overall sector
and Information Management Directorate. efficiency and performance is seen as having
improved by trade sector staff across all levels
65. Survey results indicate that BPR reforms had of government.
a significant impact for the trade sector, which
is consistent with its particular role. Especially 68. A revised Citizens Charter is being prepared but
federal and regional level staff agree that was not yet published by March 2016. Meanwhile,
processes were dissolved or changed, and that the Ministry of Trade sees itself engaging more
processing times were reduced. frequently with representatives of the private
sector. Meetings with the Chamber of Commerce
66. With regards to the BSC, this is seen as a good and with sectoral associations are being held
idea in principle, but posing challenges in terms regularly at quarterly intervals, or more frequently
of actual implementation. One key challenge is when needed. As part of those consultations, draft
the availability of staff capable of formulating legislation is also being shared with stakeholders.
strategic plans, setting out a monitoring framework, Currently, the MoT and trade offices do not yet
and then actually monitoring implementation and operate hotlines to receive complaints; but plans
the achievement of targets. For the assessments exist to establish such a mechanism.
of individual staff, the lack of automation poses
a problem. Staff assessments have been carried 69. The Public Wing representatives for the trade
out for 2014 and 2015. The Ministry of trade sector are mainly chambers of commerce as
is engaging in discussions with the Ministry of well as some other associations. Chambers of
Public Service of how best to develop BSC plans commerce exist at the federal and regional levels,
and documents, as well as on the automatization as well as in over 500 cities. For over a decade,
of evaluations for individual staff. According to there have been various efforts at developing
survey results, at the federal level, actions taken public-private dialogue, with support from various
in response to poor performance are training or development partners, and an MoU was signed
113

between the federal Chamber of Commerce and is very much consistent with findings from other
the Ministry of Trade. There is some concern countries regarding performance management, as
about maintaining a sufficiently active dialogue, as reflected in the literature on civil service reforms
well as about the follow up to suggestions raised. (Hood 2006; Pollitt; etc.).

70. Overall, the range of civil service reform 71. As for the other sectors considered for this
tools – BSC, Change Army, Citizens Charter, good report, the federal Ministry of Trade faces
governance plans, and Public Wing engagements significant resource constraints. As a regular
– is not seen as an excessive burden; however, ministry, Ministry of Trade offers standard
there are multiple reporting requirements and salaries starting at 2,800 Birr/month and up to
requests for these plans and strategies that are 5,700 Birr/month at the director level (~130-263
seen as burdensome as well as repetitive. Also, USD). Consequently, positions often cannot be
given the institutional mandates, it is not always filled with qualified staff. In addition, the office
easy to set clear and quantifiable targets against space of the ministry is particularly constrained
which performance can be measured. This concern and old.

Annex 2
Methodology Notes for the
Civil Service Survey

72. This chapter covers the methodology of the tools which will allow us to better understand the
questionnaire development; followed by details of incentive environments which lead to different
the translation, the interview format, enumerator types of behavior and the determinants of service
training, team assignment and quality assurance; delivery in the civil service. What distinguishes
and how the sample was selected. civil servant surveys from other types of surveys
is not only their scale and scope, but the clear
73. The aim of the Ethiopia Civil Servant Survey shift from form to function. Much of the previous
is to gather micro-level data on the perceptions work in civil service reform defaulted to optimal
and experiences of civil servants, and on the key ‘forms’ for the civil service motivated by theory.
restraints to civil servants performing their duties Through the creation of improved survey modules,
to the best of their abilities, and to the provision the survey builds a foundation for the systematic
of public goods. This civil servant survey aims collection of data directly from individuals with
to contribute to the development of diagnostic first-hand experience.
 114

Building on existing civil servant surveys

74. The Ethiopia Civil Servant Survey is part of the issues. The survey follows and expands upon
‘Strengthening Research on the Civil Service’ (SRCS) similar work in Nigeria (Rasul and Rogger 2010)
initiative, which aims to develop the evidence and Ghana (Rasul, Rogger, and Williams 2015).
base for public sector reform by understanding
the characteristics of public officials and the 76. Survey of Civil Servants: Nigeria. In 2010 a
systems and organizations in which they work. research team, in partnership with the Nigerian
The SRCS aims to develop improved methods of government, visited institutions across the country
data collection and analysis on the civil service, to interview civil servants to answer some of the key
based on a complementary set of quantitative and questions arising from the public administration
qualitative approaches. literature and the debate on why bureaucracies
in the developing world often fail to fulfil their
75. By experimenting with survey design, the duty of providing critical services that support
project aims to create a new flagship survey on the capabilities of its people. To answer these
civil service staff and organizations around the questions -Why do civil services fail to deliver
world. The project has developed several civil public goods so much of the time? What are the
servant surveys across sub-Saharan Africa to key bottlenecks to effective service delivery?
assist governments in diagnosing key reform What is the experience of a civil servant working
issues. To inform these efforts, existing surveys in a developing country bureaucracy? - 4148 civil
of public officials are continually reviewed to assist servants from 63 organizations across Nigeria at
in the development of a database of questions all three tiers of government were interviewed. The
and methodologies. The Ethiopia Civil Servant survey was split into a quantitative questionnaire
Survey questionnaire draws extensively from this and a qualitative discussion process. This first
database, but expands upon it by including a set questionnaire forms of the basis of the Ethiopian
of modules aimed at addressing country-specific civil servant survey.

Questionnaire structure

77. To obtain reliable information on public-sector aims to understand the challenges faced by public
organizations, it was important to recognize that officers in their daily duties. It includes questions
protocol and language-use in civil services are on the demographics and work history of officers,
country-specific. The questionnaire was therefore how officers entered the service and their current
created in consultation with members of the relationship to it. It asks about their job – how
Ministry of Public Service and Human Resource satisfied they are, what they expect their career
Development (MPSHRD), a World Bank qualitative trajectory to be in the next few years. It asks about
mission from 7th to 17th March 2016, and a number their local environment, and the engagements they
of pilot interviews conducted to ensure question face daily.
viability, as well as to ensure question wording
and phrasing remained relevant and suitable to 79. In order to ensure questions remained relevant
the Ethiopian context. to each broad level of staff within the Ethiopian
civil service, the questionnaire was split into
78. The questionnaire aims to gather information three tracks: an employee-level track, a director-
on the experience of civil servants in the service. It level track and a head-of-organization (political
115

appointee)-level track. While some modules were government reforms, and how these have
covered in all three tracks, some modules were impacted operations on the ground in each
designed specifically to be administered to specific of the organizations.
tracks only, and not to all sampled individuals.
• Information: This section aims to explicitly
80. Below is a brief outline of all the sections covered assess the level of information that officials
by all three tracks and the purposes behind them: (directors and employees only) have about
the demographic conditions of their
• Demographics and Work History: The jurisdiction and the state of services in their
purpose of this section is to obtain basic respective sectors.
characteristics of the official. This allows us
to separate the experience of the service by • Information technology: The purpose of this
different groups. It also investigates the career section is to investigate information systems
trajectory of the individual civil servant, and in the organization, the type of information
thus what the context of their experience of available and how accessible it is.
the service has been to date.
• Public service motivation and locus of control:
• Time Use and Bottlenecks: This section This section uses the Perry (1996) Public
explores how the official uses their time Service Motivation Scale and Levenson’s
during a typical work day or week, and what IPC scale (1981) to measure motivation and
the obstacles they face at work are. This behavioural characteristics of civil servants.
module incorporates experimental methods
for the measurement of time use in order to 81. Below is a brief outline of the sections covered
help build the knowledge base of how best to by the employee-level track only:
measure this aspect of civil servant life.
• Recruitment and Selection: The aim of this
• Stakeholder Engagement: This section section is to identify the criteria on which
explores who the official interacts with in recruitment into the service is based.
their day-to-day. It aims to understand their
current relationships in the service, and how • Attitude: This section explores officers’
this determines their experience. attitude to their job and their organization;
their feedback sources; how satisfied they
• Reforms: This section aims to give us a are; what they believe drives promotion in the
sense of how individual officials feel about organization; their position relative to career
 116

goals, and how much they identify with the • Management practices: This section determines
mission and objectives of their directorate, the nature of core management practices
their organization and the service as a whole. from the frequency of monitoring, through
The purpose of this section is to get a sense what is done with that data, to procedures to
of officers’ motivation and the sources from implement targets.
which they derive it.
83. Below is a brief outline of the only section of
82. Below is a brief outline of the sections covered the survey administered to political appointees
by the director-level track only: only through the head of organization-level track:

• Turnover: This section investigates the • Benchmarking: A short section to assess

structure and nature of turnover in the extent to which woreda operations
the organization from a management benchmark against indicators measuring
perspective. institutional quality.

Questionnaire design by module

Demographic and work history

84. Objective: Collect data on basic through, and experience in, the civil service is
characteristics of the individuals that make up extremely limited. The aim of this module is to
the civil service. collect basic statistics on the individuals that
make up the civil service. This will not only provide
85. Rationale: Designing effective policy and a snapshot of the composition of the civil service,
implementing it efficiently and fairly requires but it will allow for the investigation of if and how
bureaucrats and the organizations they work in individual characteristics and experiences of the
to have the requisite capacity, incentives and civil service affect perceptions, motivation, and
motivation. The question “what makes a good performance.
civil servant?” is thus an important one. Despite
their importance, existing data on the basic 86. Methodology: Basic questions on age, role,
characteristics of civil servants and their trajectory tenure and educational background.

Management practices

87. Objective: Adapt the existing World productivity and management practices vary
Management Survey (WMS) instrument to the with bureaucrat characteristics (measured using
public sector (civil service) to measure the quality additional questions on their tenure, employment
of management practices within and across civil history, intrinsic motivation, and perceptions of
service organizations. organizational corruption).

88. Rationale: To measure the quality, and 89. Methodology: 1) Review the public
degree of adoption, of 18 management practices administration literature on management (e.g.
in the civil service sector to shed light on how Rose-Ackerman 1986, Wilson 1989), as well as
the correlation between bureaucratic output/ the WMS methodology (Bloom and Van Reenen,
117

1997) to determine how to approach the creation organizations in undertaking the questionnaire
of an adapted instrument for the civil service development process. At its inception, a number
setting. 2) Create a new instrument measuring 18 of pilots using semi-structured interviews like
key management practices relevant to the civil those used by Bloom and Van Reenen were
service context, measurable on a scale of 1-5, and held to outline key similarities and deviations
covering questions on operations, performance from the original WMS methodology (Rasul and
tracking, target setting and people management. Rogger 2016). Adapting the Nigerian/ Ghanaian-
setting instruments to the Ethiopian context
90. The management section (director track only) required only minimal changes in terminology
was based on an adapted survey methodology (e.g. changing the term ‘unit’ to the more service-
described in Bloom and Van Reenen (2007) relevant ‘directorate’).
and previously employed in the manufacturing,
retail, education and healthcare sectors. These 92. For each organization the focus was on the
dimensions are investigated through open ended two dimensions of management practice most
questions measuring the adoption of 18 basic focused in earlier academic work: (i) the autonomy
management practices, where the degree of provided to bureaucrats; (ii) the provision of
adoption is evaluated on a scale of 1 to 5, where incentives and monitoring of bureaucrats. The
a higher score indicates a higher level of adoption. autonomy index captures the extent to which: (i)
bureaucrats input into policy formulation and
91. The management section was first adapted implementation processes; (ii) the flexibility with
to the civil sector setting in Nigeria in 2010, which bureaucrats can be reorganized to respond
and then applied to the Ghanaian civil service to best practice and project peculiarities. The
in 2015, taking account of insights from the incentives/monitoring-based management index
public administration literature. Adapting the captures the extent to which an organization
WMS instrument to the civil service setting (first collects indicators of project performance, how
in Nigeria, then in Ghana, and most recently these indicators are reviewed, and whether
in Ethiopia) required extensive consultation bureaucrats are rewarded for achievements
with members of the relevant collaborating reflected in these indicators.

