http://ijhpm.com
Int J Health Policy Manag 2019, 8(1), 28–39 doi 10.15171/ijhpm.2018.88
Original Article
Decentralisation and Management of Human Resource 
for Health in the Health System of Ghana: A Decision 
Space Analysis
Anthony Mwinkaara Sumah1, Leonard Baatiema2,3* 
Abstract Article History:
Background: The implications of decentralisation on human resource for health management has not received adequate Received: 10 August 2017
research attention despite the presupposition that the concept of decentralisation leads to the transfer of management Accepted: 8 September 2018
authority and discretion for human resource management from national levels to subnational levels. This study aims at ePublished: 23 September 2018
investigating the extent to which decentralisation practice transfers management autonomy and discretion to subnational 
units, and the effect of the level of decision space on human resource management in the health sector.
Methods: A mixed methods study design was adopted employing a cross-sectional survey and  a document analysis. 
The respondents included health managers from the regional, district and hospital administrations as well as facility 
managers from the community-based health planning and services zones. A decision space framework was employed to 
measure management autonomy and discretion at various management levels of the study region. For the quantitative 
data, descriptive statistical analysis was used to analyse and report the data whilst the qualitative data was content-
analysed. 
Results: The study reported that in practice, management authority for core human resource functions such as 
recruitment, remuneration, personnel training and development are centralised rather than transferred to the subnational 
units. It further reveals that authority diminishes along the management continuum from the national to the community 
level. Decentralisation was however found to have led to greater autonomy in technical supervision and performance 
appraisal. The study also reported the existence of discrepancy between the wide decision space for performance 
assessment through technical supervision and performance appraisal exercised by managers at the subnational level and 
a rather limited discretion for providing incentives or rewards to staff.
Conclusion: The practice of decentralisation in the Ghanaian health sector is more apparent than real. The limited 
autonomy and discretion in the management of human resource at the subnational units have potential adverse 
implications on effective recruitment, retention, development and distribution of health personnel. Therefore, further 
decision space is required at the subnational level to enhance effective and efficient management of human resource to 
attain the health sector objectives.
Keywords: Decentralisation, Decision Space, Human Resource Management, Health System, Ghana
Copyright: © 2019 The Author(s); Published by Kerman University of Medical Sciences. This is an open-access article 
distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/
by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is *Correspondence to:
properly cited. Leonard Baatiema  
Citation: Sumah AM, Baatiema L. Decentralisation and management of human resource for health in the health system Email:
of Ghana: a decision space analysis. Int J Health Policy Manag. 2019;8(1):28–39.  doi:10.15171/ijhpm.2018.88 baatiemaleonard@gmail.com
Key Messages 
Implications for policy makers
• The effect of decentralisation on the management of human resource for health is largely dependent on the form of decentralisation adopted 
and the context within which it is applied.
• To realise the benefits of decentralisation, there is the need to transfer authority and discretion for managing human resource from centralised 
bodies to localised levels. 
• Limited or lack of decision space at the localised level regarding personnel recruitment, remuneration, retention, distribution and development 
could result in adverse implications including shortage of staff.
• Wide decision space for performance assessment at the localised level should be accompanied with increased authority and discretion to 
provide rewards.
Implications for the public
Decentralising the management of human resource in the health sector has the potential to improve quality, efficiency, equity, innovation, and 
access to healthcare delivery as well as enhance local participation in the health sector decision-making. To realise these benefits, the authority and 
discretion for managing human resource should be transferred to the localised levels or units. Where there is limited or no transfer of this authority 
and discretion to the localised units as reported in this study, adverse human resource implications could result. These may include the persistent 
shortage of staff, high staff turnover, inadequate technical skills, inequitable distribution of staff and inconsistency between performance and rewards 
system.
Full list of authors’ affiliations is available at the end of the article.
Sumah and Baatiema
Background of how decentralisation has impacted on the management of 
Arguably, human resource is one of the most important human resource in Ghana’s health sector are limited. This 
resource in a labour-intensive health sector. Consequently, study, therefore, seeks to investigate the amount of choice that 
health sector organisational performance is inextricably is transferred from the central MoH to the subnational health 
linked to the effectiveness of its workforce which is further systems and its attendant implications for the management 
a function of the existing human resource management of human resource. The study findings have the potential to 
policies and practices.1,2 In an effort to improve health provide useful insights on the effect of decentralisation on the 
systems performance, major health reforms have taken management of human for health. Also, besides bridging this 
place globally and decentralisation is considered one of knowledge gap, the study will equally provide useful lessons 
such reforms aimed at improving the efficiency and equity for future decentralisation attempts or reforms in the health 
of health systems as well as responding to local preferences system in Ghana and beyond. To provide adequate background 
of populations.3 Besides the view that decentralisation is a to examine the human resource management implications of 
means of responding to system-wide performance challenges, decentralisation, the concepts of decentralisation and decision 
it is also argued that decentralisation is a means of improving space are examined in detail in the ensuing paragraphs.