Turnover (Public Sector Staff Turnover Study 2014)

93. Objective: Recapture the state of turnover in 95. Methodology: The items of this module were
the civil service. directly transported from the Public Sector Staff
Turnover Study (2014), which itself was designed
94. Rationale: To add weight to the findings of to measure actual turnover and intentions to quit.
the Public Sector Staff Turnover Study (2014) and The Ethiopian civil servants survey uses only the
to provide additional instruments for validation items measuring actual turnovers and the drivers
checks. in order to allow validation checks.
 118

Recruitment and selection

96. Objective: Investigate individuals’ perception individual’s perceptions on the fairness of the
on the recruitment and selection strategies used system and what attracted those individuals to
by their organizations. the job in the first place, these questions can tell
us something about individuals’ motivation to join,
97. Rationale: Recruitment and selection has and stay, in their organization.
an important role to play in ensuring employee
performance and can play a pivotally important 98. Methodology: Questions on how individuals
role in shaping an organization’s effectiveness learned about open positions and were selected
and performance. The purpose of this section into those positions; and the factors that influence
is to investigate individuals’ perceptions on how how staff are treated. Questions were formulated
they believe selection and reward decisions are in such a way as to elicit what practices are in
made in their organization. By providing data on place without biasing the respondent.

Attitude

99. Objective: Collect data on individuals’ attitude career goals and relative performance, value of
towards different aspects of their job in the civil the public sector relative to the private sector,
service, and how this affects their satisfaction satisfaction of financial and non-financial rewards,
and motivation levels. The questions go beyond mission alignment and trust levels. This section
traditional issues of job satisfaction and focus emphasizes process-focused motivation which
on individuals’ feelings and beliefs regarding produces judgments, experiences, and behaviors
satisfaction, value, mission and trust within the congruent with an emphasis on process (feedback,
organization. mission alignment, value) rather than outcome
(salary, promotions). Only one question directly
100. Rationale: Attitude is of great interest to the asks respondents to rate their motivation, all other
government as a method to improve public service questions in this module present respondents
delivery. Two of the recent major reforms, the BSC with preference/ choice questions to prevent
and the Change Army, have focused on employee answers from being biased by social desirability
attitude. The extent to which attitude could be a concerns. While most questions provide choice
significant drive of service delivery needs to be options for the respondent, a number of open
addressed in order to say anything about these ended questions have been included to ensure
initiatives and similar future initiatives. Apart from the module captured the strength of motivation,
important questions about job satisfaction, this as well as the direction of motivation.
module aims to investigate deeper issues related
to individuals’ attitude and intrinsic motivation, 102. We supplement these questions on attitude
their sources of feedback, and how they perceive through the investigation of BSC scores, which
their career, the public service, and their value to allows us to measure motivation in terms of level
and fit within their organization. of performance at a goal-related task. The purpose
of this is to determine the extent to which attitude
101. Methodology: Attitude and motivation is scores on the BSC and positively associated with
measured in relative terms: feedback sources, performance scores on the BSC.
119

103. This section is also supplemented with the use between stated motivation and intrinsic motivation
of the Perry (1996) Public Service Motivation scale and perceptions of control as measured by these
and the Locus of Control scale, to determine the link standardized scales, see 12 for further details.

Time use and bottlenecks

104. Objective: 1) Methodological experiment: to methods were identified: (i) direct questioning
investigate the strengths of 5 different time use data (1 week, short version); (ii) direct questioning
collection methods to measure how civil servants use (1 week, short version); (iii) direct questioning
their time during working hours. 2) To understand (1 month, short version); (iv) graphical method;
some of the key bottlenecks civil servants encounter (v) time use diary. For all employee-level civil
in their day-to-day which prevent them from carrying servants outside of Addis Ababa one randomly
out their tasks effectively. chosen time use method was administered and
all data was collected during the interview.
105. Rationale: During the qualitative phase Addis-based respondents were provided with a
of the project, the inefficient use of time and template time use diary, to be filled in daily for
the ineffectiveness of meetings were raised as the duration of one week and collected by the
concerns by the MPSRHD and key stakeholders. enumerator at the end of the week. 2) Bottlenecks:
Time use modules were supplemented with
106. Methodology: 1) Time use: Review of existing questions on the effectiveness of meetings and
time use research methodologies to identify perceived bottlenecks to explore what some of
feasible methods to collect civil servant time the obstacles civil servants face at work on a
use data. 5 different time use data collection day-to-day basis are.

Information

107. Objective: Investigate the degree to which 108. Rationale: Information has been cited as a
information flows within and across civil service severe constraint by civil service organizations
organization by assessing civil servant awareness in Ethiopia. Addressing this concern can only be
and knowledge of conditions in their jurisdiction – done by understanding: the stock and distribution
demographic conditions and sector-specific conditions. of information among and between different types
 120

of civil servants, how information is disseminated, the extent to which civil servants have knowledge
how it is used as a strategic tool, and what the of basic information they are assumed/ expected
barriers against effective information flows are. by the service to be familiar with. Basic questions
on the tenets of the proclamation were asked. This
109. Methodology: 1) Information booklet: an provides a benchmark measure of information
information booklet, containing information on the with which to compare jurisdiction-specific items
demographic and service indicators measured in the of information. 3) Questions on demographic,
survey, was sent to randomly selected organizations jurisdiction-specific, verifiable items of information:
to observe if and how far the information was these items measure the official’s knowledge of the
disseminated throughout each organization. population, unemployment rate, and percentage of
Furthermore, the comparison of responses to the rural inhabitants of the jurisdiction. The responses
information questions between those organizations can be measured against the official census data.
that received the booklet and those that did not The understanding of such characteristics could
will provide a measure of the extent to which be key in the formulation of appropriate policies
information flows to and through organizations. and determine the quality of service delivery. 4)
Correlating randomly induced changes to the Questions on sector-specific, jurisdiction-specific,
stock of information to changes in productivity or verifiable items of information: these items measure
changes in service delivery will also provide insight the official’s knowledge of sector-specific indicators
into the extent to which improved information can in the jurisdiction, which can be measured relative
indeed lead to improvements in service delivery to official administrative data. Understanding the
2) Questions on uniform, verifiable items of existing state of service delivery could be vital in
information, constant across all civil servants: the directing policies towards the right areas, and affect
Civil Service Proclamation was chosen to determine the overall levels of service provision.

Information technology

110. Objective: Understand the state of information are integrated into processes, and the barriers to
technology and information management systems, full integration, will provide insight into the extent
the frequency of use, barriers to access and use, to which the functioning of systems facilitates
possible effects, and integration into the BSC. effective information and service delivery.

111. Rationale: Information is considered a key 112. Methodology: Respondents are asked
constraint to service delivery in Ethiopia. In directly about the existence and frequency of
addition to understanding the level of information use of information systems, their perceptions of
within organizations, understanding the the major challenges towards integrating these
information management systems and how they systems into processes, and their possible effects.

Stakeholder engagement

113. Objective: To investigate the extent to which 114. Rationale: It is increasingly expected that
civil servants interact with individuals or entities to implement policy and deliver public goods
outside their own organization. effectively all those stakeholders concerned
121

should be engaged during the policy-making, policy and the provision of public goods.
design and implementation process to maximize
the quality of information disseminated and 115. Methodology: Respondents are asked to
received and its effectiveness. Nevertheless, consider key projects or tasks they worked on
the types of engagement are important and an in the last year, and questions are asked on the
imbalance of the relative influence of different type and degree of engagement with different
stakeholders can negatively impact the design of stakeholders, public, private and community.

Reforms

116. Objective: To investigate general attitudes place when reforms were implemented there may
towards recent government reform efforts: exist a problem of recall; (iii) asking individuals
whether reform has been helpful in improving about counterfactuals would not provide reliable
service delivery and how; which aspects of reform information that is representative of civil servants’
to date have been lacking, in terms of awareness attitude towards reform; (iv) asking questions
and implementation; and whether civil servants about the BSC (civil servants often receive a
view the prospect of further reform favorably. high BSC score and as such are likely to view it
favorably) or the BPR (which resulted in the large-
117. Rationale: Before embarking upon further scale removal of employees and as such is likely to
reform, the government of Ethiopia has called for be viewed unfavorably, regardless of other aspects
a study on the status of civil service reforms: how of the reform) could give us skewed results,
past reform has been received by civil servants; (v) conversely, if civil servants are not directly
the elements of reform design and implementation affected by reforms in their day-to-day, they are
that have limited acceptance and effectiveness. less likely to have a strong opinion on that reform
The aim of the module is thus not only to know resulting in an ‘average’ opinion which has less
if civil servants thought a reform was positive or clear implications for reform. In order to address
negative, but why it was positive or negative. these concerns the reform module was split into
two tracks: (i) the employee- and director-level
118. Methodology: Addressing the broader tracks covered reform in general terms (e.g. “What
question of whether reform in general has helped percentage of your activities in the civil service
solve the problem of service delivery requires has been substantially affected by the following
an understanding of the general feeling among reforms?”); (ii) the head of organization (assumed
civil servants toward reform and whether there to have a better understanding of past reform by
is an appetite for reform moving forward. It was nature of their position) track covered five key
considered that evaluating the true impact of reforms in detail: Business Process Reengineering
specific reforms would be more complicated (BPR), Balance Scorecards (BSC), Change Army,
among employees for the following reasons: (i) Public Wing, and Citizens’ Charter. In particular,
sample censoring (the large increase in the number regarding the BPR and the issue of distant recall,
of civil servants in the years since reforms were heads of organization were explicitly asked where
implemented resulted in a smaller sample of civil they were working at the time of implementation,
servants who can comment on the impact of such in order to identify the possibility of differential
reforms); (ii) among those civil servants who were in responses based on direct experience.
 122

Woreda and city benchmarking

119. Objective: Incorporate existing measures of data and understand why certain trends or
institutional characteristics at the woreda level changes have been experienced ensure a greater
to provide up-to-date information on woreda- understanding of the quality of the data and/or
level operations, to provide a source of validation the trends in woreda-level operations.
checks, and to assess changes in institutional
characteristics over time. 121. Methodology: 11 core items were identified
from the latest questionnaire of the WCBS (round
120. Rationale: The Woreda and City Benchmarking 5) to be directly relevant to woreda-level operations
Survey (WCBS) has been ongoing in Ethiopia since in terms of financial processes and audits; human
2005, covering 5 rounds of survey until 2012/13. resource management; participatory planning;
This module of the survey incorporated items community consultations and committees. The
from the WCBS directly in order to observe the WCBS items themselves have been developed over
changes in institutional features at the woreda previous rounds of surveying, leading to round 3
level over time, independent of the phrasing (2009/10) and round 5 (2012/13) forming a panel
of questions. Furthermore, using existing and dataset in terms of variables.
available data sources to validate the survey

Public service motivation and locus of control

Public service motivation (PSM) Locus of control (LOC)

122. Objective: Apply Perry’s Public Service 125. Objective: Create a locus of control measurement
Motivation (PSM) scale (1996) to Ethiopian civil scale for civil servants.
servants at all three tiers of government.
126. Rationale: To measure the locus of control of
123. Rationale: To measure the degree of PSM civil servants to better understand their beliefs
among Ethiopian civil servants and investigate the about what determines the events that affect
interplay between intrinsic motivation (measured their professional lives. Understanding the locus
through the PSM scale) and extrinsic motivation of control of civil servants will give us an insight
(measured using questions on incentives, tenure into how civil servants perceive control systems
and perceptions of organizational corruption). in their organizations and how this affects their
motivation and performance levels.
124. Methodology: An unaltered version of Perry’s
original scale is used. The PSM construct is associated 127. Methodology: 1) Review of locus of control
conceptually with six dimensions: attraction to literature and existing measurement scale to
public policy making, commitment to the public determine how to approach the creation of an
interest, civic duty, social justice, self-sacrifice, and adapted scale for the civil service setting. 2) Create
compassion. The scale uses Likert-type items for a new scale by combining an adapted version of
each dimension. Self-administered questionnaire. an existing scale and a new sub-scale inspired
123

by the I-E scale. Levenson’s IPC scale (1981) was original rationale. The scale was also adapted in
identified as the most appropriate scale on which such a way that it would be relevant to all civil
to build a bureaucracy-specific scale of locus of servants equally across hierarchies and sectors.
control. Self-administered questionnaire.
129. Based on our knowledge of the literature on
128. In order to adapt the IPC Scale to a civil public service, it was determined that there was
service setting, each item in the scale was one key aspect of a civil servant’s life that would
assessed in terms of how it could be applied to not be captured by the IPC scale: institutional
the professional life of a civil servant. In those or system constraints. Civil servants typically
instances where items were applicable to this operate in environments regulated by strict rules
setting, no changes were made. In those instances and rigid structures which significantly impacts an
where items did not apply to the professional individual’s ability to use their discretion in how
setting of a civil servant the wording was adapted, they act and behave in the workplace. In order
but without changing the underlying objective of to capture systemic control over civil servants’
those items. Some items required only a small locus of control, a fourth sub-scale was added:
change, such as replacing the word ‘friends’ with Power of the System. To ensure consistency, it
‘colleagues’, whereas other items required a more was determined that this fourth scale should
substantial change. In order to ensure adapted complement the other three scales and should
items continued to capture the basic original be inspired by Rotter’s scale. Eight new items were
ideas, careful consideration was given to the included in this scale, each of which corresponds
original rationale behind each item. Using this as a to one of the pre-existing three-item sets in the
baseline, items were adapted to reflect the specific IPC scale, preserving the parallelism across the full
context of the bureaucracy without changing the adapted scale (becoming four-item sets).

Translation

130. Once developed in English, the questionnaire those in more remote or developing areas. The
was translated into Amharic by a translator translation went through a series of iterations,
with experience in the civil service, in order to including consultations with the MPSHRD, the
align the vocabulary of the document with that enumerators, and the Task Team Leader. The
frequently used in the civil service, particularly translated version of the document was used
those around the processes and reforms. The in a number of the practice interviews and was
questionnaire needed to be in Amharic to used in the pilot interviews, to verify the validity
generate greater understanding, familiarity, and of the translation and to provide an opportunity
comfortability of the respondents, particularly for further feedback from actual respondents.

Method of interview

131. Face-to-face interviews were considered the • Length of the survey: civil servant surveys
best enumeration methodology, using SurveyCTO tend to be quite long, so it is important to
software through Android tablets, for this type of ensure the respondent remains comfortable,
survey for various reasons: engaged and focused throughout the entire
 124

duration, which is more easily achieved in lot more difficult for them to pretend they
person than over the phone. are not there!).

• Complexity of the modules: civil servant 132. In addition, due to their specific nature of
surveys include innovative and potentially questioning and aiming to minimize the duration
complex modules which require the of the survey, paper forms were used for: the
enumerator to gauge the level of understanding Graphical Time Use option (measuring how civil
of the respondent, which is also more easily servants use their time with a pie chart); the Time
achieved in person. Use Diary (for civil servants based near Addis
Ababa for logistical reasons, measuring their
• Confidentiality and respondent ease: time use during an actual week with the use of an
it is significantly easier to achieve a hourly diary); the Perry Public Service Motivation
rapport between the respondent and scale and the Locus of Control scale.
the enumerator in person, and this is
of particular importance in relation to 133. The questionnaire was timed to take
concerns about confidentiality. approximately 1.5 hours for all sections, including
the self-administered forms. The respondents
• Commitment: respondents are more likely were first presented with an introduction into
to commit to starting and completing the purpose of the survey and the confidentiality
the interview if they know someone has agreement, and were able to opt out of the survey
travelled to interview them (and it is also a if they did not feel comfortable.

Enumerator training

134. Enumerators underwent comprehensive • Interviewing techniques: ensuring no prior

training by the field coordinators from Monday conceptions are taken to the interview, that
6th June to Saturday 11th June 2016, during which the interview is conducted professionally,
pilot surveys were also conducted at federal without bias, and to assert that the civil
institutions. The training consisted of: servants’ responses are confidential and
the purpose of the interview is to purely
• Project overview: providing an overview of capture their thoughts, opinions and beliefs,
the background, objectives and importance / rather than what the official documents /
relevance of the project, in order to generate guidelines state. Another major part of
and maintain interest, professionalism, and a the technique is to effectively manage the
sense of being involved in something bigger. duration of the interview.

• Core modules: specific training on items of • Tablet use: enumerators were provided
the survey, with a particular focus on: attitude, with specific training on how to use the
the World Management Survey, and the time- tablets and the SurveyCTO software.
use experiments. This section is to ensure a Practice sessions were designed in order
standardized understanding of the survey to ensure comfortability with the tablets
items across all enumerators, for the purpose and software.
of data quality.
125

• Research standards and confidentiality: across organizations.

generating a commitment to the scientific
method and the understanding of the need for • Practice interviews and pilots: in order to put
standardized, true measures of the situation the training into practice and to observe the
of the civil service was a major part of the understanding of the enumerators, a series of
training. The major parts of this training were practices interviews were conducted during
to: highlight the need for the survey to be the training week and guidance given by the
introduced and administered in the same coordinators on what could be improved.
way across respondents and organizations; Furthermore, interviews with actual civil
garner understanding of the requirement to servants in the field were organized for the
make respondents feel equally comfortable enumerators, observed by the coordinators,
with the survey and the confidentiality of the to apply their training in the actual context.
responses; provide each respondent with The practice interviews were important to
roughly the same amount of time to fill in the highlight areas that needed further training
questionnaire. This element of the training and to also highlight any areas where the
is vital to allow comparability of responses questionnaire could be improved.

Team makeup

135. Enumerators were assigned to teams of correctly labelled and all parts of the survey
4 during the training week, based on their were filled in; and to oversee the uploading of
performance and team fit. An organizational completed surveys to the SurveyCTO server. Each
supervisor was assigned within each team, with the team was assigned a list of jurisdictions, designed
responsibility for: organizing the daily activities; to maximize logistical efficiency. During the first
informing the organizations of the timing and week of out-of-Addis-Ababa enumeration, each
logistical requirements; provide enumerators team was assigned an external coordinator to
with the correct respondent codes and ensure ensure that the necessary checks, discussions,
that they have all of the correct materials for and general approaches as discussed during
the day. A survey supervisor was assigned to sit the training phase became systematized.
with the enumerators at the end of the day / Furthermore, the enumerators had a direct line
during travel days to discuss the questionnaire, and continuous communication with the TTL in
in particular the management scores, and general Addis Ababa, able to reach enumerators in cases
challenges; to ensure that all paper forms were where extra support was required.
 126

Project and data quality assurance

136. Throughout the enumeration period, the any irregularities in the data (primarily slight
data, through the online server, was followed identification-number errors). The paper forms
by the field coordination team to ensure that were entered by a data-entry team and a team
the project was going as planned and that the of analysts were hired to double check the data
data looked reasonable. After all interviews were entry. The data-entry passed a further stage of
conducted, a debriefing session was held between cleaning during the analysis and merging phase
the enumerators and the field coordinator to and more than 95% of forms have been perfectly
identify any possible issues with the data that matched to the survey data. Data checks show that
may not be visible in just the data itself and to the location and data identifiers correspond to the
understand general challenges. During the data- agreed itineraries and that the team identifiers
cleaning stage, much quality assurance and back- do not show any specific explanatory variation,
and-forth was conducted between the analysis suggesting that the data is unlikely to be biased
team and the field team in order to reconcile by team-specific effects.