performance in specific functional management areas such as 
human resources management.1 Concept of Decentralisation
A considerable amount of literature has been published Decentralisation is conceptualised as the transfer of authorities 
on the impact of decentralisation reforms on health systems. from central government bodies to lower levels within the 
Much of this literature focussed on health system change and public sector or to autonomous institutions.20 Essentially 
performance with emphasis on efficiency and equity in health 3 typologies of the concept exist in literature. As such 
service delivery and financing.4-9 As such, from the review decentralisation may be political, fiscal or administrative.21 
of the relevant literature, the impact of decentralisation Some researchers however propose a fourth typology – 
reforms on the management of human resource for health market decentralisation to account for the direction of public 
has not received adequate attention. Moreover, studies enterprise decentralisation in the form of privatisation and 
which have attempted to address this dearth of knowledge deregulation.22 Political decentralisation extends decision-
tended to focus on certain aspects such as human resource making power governing public institutions to citizens 
attraction and retention, policy restructuring, staff workload, at the local level while fiscal decentralisation relates to 
human resource stock, quality of care, staff knowledge on subnational ability to control financial resources, including 
decentralisation reforms and capacity of local authorities to revenue generation and allocation of funds.21 Administrative 
implement decentralisation among others.2,10-15 Nonetheless, decentralisation on the other hand is the transfer of 
the findings from these studies regarding the effect of responsibility for the planning, financing and management 
decentralisation are mixed. More importantly, although it of certain public functions from the central government 
is generally anticipated that decentralisation leads to the and its agencies to field units of government agencies, 
expansion of decision space at the local level,10,16,17 literature subordinate units or levels of government, semi-autonomous 
on the level of decision space available to local authorities public authorities or corporations, or area-wide, regional or 
in practice particularly with regards to human resource functional authorities. From the perspective that the transfer 
management remains poorly understood. The above situation of authority under administrative decentralisation may take 
persists notwithstanding the fact that, the control of human several forms along a continuum of lesser and greater degrees, 
resource management in the health sector is a major factor administrative decentralisation can be categorised into 3 main 
in local decision space that has far reaching effects on health typologies or forms namely deconcentration, devolution 
sector performance.5 and delegation. That is, administrative decentralisation may 
Ghana is an example of a country that has institutionalised therefore be characterised as deconcentration of authorities 
decentralisation as part its health sector reforms. from the central to local levels within a ministry or department 
Consequently, fiscal, administrative and political authority for of health, delegation of authorities to semi-autonomous bodies 
the delivery of healthcare is expected to be transferred from or devolution of responsibilities to autonomous or separate 
the central Ministry of Health (MoH) to Ghana health service local governments.21 A fourth variation of decentralisation 
(GHS) and other alternate institutions in an effort to ensure that has emerged in the literature is privatisation which 
efficiency, equity, service innovations, quality, transparency, involves the transfer of government functions to voluntary 
and accountability in the health sector.5,18 Through the organisations or private enterprise.20,23
passing into law of GHS and Teaching Hospital Act 1996, (Act Depending on the existing political, public administration 
525), Ghana established the legal framework to operationalise and health system structures in place, decentralisation may 
decentralisation in the health sector. The Act sought to also take several forms across and within health systems or 
delegate governance of healthcare from the central MoH to the countries.23 They may vary based on size of the sub-national 
GHS and teaching hospitals while applying deconcentration governments, level of decision-making autonomy or resource 
within the GHS and its subnational bodies. Two decades after allocation responsibility. These variations have given reason 
its implementation, few studies5,19 have attempted to examine to measure the practice of decentralisation, an underlying 
its implication on the performance of the health sector. To our rationale for this study.