Sample selection

137. To provide a large sample for statistical below outline exactly how these individuals were
analysis, while remaining within budget, the distributed at the sampling stage.
Ethiopian civil servants survey focused on the three
major policymaking tiers of government: Federal; 138. Federal offices – 66 individuals per ministry
Regional; and Woreda. The Ministry of Public Sector – 8 directorates – Finance, Planning, Resource
and Human Resource Development identified the Mobilization, 5 Randomly-Selected Service Delivery
5 core sectors that the survey should include: Directorates:
agriculture, education, health, revenue, and trade.
The government of Ethiopia also requested that all • The Head of Organization (or Deputy Head)
9 regional governments and 2 city administrations • The director of the Finance office
be included in the survey (11 ‘region-level’ • The director of the Planning office
governments). The decision was made then to plan • The director of the Resource Mobilization office
to interview a sufficient number of individuals from • The director of Service Delivery Directorate 1
each of those tiers and allocate the remaining funds • The director of Service Delivery Directorate 2
to woreda-level interviews. With this methodology, • The director of Service Delivery Directorate 3
with the funds available, 70 woredas were included • The director of Service Delivery Directorate 4
in the target sample at the planning stage. At the • The director of Service Delivery Directorate 5
Federal level 330 individuals were planned to be • 57 randomly selected employees from the
interviewed; 550 at the Region level; and 1615 at above directorates
the Woreda level. Within each region 50 individuals
were targeted to be interviewed, except in Addis 139. Note, in some cases, where we do not know
Ababa, where the target was 40 due to not having the directorates of the individuals, we ensure
an agriculture bureau, and except in Oromiya, that there is at least one individual from each
where, due to additional funds becoming available, directorate (randomly selected), and then simply
the target became 60. Within each woreda, 25 randomly select among all professional staff in
individuals were planned to be sampled. The details the Ministry.
127

140. Regional offices – 10 individuals per bureau 142. For Oromiya regional bureau, due to additional
– 3 directorates – Finance, Planning, Randomly- funds made available, 12 individuals from each
Selected Service Delivery bureau were targeted for sampling
• The Head of Organization (or Deputy Head)
• The director of the Finance office 143. Woreda offices – 5 individuals per office
• The director of the Planning office • The Head of Organization (or Deputy Head)
• The director of one randomly selected Service • One randomly selected director (many woredas
Delivery directorate have only one director)
• 6 randomly selected employees from the • 3 randomly sampled individuals
above directorates
144. Stratified randomization was conducted to
141. In cases where we do not know the directorates, select 70 woredas from the 9 regional states in a
we will ensure that one randomly selected individual way that is proportional to the size of the region
is sampled from each of the above directorates (in terms of number of woredas as per the 2007
and then 3 randomly selected individuals from the census). The table below outlines exactly how this
entire Bureau (that are professional staff). was done.

Table A.2.1: Total number of woredas per region and number of woredas sampled

Total number of Sample Target Sample

woredas (2007 Census) weights sample

Tigray 47 6% 70 5

Afar 30 4% 3

Amhara 139 19% 13

Oromia 277 38% 26

Somali 53 7% 5

Benishangul Gumuz 20 3% 2

SNNPR 145 20% 14

Gambela 13 2% 2

Harar 1 0% 0

Addis Ababa 10 1% 0

Dire Dawa 1 0% 0

Special EA 4 70

740

736 excluding Special EAs

Source ethphcensusatlas_data.xls, sheet ‘Spatial Units’.

 128

145. The following woredas, however, were removed will be surveyed and not woredas, due to the
from the sample due to security or logistic relevance to policy and the recent large number
infeasibility and replaced with the woredas noted of changes in woreda structure in Addis Ababa.
“included instead”. The “included instead” woredas
were simply next on the list of the randomly 146. Pre-survey exclusions: the following list
selected woreda list per region. For Addis Ababa of woredas were identified as security risks or
City Administration, the city administration alone logistically infeasible prior to the survey.

Table A.2.2: Pre-survey changes to woredas based on identified security risks or infeasible logistics

Region Wereda Note

Benishangul Gumuz KURMUK WEREDA Security

Benishangul Gumuz YASO WEREDA Included instead

Gambela WANTAWO WEREDA Security

Gambela JOR WEREDA Security

Gambela GAMBELLA ZURIYA WEREDA Included instead

Gambela ABOBO WEREDA Included instead

Oromia WOLMERA WEREDA Not on map

Oromia MOYALE WEREDA Logistically infeasible

Oromia MANA WEREDA Not on map

Oromia ARERO WEREDA Logistically infeasible

(Excluded as a replacement for the above)

Oromia HITOSA WEREDA Included instead

Oromia YEMALOGI WELEL WEREDA Security

(Excluded as a replacement for the above)

Oromia GIDA KIREMU WEREDA Included instead

Oromia MESELA WEREDA Included instead

SNNPR KOCHIRE WEREDA Unable to locate

SNNPR BASKETO WEREDA Included instead

Somali MULO WEREDA Security

Somali HARSHIN-WEREDA Included instead

129

147. During-survey exclusions: the following list logistically infeasible during to the survey.
of woredas were identified as security risks or

Table A.2.3: Woredas dropped during the survey period due to security challenges

Region Wereda Note

Amhara GONDAR ZURIYA WEREDA Conflict broke out during the enumeration period

Oromia ARSI NEGELE WEREDA Conflict broke out during the enumeration period

Oromia BEDENO WEREDA Conflict broke out during the enumeration period

Somali AFDEM WEREDA Adversely affected by flooding. Replaced with

AWUBERE WEREDA (the only remaining woreda

which the regional bureau staff were willing to

travel with the enumerators)

Table A.2.4: Final sample list of woredas and original itineraries

Region Wereda

Afar TELALAK WEREDA

Afar AWASH FENTALE WEREDA

Afar TERU WEREDA

Amhara AWABEL WEREDA

Amhara DAWA CHEFA WEREDA

Amhara DEBRESINA WEREDA

Amhara ENARJ ENAWGA WEREDA

Amhara TACH ARMACHOHO WEREDA

Amhara BASONA WERANA WEREDA

Amhara JABI TEHNAN WEREDA

Amhara KUTABER WEREDA

Amhara JILE TIMUGA WEREDA

Amhara GIDAN WEREDA

Amhara DEJEN WEREDA

 130

Region Wereda

Amhara SIMADA WEREDA

Benishangul Gumuz DIBATE WEREDA

Benishangul Gumuz YASO WEREDA

Gambela GAMBELLA ZURIYA WEREDA

Gambella ABOBO WOREDA

Oromia SERU WEREDA

Oromia DIGLUNA TIJO WEREDA

Oromia JARDEGA JARTE WEREDA

Oromia JELDU WEREDA

Oromia BEGI WEREDA

Oromia NONO WEREDA

Oromia SIRARO WEREDA

Oromia TOLE WEREDA

Oromia DEDESA WEREDA

Oromia GORO GUTU WEREDA

Oromia KOFELE WEREDA

Oromia BABILE WEREDA

Oromia ALE WEREDA

Oromia WERE JARSO WEREDA

Oromia GUDURU WEREDA

Oromia BAKO TIBE WEREDA

Oromia MIDGA TOLA WEREDA

Oromia AMIGNA WEREDA

Oromia HARO MAYA WEREDA

Oromia WADERA WEREDA

Oromia TIKUR ENCHINI WEREDA

Oromia HITOSA WEREDA

Oromia GIDA KIREMU WEREDA

Oromia MESELA WEREDA

SNNPR BONA ZURIYA WEREDA

SNNPR DECHA WEREDA

SNNPR DOYO GENA WEREDA

131

Region Wereda

SNNPR WENAGO WEREDA

SNNPR SHEBEDINO WEREDA

SNNPR HWASSA ZURIYA WEREDA

SNNPR GOMIBORA WEREDA

SNNPR AMARO WEREDA

SNNPR BENA TSEMAY WEREDA

SNNPR ANALIMO WEREDA

SNNPR KUCHA WEREDA

SNNPR DALE WEREDA

SNNPR CHERE WEREDA

SNNPR BASKETO WEREDA

Somali KEBRI BEYAH WEREDA

Somali ERER WEREDA

Somali JIJIGA WEREDA

Somali AWUBERE WEREDA

Tigray TAHTAY KORARO WEREDA

Tigray GULO MEHEDA WEREDA

Tigray HINTALO WAJIRAT WEREDA

Tigray WERE LEHE WEREDA

Tigray EROB WEREDA

Table A.2.5: Original team itineraries

Itinerary plan of Group 1 (39 Days)

Region Date Activity Woreda

Numbers

Oromia Sunday, 19, June, 2016 Travel from AA to Filiklik town of Were Jarso
Woreda (Oromia RS_North Shewa Zone)

Oromia Monday, 20, June, 2016 Data gathering in Were Jarso woreda offices 1

Amhara Tuesday, 21, June, 2016 Travel to Dejen Woreda (Amhara RS_East 2
Gojam Zone) and data gathering in the
woreda offices
 132

Itinerary plan of Group 1 (39 Days)

Region Date Activity Woreda

Numbers

Amhara Wednesday, 22, June, 2016 Travel to Lumame town of Awabel Woreda
(Amahara RS_East Gojam Zone) and data
gathering in the woreda offices

Amhara Thursday, 23, June, 2016 Data gathering in Awabel woreda offices 3

Amhara Friday, 24, June, 2016 Travel to Debrework town of Enarj Enawga 4
Woreda (Amhara RS_East Gojam Zone) and
data gathering in the woreda offices

Amhara Saturday, 25, June, 2016 Travel to Jabi Tehnan Woreda

Amhara Monday, 27, June, 2016 Data gathering in Jabi Tehnan woreda offices 5

Benishangul Gumuz Tuesday, 28, June, 2016 Travel to Dibate Woreda (BGRS_Metekel Zone)

Benishangul Gumuz Wednesday, 29, June, 2016 Data gathering in Dibate woreda offices 6

Amhara Thursday, 30, June, 2016 Travel to Bahir Dar city (Amahara RS) and
begin data collection at Regional offices

Amhara Friday, 01, July, 2016 Data gathering in Regional bureaus

Amhara Saturday, 02, July, 2016 Travel to Gonder Zuria Woreda

Amhara Monday, 04, July, 2016 Data gathering in Gonder Zuria woreda offices 7

Amhara Tuesday, 05, July, 2016 Travel to Tach Armachiho Woreda

Amhara Wednesday, 06, July, 2016 Data gathering in Tach Armachiho woreda offices 8

Tigray Thursday, 07, July, 2016 Travel to Tahtay Koraro Woreda (Tigray
RS_North Western Tigray Zone)

Tigray Friday, 08, July, 2016 Data gathering in Tahtay koraro woreda offices 9

Tigray Saturday, 09, July, 2016 Travel to May Kinetal town of Were Lehe
Woreda (Tigray RS)

Tigray Monday, 11, July, 2016 Data gathering in Were Lehe woreda offices 10

Tigray Tuesday, 12, July, 2016 Travel to Zala Anbessa town of Gulo
Meheda Woreda (Tigray RS)

Tigray Wednesday, 13, July, 2016 Data gathering in Gulo Meheda woreda offices 11

Tigray Thursday, 14, July, 2016 Travel to Alitena town of Erob Woreda
(Tigray RS) and start data gathering

Tigray Friday, 15, July, 2016 Finalise data gathering in Erob woreda offices 12

Tigray Saturday, 16, July, 2016 Travel to Makalle City (Tigray RS)
133

Itinerary plan of Group 1 (39 Days)

Region Date Activity Woreda

Numbers

Tigray Monday, 18, July, 2016 Data gathering inTigray Regional State bureaus

Tigray Tuesday, 19, July, 2016 Data gathering inTigray Regional State bureaus

Tigray Wednesday, 20, July, 2016 Travel to Adigudom town of Hintalo Wajirat 13
Woreda (Tigry RS) and data gathering in the
woreda offices

Amhara Thursday, 21, July, 2016 Travel to Gidan Woreda (Amahara RS)

Amhara Friday, 22, July, 2016 Data gathering in Gidan woreda offices 14

Amhara Saturday, 23, July, 2016 Travel to Simada Woreda (Amahara RS_
South Gonder Zone) and data gathering in
the woreda offices

Amhara Monday, 25, July, 2016 Data gathering in Simada woreda offices 15

Afar Tuesday, 26, July, 2016 Travel to Teru Woreda (Afar RS)

Afar Wednesday, 27, July, 2016 Data gathering in Teru woreda offices 16

Addis Ababa Thursday, 28, July, 2016 Back to Addis

Itinerary plan of Group 2 (39 Days)

Region Date Activity Woreda

Numbers

Amhara Sunday, 19, June, 2016 Travel to Kutaber Woreda (Amahara RS_
South Wollo Zone)

Amhara Monday, 20, June, 2016 Data gathering in Kutaber woreda offices 1

Amhara Tuesday, 21, June, 2016 Travel to Chefa Gula Woreda (Amahara RS_
oromo Zone)

Amhara Wednesday, 22, June, 2016 Data gathering in Chefa Gula woreda offices 2

Amhara Thursday, 23, June, 2016 Travel to Jile Timuga Woreda (Amahara 3
RS_oromo Zone) and data gathering in the
woreda offices

Amhara Friday, 24, June, 2016 Travel to Debresina Woreda (Amahara

RS_South wollo Zone)

Amhara Monday, 27, June, 2016 Data gathering in Debresina woreda offices 4
 134

Itinerary plan of Group 2 (39 Days)

Region Date Activity Woreda

Numbers

Amhara Tuesday, 28, June, 2016 Travel to Gudo Beret of Basona Werana 5
Woreda (Amahara RS_North Shewa Zone)
and data gathering in the woreda offices

Addis Ababa Wednesday, 29, June, 2016 Back to Addis

Afar Thursday, 30, June, 2016 Travel to Awash Fentale Woreda (Afar RS)

Afar Friday, 01, July, 2016 Data gathering in Awash Fentale woreda 6
offices

Afar Saturday, 02, July, 2016 Travel to Telalak Woreda (Afar RS)

Afar Monday, 04, July, 2016 Data gathering in Telalak woreda offices 7

Afar Tuesday, 05, July, 2016 Travel to Semera town (Afar RS) and begin
data collection in Afar Regional Offices

Afar Wednesday, 06, July, 2016 Finalise Data gathering in Afar regional
Bureaus

Oromia Thursday, 07, July, 2016 Travel to Mesela woreda (Oromia RS_west
Hararghe Zone)

Oromia Friday, 08, July, 2016 Data gathering in Mesela woreda offices 8

Oromia Saturday, 09, July, 2016 Travel to Karamile town of Goro Gutu
Woreda (oromia RS_ East Hararghe zone)

Oromia Monday, 11, July, 2016 Data gathering in Goro Gutu woreda offices 9

Dire Dawa Tuesday, 12, July, 2016 Travel to Dire Dawa city and begin data
collection at Regional Offices

Dire Dawa Wednesday, 13, July, 2016 Data gathering in the regional bureaus

Somali Thursday, 14, July, 2016 Travel to Erer woreda (Somali RS) and data 10
gatering in the woreda offices

Somali Friday, 15, July, 2016 Travel to Afdem Woreda and data gathering 11
in the woreda offices

Oromia Saturday, 16, July, 2016 Travel to Bedeno Woreda

Oromia Monday, 18, July, 2016 Data gathering in the offices of Bedeno 12
woreda offices

Oromia Tuesday, 19, July, 2016 Travel to Haromaya Woreda and data 13
gathering in the woreda offices

Harar Wednesday, 20, July, 2016 Travel to Harar city and data gathering in 14
the regional offices
135

Itinerary plan of Group 2 (39 Days)

Region Date Activity Woreda

Numbers

Oromia Thursday, 21, July, 2016 Travel to Midega Tola Woreda and data
gathering in the woreda offices

Oromia Friday, 22, July, 2016 Travel to Babile Woreda and data gathering 15
in the woreda offices

Somali Saturday, 23, July, 2016 Travel to Jijiga

Somali Monday, 25, July, 2016 Data gathering in Somali Regional Offices

Somali Tuesday, 26, July, 2016 Finalise data collection in Regional Offices

Somali Wednesday, 27, July, 2016 Data gathering in Jijiga zuria woreda offices 16

Somali Thursday, 28, July, 2016 Travel to Kebri beyah Woreda and data 17
gathering in the woreda offices

Somali Friday, 29, July, 2016 Travel to Harshin Woreda and data 18
gathering in the woreda offices

Addis Ababa Saturday, 30, July, 2016 Back to Addis

Itinerary plan of Group 3 (41 Days)

Region Date Activity Woreda

Numbers

SNNPR Sunday, 19, June, 2016 Travel from AA to Gembora Woreda

(SNNPRS_Hadiya Zone)

SNNPR Monday, 20, June, 2016 Data gathering in the sampled 5 offices of 1
Gembora woreda

SNNPR Tuesday, 21, June, 2016 Travel to Foniko town of Analimo Woreda 2
(SNNPRS_Hadiya Zone) and data gathering
in the woreda offices

SNNPR Wednesday, 22, June, 2016 Travel to Doyo Gena Woreda (SNNPRS_ 3
Kembata Zone) and data gathering in the
woreda offices

SNNPR Thursday, 23, June, 2016 Travel to Selam ber town of kucha Woreda
(SNNPRS_Gamo Gofa Zone)

SNNPR Friday, 24, June, 2016 Data gathering in Kucha woreda offices 4
 136

Itinerary plan of Group 3 (41 Days)

Region Date Activity Woreda

Numbers

SNNPR Saturday, 25, June, 2016 Travel to Laska town of Basketo special
Woreda (SNNPRS)

SNNPR Monday, 27, June, 2016 Data gathering in Basketo S. woreda offices 5

SNNPR Tuesday, 28, June, 2016 Travel from Basketo to Key Afer town of Bena
tsemay Woreda (SNNPRS_South omo Zone)

SNNPR Wednesday, 29, June, 2016 Data gathering in bena tsemay woreda offices 6

SNNPR Thursday, 30, June, 2016 Travel from Bena tsemay to Hawassa city
(SNNPRS) and begin Regional office data
colection

SNNPR Friday, 01, July, 2016 Finalise data gathering in Regional Bureaus

SNNPR Monday, 04, July, 2016 Data gathering in Hawassa zuria woreda offices 7

SNNPR Tuesday, 05, July, 2016 Travel to Leku town of Shebedino Woreda 8
(SNNPRS_Sidama Zone) and data gathering
in the woreda offices

SNNPR Wednesday, 06, July, 2016 Travel to Yirgalem town of Dale Woreda 9
(SNNPRS_Sidama Zone) and data gathering
in the woreda offices

SNNPR Thursday, 07, July, 2016 Travel to Wenago Woreda (SNNPRS_Gedeo 10

Zone) and data gathering in the woreda offices

SNNPR Friday, 08, July, 2016 Finalise data gathering in wenago woreda
offices

SNNPR Saturday, 09, July, 2016 Travel to Kele town of Amaro Special
Woreda (SNNPRS)

SNNPR Monday, 11, July, 2016 Data gathering in Amaro S. woreda offices 11

SNNPR Tuesday, 12, July, 2016 Travel to Bona town of Bona zuria Woreda
(SNNPRS)

SNNPR Wednesday, 13, July, 2016 Data gathering in bona zuria woreda offices 12

SNNPR Thursday, 14, July, 2016 Travel to Kele Chere Woreda (SNNPRS) and
start data gathering

SNNPR Friday, 15, July, 2016 Finalise data gathering in Kele Chere 13
Woreda offices

SNNPR Saturday, 16, July, 2016 Travel to Wadera Woreda (Oromia RS_Guji
zone)

SNNPR Monday, 18, July, 2016 Data gathering in Wadera woreda offices 14
137

Itinerary plan of Group 3 (41 Days)

Region Date Activity Woreda

Numbers

SNNPR Tuesday, 19, July, 2016 Travel to Arsi Negelle Woreda (Oromia
RS_West Arsi zone)

Oromia Wednesday, 20, July, 2016 Data gathering in Arsi Negelle woreda offices 15

Oromia Thursday, 21, July, 2016 Travel to Rope town of Siraro Woreda 16
(Oromia RS_West Arsi Zone) and data
gathering in the woreda offices

Oromia Friday, 22, July, 2016 Travel to Kofele Woreda (Oromia RS_West 17
Arsi Zone) and data gathering in the woreda
offices

Oromia Saturday, 23, July, 2016 Travel from kofele to Tijo town of Digluna
Tijo Woreda (Oromia RS_ Arsi Zone)

Oromia Monday, 25, July, 2016 Data gathering in Digluna Tijo woreda offices 18

Oromia Tuesday, 26, July, 2016 Travel to Eteya town of Hitosa Woreda 19
(Oromia RS_Arsi Zone) and data gathering
in the woreda offices

Oromia Wednesday, 27, July, 2016 Travel to Adele town of Amigna Woreda 20
(Oromia RS_ Arsi Zone) and data gathering
in the woreda offices

Oromia Thursday, 28, July, 2016 Travel to Seru Woreda (Oromia RS_ Arsi Zone)
and data gathering in the woreda offices

Oromia Friday, 29, July, 2016 Data gathering in Seru woreda offices 21

Addis Ababa Saturday, 30, July, 2016 Back to Addis

Itinerary plan of Group 4 (40 Days)

Region Date Activity Woreda

Numbers

Oromia Sunday, 19, June, 2016 Travel to Jeldu Woreda (Oromia RS_West
Shewa zone)

Oromia Monday, 20, June, 2016 Data gathering in Jeldu woreda offices 1

Oromia Tuesday, 21, June, 2016 Travel to Enchini town of Tikur Enchini 2
Woreda (Oromia RS_West Shewa Zone) and
data gathering in the woreda offices
 138

Itinerary plan of Group 4 (40 Days)

Region Date Activity Woreda

Numbers

Oromia Wednesday, 22, June, 2016 Travel to Bako town of Bako Tibe Woreda 3
(Oromia RS_West Shewa Zone) and data
gathering in the woreda offices

Oromia Thursday, 23, June, 2016 Travel to Kombolcha town of Guduru

Woreda (Oromia RS_Horo Guduru zone)

Oromia Friday, 24, June, 2016 Data gathering in Guduru woreda offices 4

Oromia Saturday, 25, June, 2016 Travel to Alibo town of Jerdga Jerte
Woreda (Oromia RS_Horo Guduru zone)

Oromia Monday, 27, June, 2016 Data gathering in Jerdga Jerte woreda offices 5

Oromia Tuesday, 28, June, 2016 Travel to Kiremu town of Gida kiremu
Woreda (Oromia RS_East Wellega zone)

Oromia Wednesday, 29, June, 2016 Data gathering in Gida Kiremu woreda offices 6

Oromia Thursday, 30, June, 2016 Travel to Begi Woreda (Oromia RS_ West
Wellega zone)

Oromia Friday, 01, July, 2016 Data gathering in Begi woreda offices 7

Benishangul Gumuz Saturday, 02, July, 2016 Travel to Assosa city (Benshangul Gumuz RS)

Benishangul Gumuz Monday, 04, July, 2016 Data gathering in the regional bureaus

Benishangul Gumuz Tuesday, 05, July, 2016 Finalise data gathering in the regional
bureaus

Benishangul Gumuz Wednesday, 06, July, 2016 Travel to Yaso Woreda (BGRS_ Kamashi zone)

Benishangul Gumuz Thursday, 07, July, 2016 Data gathering in Yaso woreda offices 8

Gambella Friday, 08, July, 2016 Travel to Gambella city (Gambella RS) and
begin Data Collection in Regional Offices

Gambella Monday, 11, July, 2016 Data gathering in the regional bureaus

Gambella Tuesday, 12, July, 2016 Data gathering in Gambella zuria woreda 9
offices

Gambella Wednesday, 13, July, 2016 Travel to Abobo Woreda and data gathering 10
in the woreda offices

Oromia Thursday, 14, July, 2016 Travel to Gore town of Ale Woreda and
start data gathering in the woreda offices

Oromia Friday, 15, July, 2016 Finalise data gathering of Ale Woreda and 11
travel to Denbi town of Dedesa Woreda
(Oromia RS_Illu Ababora Zone)
139

Itinerary plan of Group 4 (40 Days)

Region Date Activity Woreda

Numbers

Oromia Monday, 18, July, 2016 Data gathering in the offices of Dedesa woreda 12

SNNPR Tuesday, 19, July, 2016 Travel to Chiri town of Decha Woreda
(SNNPRS_ Kaffa Zone)

SNNPR Wednesday, 20, July, 2016 Data gathering in the offices of Decha 13
woreda

Oromia Thursday, 21, July, 2016 Travel to Silk amba town of Nono Woreda
(Oromia RS_ West Shewa zone)

Oromia Friday, 22, July, 2016 Data gathering in Nono woreda offices 14

Oromia Saturday, 23, July, 2016 Travel to Bantu town of Tole Woreda
(Oromia RS_ South West Shewa zone)

Oromia Monday, 25, July, 2016 Data gathering in Tole woreda offices 15

Addis Ababa Tuesday, 26, July, 2016 Back to Addis

Oromia Wednesday, 27, July, 2016 Data gathering in Oromia Regional Offices,
based in Addis Ababa

Oromia Thursday, 28, July, 2016 Data gathering in Oromia Regional Offices,
based in Addis Ababa

Addis Ababa Friday, 29, July, 2016 Data gathering in AA city administration
Bureaus
 140

Bibliography
Bloom, N., and J. Van Reenen. 2007. “Measuring and Explaining Management Practices across Firms and
Countries.” Quarterly Journal of Economics 122: 1351–1408.

Levenson, H. 1981. “Differentiating among Internality, Powerful Others, and Chance.” In Research with the
Locus of Control Construct, edited by H. M. Lefcourt. Vol. 1, 15–63. New York: Academic Press.

Perry, J. 1996. “Measuring Public Service Motivation: An Assessment of Construct Reliability and Validity.”
Journal of Public Administration Research and Theory 6: 5–22.

Rasul, I., and D. Rogger. 2017. “Management of Bureaucrats and Public Service Delivery: Evidence from the
Nigerian Civil Service.” Economic Journal.

Rasul, I., D. Rogger, and Williams, M. 2015. “The Effectiveness of Government Bureaucracy: A Study of the
Ghanaian Civil Service.”

Rose-Ackerman, S. 1986. “Reforming Public Bureaucracy Through Economic Incentives?” Journal of Law,
Economics and Organization 2: 131–61.

Wilson, J. 1989. Bureaucracy: What Government Agencies Do and Why They Do It. New York: Basic Books.
141

Annex 3
Methodology for Graphs and
the Management Practice
Indices

This annex provides further detail on the methodology behind the figures in the report and how they
can be interpreted. The first part of the annex explains the methodology and interpretation of each
type of figure used in the report; the second part of the annex details the creation of the management
practice indices, based on the World Management Survey methodology.

A.3.1 Methodological notes on individual graphs

Figure 2.1

Methodology

This graph uses historical data from the civil service statistical abstract and the World Development
Indicators to plot the total number of civil servants against the total population.

How to interpret the output of this graph

The left-hand axis refers to the total population and corresponds to the red line. The right-hand axis
refers to the total number of civil servants and corresponds to the blue line. The x-axis refers to the
years in the Gregorian calendar.

Figure 2.2

Methodology

This graph uses historical fiscal data from the Government of Ethiopia to plot the spending on salary
to the public sector as a percentage of gross domestic product (GDP).
 142

How to interpret the output of this graph

The blue line refers to the left-hand axis, which represents the percentage of GDP that is spent on
public-sector wages. The x-axis refers to the years in the Gregorian calendar.

Figure 2.3

Methodology

This graph combines historical data on the wage bill, GDP, total government revenue, and total
government expenditure for groups of countries to show the wage bill to the public sector as a
percentage of GDP and government expenditure and revenue.

How to interpret the output of this graph

The bars refer to the bottom axis of percentages: the grey bar represents the wage bill as a percentage
of GDP; the orange bar represents the wage bill as a percentage of government revenue; the blue bar
represents the wage bill as a percentage of government expenditure.

Figure 2.4

Methodology

This graph uses historical data from the civil service statistical abstracts to plot the number of additional
civil servants per year from 1992/93 to 2013/14.

How to interpret the output of this graph

The x-axis refers to the years in the Gregorian calendar. The y-axis refers to the number of civil servants
added in the respective year. The bars show the growth, in terms of the number of civil servants, in
the civil service each year.

Figure 2.5

Methodology

This graph uses 2014 data from the World Bank Urban Labor Market study to show the distribution of
the education level of staff within the public sector and the private sector.
143

How to interpret the output of this graph

The y-axis represents the percentage of staff. The yellow segment within each bar represents the
percentage of staff that have a post-secondary qualification; the grey segment represents the percentage
of staff that have a secondary education qualification; the orange segment, the percentage of staff that
have a primary education qualification; and the blue segment, the percentage of staff with no formal
recorded education. The first bar refers to the private sector; the second bar to the public sector.

Figure 3.1

Methodology

This graph uses official documents to produce a timeline of key reforms and changes within the civil service.

How to interpret the output of this graph

Each box represents the key reform or programme, and the x-axis refers to the year in which the reform
was initiated (Gregorian calendar).

Figures 3.3; 3.4

Methodology

These figures use data from the Ethiopian Civil Servants Survey (ECSS) 2016. The figures calculate the
percentage of respondents by tier (sector) by counting the number of civil servants that selected the
respective reform as most successful and dividing this by the total number of civil servants in that tier (sector).

How to interpret the output of this graph

The x-axis refers to the percentage of respondents selecting the respective reform as especially
successful. The y-axis represents the tier of government (federal, regional, or woreda-level) for Figure
3.3; and the sector (agriculture, education, health, revenue, or trade) for Figure 3.4. The blue segment
of the bar refers to the proportion of civil servants that selected the BSC as especially successful; the
orange bar, the percentage that selected the Change Army; the grey bar, the percentage that selected
the Citizens Charter; the yellow bar, the percentage that selected ‘Other’.
 144

Figures 3.5

Methodology

This figure uses data from the Ethiopian Civil Servants Survey (ECSS) 2016. The figure calculates the
percentage of respondents by sector by counting the number of civil servants that selected ‘Agree’ or
‘Strongly Agree’ to the stated question and dividing this by the total number of civil servants in that sector.

How to interpret the output of this graph

The x-axis refers to the proportion69 of civil servants that agreed or strongly agreed to the statement.
The y-axis refers to the sector in which the civil servant currently works (agriculture, education, health,
revenue, or trade).

Figures 3.6 and 3.7

Methodology

These figures use data from the Worldwide Governance Indicators, 2016. The graphs show the
government effectiveness by income blocks (3.6) and by country (3.7).

How to interpret the output of this graph

The x-axis refers to the country or group. The y-axis refers to the government effectiveness index, on a
range from 0 to 100, with higher values indicating greater effectiveness. Each bar within the country or group
referred to in the x-axis represents a different year of data (1998, 2004, 2009, 2014). Further information
on how the index is generated can be found under http://info.worldbank.org/governance/wgi/.

Figures 3.8

Methodology

This figure uses data from the ECSS 2016 based on the World Management Survey, see section A.3.2 below
for further details. This figure represents the raw management scores (and not the Z-Scores) for each
aggregate component of the management practices index. The aggregation is done as per the table below,
the averaging across each of the raw scores within the component.

69 This can be interpreted as a percentage if it is multiplied by 100. For example, a proportion of 0.2 means that 20% of
civil servants within the respective sector agreed or strongly agreed to the respective statement.
145

How to interpret the output of this graph

Each bar in the graph refers to each aggregate component of the management practices index
(monitoring, staffing, roles, staff involvement, targeting, flexibility, and incentives; see table below).
Each individual item (question) within the management practices index is scored from 1 (the lowest)
to 5 (the highest). Therefore, a higher score represents a higher average score across all of the items
within the aggregate component. See the section below for details of the individual items that make
up each aggregate component and examples of what constitutes a low score and a high score.

Figures 3.9, 3.11, 3.13, 3.14, 3.15, 3.16, 3.17, 3.18

Methodology

This figure uses data from the ECSS 2016 based on the methodology used for the World Management
Survey, see section below for further details. This figure represents the raw management scores (and not
the Z-Scores) for the respective aggregate component of the management practices index (monitoring,
staffing, roles, staff involvement, targeting, flexibility, and incentives). The aggregation is done as per
the table below, the averaging across each of the raw scores within the component.

How to interpret the output of this graph

The x-axis refers to the score from 1 to 5 in the management component. The y-axis refers to the
percentage of directors. The blue bar is a histogram which plots the exact percentage of respondents
per score; the green line is a smooth version of the histogram, to show what the distribution of scores
would look like if the scores on the x-axis were broken up into smaller and smaller partitions.

Figures 3.10, 3.12

Methodology

This figure uses data from the ECSS 2016, using the management practices module asked to directors
(based on the World Management Survey), and a measure of the intensity with which the BSC reforms
were introduced within the organization. The intensity of the BSC reform is based on the organization-
level average response to the following question, asked to all civil servants: “What percentage of your
activities in the civil service has been substantially affected, both positively and negatively, by the
following reforms – BSC?”.

How to interpret the output of this graph

The x-axis refers to the organization-level average response to the percentage of activities affected
by the BSC. The y-axis refers to the organization-level average management Z-Score (see the next
section) for the respective aggregate component of management practices (performance targeting,
 146

monitoring, performance incentives, roles, flexibility, staff involvement, and staffing – see section A.3.2
below for further details). The Z-Score in the y-axis can be interpreted as a ‘one-standard-deviation’
change – that is a Z-Score of -1 in the performance targeting component means that the organization
is one standard-deviation lower than the average score across all organizations. A standard deviation
is the average deviation from the mean within the distribution (i.e. a standard deviation of 2 means that
the scores are on average 2 units above or below the mean). The R-Squared represents the proportion
of the variation in the management practice component that is explained by the variation in the BSC
reform intensity (an R-Squared of 0.3199 means that 32% of the variation in the monitoring score
across organizations is explained by the percentage of activities in the organization that have been
affected by the BSC).

Figures 4.1, 5.1

Methodology

This figure uses data from the ECSS 2016. This makes use of only the top-ranked responses.
This takes the number of times the category is selected as the most important determinant of
promotion and divides this number by the number of times any category is selected as the most
important determinant. I.e. this is the number of ‘Rank 1’ responses for that category divided by the
total number of ‘Rank 1’ responses over all categories. Figure 4.1 refers to employees only. Figure
5.1 refers to employees and directors.

How to interpret the output of this graph

The x-axis refers to the percentage of ‘Rank 1’ responses that select the respective category. The y-axis
is a list of the options. Therefore, a figure of 57% for Merit/performance in Figure 4.1 means that, of
all of the ‘Rank 1’ responses, 57% are within Merit/performance.

Figures 4.2, 4.3, 4.4

Methodology

This figure uses data from the ECSS 2016. This makes use of any time the category is selected within
the top-3 most important determinants. This graph shows the proportion of times that the respective
category is within the top-3 most important determinants out of all responses. I.e. this is the total
number of times the category is ranked 1-3, divided by the total number of civil servants that responded
to this question. All figures refer to questions asked to employees only.

How to interpret the output of this graph

The x-axis and bars refer to the grouping of responses (by tier or sector). The y-axis refers to the
proportion of civil servants that selected the respective criterion (the y-axis label) in one of the top-3
147

most important determinants. Therefore, a value of 0.7 for Merit/performance in Figure 4.2 for Amhara
among the region-level civil servants means that 70% of civil servants in the Amhara regional bureaus
selected Merit/performance within the top-3 determinants of promotion. By the same reasoning,
just over 60% of woreda-level civil servants in Afar selected ‘Merit/performance’ within the top-3
determinants of promotion.

Figures 4.5, 4.8, 4.10, 4.11, 5.6, 5.7, 5.14, 5.21

Methodology

This figure uses data from the ECSS 2016. This is the percentage of respondents that select the
respective response for that specific question, grouped by tier of government. All histograms in the
report are of the same nature. Figure 4.8 was asked to employees and directors. All other figures refer
to questions asked to employees only.

How to interpret the output of this graph

Each sub-plot refers to the tier of government of the respondents. The y-axis refers to the percentage
of respondents within that tier of government. Each bar reflects the percentage of civil servants within
that tier of government that selected that option (Agree / Strongly Agree etc) for the respective question.

Figures 4.6, 4.7

Methodology

This figure uses data from the ECSS 2016. This is the number of selections for the respective category
divided by the total number of selections (directors could select multiple options, with no ordering /
ranking), converted into a percentage. Figure 4.6 refers to a survey question asked to directors only.
Figure 4.7 to employees only.

How to interpret the output of this graph

The y-axis refers to each category or driver. The x-axis is the percentage of selections attributed to
the respective category.
 148

Figures 4.9, 5.15, 5.22, 5.23, 5.24, 5.25

Methodology

This figure uses data from the ECSS 2016. This is the proportion of respondents that select the respective
response for that specific question, grouped by sector of government (x-axis) and tier of government
(different bars). Each respondent can select multiple options, so this figure is measuring the proportion
of times that the respective criterion is chosen. Figure 4.9 was asked to employees only.

How to interpret the output of this graph

Each sub-plot refers to different criteria (‘MIS Directorate does not have any useful data’; and ‘Network
problems’). The x-axis refers to the sector in which the civil servants work. The y-axis refers to the
proportion of respondents within that sector of government that selected this option. Each bar
reflects the percentage of civil servants within that tier of government that selected that option for the
respective question.

Figure 5.2

Methodology

This figure uses data from the Standish Group 2014 and World Bank (2015a) Digital Governance
Database and categorizes large public-sector ICT projects as failed, partially failed, or succeeded.

How to interpret the output of this graph

Each segment represents the percentage of ICT projects that fall within the respective. The percentages
are also labelled.

Figures 5.3, 5.4, 5.5, 5.8, 5.9

Methodology

This figure uses data from the ECSS 2016. This is the average response across all directors within the
respective tier of government, sector, or region to the stated question. All facilities questions were
asked to directors only.

How to interpret the output of this graph

The x-axis shows the grouping of the respondent by sector or tier. The bar reflects the tier or the specific
region of the respondent. The y-axis is the value of the average response to the stated question within
149

the specific sub-group of respondents represented by the bar. The lines that overlay the bars represent
the 95% confidence interval,70 which shows the precision of the estimate of the mean response. A
wide confidence interval usually means that there is limited variation in responses (e.g. due to a small
number of respondents for the question).

Each sub-plot refers to different criteria (‘MIS Directorate does not have any useful data’; and ‘Network
problems’). The x-axis refers to the sector in which the civil servants work. The y-axis refers to the
proportion of respondents within that sector of government that selected this option. Each bar reflects
the percentage of civil servants within that tier of government that selected that option (Agree / Strongly
Agree etc) for the respective question.

Figure 5.10

Methodology

This figure uses data from the ECSS 2016. Each bar shows the percentage of respondents that select
the respective response for that specific question, grouped by tier of government. Figure 5.10 was
asked to employees only.

How to interpret the output of this graph

Each bar color refers to a tier of government: blue bars refer to federal employees; grey bars refer to
regional employees; and white bars to woreda-level employees. The x-axis refers to the value of the
response to the question: ‘Imagine when you started your motivation was 100. What number would
you say it is now relative to that?’. The y-axis represents the percentage of respondents that chose the
respective value within that tier of government. So a blue bar that corresponds to 150 on the x-axis and
50% on the y-axis means that 50% of federal employees selected 150 as the answer to the question.

Figure 5.11

Methodology

This figure uses data from the ECSS 2016, the Ghana Civil Servants Survey (Rasul, Rogger, and Williams,
2015), the Federal Bureau of Revenue Staff Survey (Keefer, 2014) the Indonesia Public Employee Survey
of Bureaucracy Reform Survey (Banuri and Keefer, 2012), and the Nigeria Civil Servant Survey (Rogger,
2010). Note that there is some variation in the question and answer choices among the surveys in different

70 The correct interpretation of a confidence interval of 95% can be stated in terms of repeated samples: If this statistic
was repeatedly calculated from numerous randomly selected samples, the fraction of calculated confidence intervals
(which would differ for each sample) that encompass the true population parameter would tend toward 95%. Or, there
is a 95% probability that the confidence interval from some repeated future experiment on a different sample from the
same population (if appropriately selected) contains the true population parameter.
 150

countries, so the results should be considered indicative. With overall job satisfaction, we can fairly
straightforwardly compare amongst surveys in Ghana, Indonesia and Pakistan. We can make comparisons
to the other countries, but this is conditional on the comparability of the concepts assessed.

How to interpret the output of this graph

Each data point represents an organization within one of the 5 countries, separated by the color /
shape of the marker of the data point in the graph. The x-axis refers to the ranking of the organization,
in terms of the satisfaction of its workers, out of all organizations sampled within that country. The
y-axis refers to the proportion of civil servants that stated that they were satisfied with their experience
in the civil service in that organization. See table 5.1 for further information.

Figure 5.12

Methodology

This figure uses data from the ECSS 2016. This figure represents the responses to the question “What
percentage of your time is used for ad hoc tasks that are not part of your main job?”. The question
was asked to employees and directors only.

How to interpret the output of this graph

The x-axis refers to the percentage of time. The y-axis refers to the percentage of directors and
employees who exhibit the respective percentage of time. The blue bar is a histogram which plots
the exact percentage of respondents per percentage of time; the green line is a smooth version of the
histogram, to show what the distribution of scores would look like if the scores on the x-axis were
broken up into smaller and smaller partitions.

Figure 5.13

Methodology

This figure uses data from the Worldwide Governance Indicators to show the control of corruption in
Rwanda and Ethiopia from 2005 to 2015.

How to interpret the output of this graph

The x-axis refers to the WGI indicator ‘control of corruption’, on a range from 0 to 100, with higher
values indicating better control of corruption. The y-axis refers to the year and country respectively.
The green bars refer to the aggregate score for the indicator. The black lines refer to the margins of
error for each of the indicator values.
151

Figure 5.16

Methodology

This figure uses data from the ECSS 2016 based on the time-use module. The figure aggregates responses
on civil servants’ use of their working time and plots the distributions of time spent on the respective
activity. These questions were asked to employees and directors only.

How to interpret the output of this graph

The x-axis refers to the number of hours during a working week. The y-axis refers to the percentage
of directors and employees who exhibit the respective number of hours. The blue bar is a histogram
which plots the exact percentage of respondents per hours of time; the green line is a smooth version
of the histogram, to show what the distribution of scores would look like if the scores on the x-axis
were broken up into smaller and smaller partitions.

Figure 5.17

Methodology

This figure uses data from the ECSS 2016. The figure displays the responses to the following question:
How many people would you say regularly give you tasks as part of your formal work duties? Can you
describe who these people are? Are they…”. The question was asked to employees and directors only.
Respondents could select multiple options, reflecting all of the people that regularly give them tasks.
The statistic used in the graph is the percentage of responses that correspond to each specific type of
person (i.e. the total number of responses for “Directors from the organization”, for example, divided by
the total number of responses.

How to interpret the output of this graph

The x-axis refers to the percentage of responses. The y-axis to the type of individual. A bar of 45%
on the ‘Director from the organiszation’ therefore means that 45% of responses out of all responses
are within the ‘Directors from the organization’ category.

Figures 5.18, 5.19, 5.20

Methodology

This figure uses data from the ECSS 2016. This set of questions was asked to employees only. This is the
percentage of respondents that choose a response within the bins on the x-axis, to the respective question.
 152

How to interpret the output of this graph

Each sub-plot refers to the tier of government of the respondents. The y-axis refers to the percentage
of respondents within that tier of government. The x-axis is a set of bins of responses to the question:
so that if a respondent chose 17%, she would fall into the 0-25% bin. Each bar reflects the percentage
of civil servants within that tier of government that responded with a value within the bin reflected in
the x-axis.

Figure 5.26

Methodology

This figure uses data from the ECSS 2016. This question was an open-ended question (“What is the
best thing about working in the civil service?”) asked to employees only. This is the percentage of
respondents that choose a response corresponding to those on the x-axis, after having categorized
all of the open-ended responses.

How to interpret the output of this graph

The y-axis refers to the percentage of employees. The x-axis is the respective response to the question.
Each bar reflects the percentage of civil servants that responded with a response corresponding to the
categories reflected in the x-axis.

Figure A1.1

Methodology

This figure uses data from the Food and Agriculture Organization of the United Nations 2016 (based on
data reported by each country). The figure shows year-on-year values in cereal yields in Ethiopia and in
Eastern Africa as a whole, from 1992 to 2014.

How to interpret the output of this graph

The y-axis refers to the yields measured for a given year. The x-axis refers to the year. The darker
blue line refers to the Eastern Africa average cereal yield; the lighter blue-grey line to the Ethiopian
cereal yield.
153

A 3.2 Management practices

The management practices indicators are which core management practices are in place,
creating using the items developed from along seven dimensions: targeting; monitoring;
the World Management Survey, detailed in performance incentives; job roles; organization
the table below. The items were asked to flexibility; staff involvement; and staffing, such
directors 71 or heads of organizations within as the identification and retention of staff The
the civil service, who are responsible for small table provides the exact item asked to each
teams of professional civil servants. The World director and the type of answer that would
Management Survey72 items are designed for achieve the lowest score (1) and the type of
management staff to identify the extent to answer that would achieve the highest score (5).

Creation of indices

Each item is transformed into a Z-Score, which is a With the Z-Scores for each item, the aggregate
unitless measure that measures the deviation from indices of targeting, monitoring, incentives, roles,
the mean of the distribution of scores in standard flexibility, staffing, staff involvement – the standard
deviations. Therefore, a Z-Score of 1 means sub-indices used in the World Management
that the respondent/respective organization Survey – are created as means of each individual
is 1 standard deviations above the mean and item listed within the aggregate category.
a Z-Score of -2 means that the respondent/
respective organization is 2 standard deviations The overall management index is the mean of all
below the mean. individual Z-Scores of the items.

71
Also referred to as Process Owners, Managers, and Supervisors
72
http://worldmanagementsurvey.org
 154

Table World Management Survey Indicators

Aggregation Item Score 1 Score 5

Targeting Does your Directorate have The directorate does not Targets are clearly defined
a clear set of targets derived have defined targets. for the directorate, manager,
from the organization’s goals and employee levels, and are
and objectives? Are they well understood by all staff.
used to determine your work All tasks are directly derived
schedule? from the targets, which are
regularly reviewed to ensure
they remain on track.

Targeting When you arrive at work Staff do not know what their Staff have a very good
each day, do you and your roles and responsibilities are. understanding of their roles
colleagues know what and responsibilities. Their
their individual roles and own roles and goals are
responsibilities are in clearly interconnected to
achieving the organization’s those of their organization.
goals?

Targeting How are targets and Neither targets nor Targets and performance
performance measures performance measures are measures are formally
communicated to staff in communicated to staff. communicated and
your directorate? understood by all staff.

Monitoring In what kind of ways does Directorate does not track Full set of indicators are
your Directorate track performance. tracked formally and
how well it is delivering continuously. Reviews are
services? Can you give me conducted regularly and
an example? involve representative of
all directorate staff groups.
The results of the review are
formally communicate to all
directorate staff.

Monitoring Are you involved in Not involved in performance Weekly*

performance review for your review*
Directorate? If so, how often
does this occur?

Performance How would under- Poor performers stay in their Poor performers are
incentives performance be tolerated in positions (no consequences). identified through regular
your Directorate? Can you give reviews and are put on
me an example of how such a a formal performance
case would be dealt with? improvement plan
immediately. This applies to
all staff.

Performance Given past experience, have There are no consequences Bad behaviour/ breaking the
incentives members of [respondent’s for bad behaviour/ breaking rules is addressed through
organization] been disciplined the rules. concrete action. If any
for breaking the rules of the employee breaks the rules,
civil service? the underlying issues will be
identified and rectified. This
155

Aggregation Item Score 1 Score 5

applies to all employees.

Performance Does your Directorate use Staff are rewarded equally There is a formal staff
incentives performance, targets, or (or not rewarded) irrespective evaluation system and
indicators for tracking and of performance. Individual performance is rewarded
rewarding (financially or non- performance is not tracked (financially or non-
financially) the performance formally. financially). Rewards are
of its employees? given as a consequence of
well-defined and monitored
individual achievements. This
applied to all staff.

Roles When staff in your How officers carry out their Officers have complete
Directorate are given tasks in assignments is decided by autonomy in deciding how to
their daily work, how much senior managers. Officers carry out their tasks.
discretion do they have to have no say.
carry out their assignments?
Can you give me an
example?

Roles Can most staff in your Staff do not contribute Management expects all
Directorate make substantive to policy formulation, staff to contribute to policy
contributions to the nor to decisions about formulation and decisions
policy formulation and implementation. about implementation
implementation process? (formally or informally), and
considers this part of their
duties.

Roles Is the workload of achieving A small minority of The burden of the

your Directorate’s targets staff undertake the vast directorate’s work is
evenly distributed across its majority of work within the distributed equally among
different employees, or do directorate. staff. Tasks are assigned in
some groups consistently such a way that the amount
shoulder a greater burden of time required and the level
than others? of difficulty are balanced out
so no member of staff finds
him/herself overburdened.

Roles Thinking about all the Staff are allocated to tasks The right staff are always
projects that your Directorate randomly. used for a task. Allocation
has been involved in since of tasks is based on staffs’
your appointment here, documented skills and
would you say that managers competencies.
and supervisors try to use the
right staff for the right job?

Flexibility Does your Directorate The directorate uses the The directorate tailors all
make efforts to adjust to same procedures no matter procedures to the specific
the specific needs and what. needs of its stakeholders.
specific requirements of The evolution of those needs
communities, clients, or results in adaptation to plans,
other stakeholders? project and policies.
 156

Aggregation Item Score 1 Score 5

Flexibility How flexible would you say New practices are not The adoption of new ideas
your Directorate is in terms adopted/ integrated in the and practices is an integral
of responding to new and directorate. part of the directorate’s work.
improved work practices or New practices are regularly
reforms? reviewed and considered,
and once adopted and
integrated across the
directorate within 6 months.

Staff involvement How do problems in your Ad-hoc, no set process Exposing problems and
/ Contribution directorate get exposed and for improvement. Deal suggesting solutions and
fixed? with problems as they improvements is part of all
arise without following an staffs’ daily duty. Continuous
established procedureOnce improvement is part of the
fixed, no further action taken. culture of the organization.
No suggestions form staff

Staff involvement What kind of feedback do No feedback from staff. Staff provide the feedback
/ Contribution you get in staff meetings? on which action plans will be
based. Focus on both good
and bad performance. Details
of the meetings are recorded
and communicated to all staff.

Staff involvement Let’s say you’ve agreed to No action taken. No changes In addition to 4, tools can
/ Contribution a follow up plan at one of made in the operations be checked up and reported
your meetings, what would process. to the manager in charge.
happen if the plan wasn’t Meetings (formal/ informal)
enacted? are held to look into the
root causes of problems and
preventive actions are taken
for future similar task.

Staffing Do you think the Directorate does not put Senior management
management of your emphasis on talent believes that attracting
Directorate think about and developing talent is
attracting talented people important. There is a clear
to your Directorate and then system for identifying and
doing their best to keep attracting talent, developing
them? For example, by and retaining talent.
ensuring they are happy and
engaged with their work.

Staffing If two senior level staff joined No promotion system (no Promotion system is
your Directorate five years one in the organization has based on performance.
ago and one was much better been promoted for years) Organization actively
at their work than the other, The promotion system is identifies, develops and
would he/she be promoted based on tenure promotes top performers.
through the service faster? Regular assessments,
clear set of indicators and
personalised career plans
for individuals (regularly
revised).
157

Notes * This question is not score on a scale from 1 to 5, but is simply a coded response where the least frequent possible
response is “Not involved in performance review” and the most frequent possible response is “Weekly”. World
Management Indicators based on Bloom and Van Reenen (2007), adapted to the civil service following Rasul and
Rogger (2017) in Nigeria and Rasul, Rogger, and Williams (forthcoming) in Ghana. See http://worldmanagementsurvey.
org/ for further details on the World Management Survey instruments.

Image References
Cover Page. ‘Africa Partnerships Hamlin Fistula 7.’ Department of Foreign Affairs and Trade’s photostream.
Photo by Lucy Perry. Flickr.

Cover Page. ‘Upgrade training for Microsoft Windows 7 and Microsoft Office 2010’. ILRI. Photo by Liya
Dejene. Flickr.

Cover Page & Page 87. ‘Cattle being watered at the Ghibe River in southwestern Ethiopia’. ILRI. Photo
by Stevie Mann. Flickr.

Page ix. ‘Working Their Way Out Of Poverty’. DFID - UK Department for International Development’s
photostream. Photo by Gavin Houtheusen. Flickr.

Page x. ‘Cementing Ethiopia’s progress’. DFID - UK Department for International Development’s photostream.
Photo by Simon Davis. Flickr.

Page 1. ‘Workers building new irrigation system in Sebaya’. Trocaire. Photo by Jeannie O’Brien. Flickr.

Page 1. ‘Ploughing with cattle in southwestern Ethiopia’. ILRI. Photo by Stevie Mann. Flickr.

Page 6. ‘Interviewing poultry farmer in Ethiopia.’ ILRI. Photo by ILRI/Bettridge. Flickr.

Page 18. ‘Seeing cement sales soar.’ DFID - UK Department for International Development’s photostream.
Photo by Simon Davis. Flickr.

Page 19. ‘Dil Chora Hospital, Dire Dawa, Ethiopia, 2010’. US Army Africa. Flickr.

Pgae 44. ‘Classroom construction with well-built roof and walls for Arbisi children.’ ILRI. Photo by
ILRI. Flickr.

Page 45. ‘Mulusayoo teacher training’. One Laptop per Child. Photo by Daniel Drake. Flickr.

Page 55. ‘Sifting good beans from bad’. DFID - UK Department for International Development’s photostream.
Photo by Pete Lewis. Flickr.
 158

Page 55. ‘Africa Partnerships Hamlin Fistula 8’. Department of Foreign Affairs and Trade’s photostream.
Photo by Lucy Perry. Flickr.

Page 63. ‘Africa Partnership Hamlin Fistula 6’. Department of Foreign Affairs and Trade’s photostream.
Photo by Lucy Perry. Flickr.

Page 87. ‘A new generation of industrial leaders’. FID - UK Department for International Development’s
photostream. Photo by Gavin Houtheusen. Flickr.

Page 95. ‘Farmer winnowing tef in Bochessa, Ethiopia’. Photo by Ryan Kilpatrick. Flickr.

All other images used are free imags from pexels.com.