knowledge, studies which provide a systematic understanding 
International Journal of Health Policy and Management, 2019, 8(1), 28–39 29
Sumah and Baatiema
Decentralisation in the Ghanaian Context analysis is employed. In this framework, ‘decision space’ 
Decentralisation efforts in Ghana predates the attainment refers to the range of effective choice that is allowed by the 
of independence from British rule.24 Concrete steps to central authorities to be utilised by the local authorities or 
institutionalise decentralisation as a preferred governance alternatively, the ‘range of choice’ local decision-makers have 
strategy however commence post-independence with the Local available in a decentralised context.5,16,25 Beyond the choice 
Government Act, Act 54 (1961). This piece of legislation built that is allowed for decision-making at the local levels, the 
on previous ordinances to establish towns and municipalities, decision space framework further offers the opportunity to 
maintained distinction between local government and evaluate the kind of choices local officials make with their 
central government structures and operated dual hierarchical increased discretion and the implications these choices have 
structures in parallel with central government structures.24 on the performance of the health system through a decision 
The current practice of decentralisation in Ghana however space map.16 To estimate the degree of discretion or range of 
draws its legal framework from the Local Governance Act, choices local authorities have over the various functions, the 
2016 (ACT 936). This act adopted devolution as the preferred decision space frameworks illustrates this as either ‘narrow,’ 
Government of Ghana’s policy for governance. As such, the ‘moderate,’ or and ‘wide’ (see decision map in Table 1).
act provides for the creation of district assemblies and the This paper adopts this approach because it provides the 
devolution of power to the district assemblies as the highest opportunity to disaggregate the various human resources 
political decision-making bodies in the districts. To this functions over which local officials have a defined range of 
extent, Ghana has been organised into 10 regions with 216 discretion and to examine the level of choice regarding these 
districts. Local governance structure in the Ghanaian context functions and how it impacts on the overall management of 
comprises the district assemblies who are monitored and human resource. In this study we disaggregated the human 
coordinated by the regional co-ordinating councils. Below resource function according to 3 categories (employment 
the district assemblies are the area councils and the unit management, personnel administration and staff 
committees which form the last 2 levels of the governance development) as espoused by a group of researchers26 and 
structure in accordance with Act 936 (Republic of Ghana apply the decision map to these categories.
2016).
Notwithstanding the adoption of the devolution as the preferred Methods 
government of Ghana governance policy, decentralisation Study Design
in Ghana’s health sector is at variance with devolution. The The study is part of a larger mixed-methods design conducted 
health sector rather saw a delegation from the MoH to the between March and December 2013 to investigate the degree 
GHS and Teaching hospitals and a deconcentration within of decentralisation in GHS and its attendant impact on health 
the GHS as provided by ACT 525. This process has led sector performance. The use of qualitative and quantitative 
to the establishment of the GHS at the national, regional, 
district and sub-district levels giving rise to a 5-tier (national, 
regional, district, sub-district, and community-based health Table 1. Map of Decision Space
planning and services [CHPS] zones) governance structure Range of Choice
currently in practice in the GHS. The variance in the forms Functions  Narrow  Moderate  Wide
of decentralisation practiced between the local government Finance
and the health sector has resulted in a mixed model of both Sources of revenue → → →
devolution and deconcentration leading to unclear and in Allocation of expenditures → → →
some instance contradictory reform efforts.24 The situation 
Income from fees contracts → → →
has also led to dual reporting lines for the GHS as it has to 
Service organization
fulfil reporting obligations both to the hierarchy of the GHS 
Hospital autonomy → → →
and the district assembly system.24 
Insurance plans → → →
Conceptual Framework-Decision Space Payment mechanisms → → →
The study relied on decision space framework developed Contracts with private providers → → →
by Bossert for analysing health sector decentralisation.16 Required programs/norms → → →
Based on the central principle that decentralisation transfers Human resources
authority, responsibility and resources for the management Salaries → → →
of healthcare from central MoH to regional and municipal Contracts → → →
governments as well as other autonomous institutions, this Civil service → → →
framework provides the means to analyse and compare in a Access rules
consistent manner the amount of choice that is transferred Targeting → → →
from central institutions to institutions at the periphery Government rules
of health systems. It therefore provides a mechanism for Facility boards → → →
analysing decentralisation based on the set of functions Health offices → → →
and degrees of choice that are formally transferred to local Community participation → → →
officials. To measure the range of choice, decision space Source: Bossert 1998.
30 International Journal of Health Policy and Management, 2019, 8(1), 28–39
Sumah and Baatiema
methods of inquiry are increasingly used in recent years due of the regional health service. Figure provides details of the 
to their epistemological value and practical demands.27-29 organisational structure.
To this, we collected data simultaneously and conducted 
the analysis as a strategy to triangulate and corroborate the Study Participants and Sampling Method 
findings from the 3 datasets. For the quantitative aspect of the survey, participants targeted 
This larger study design principally employed a cross included managers at the lower levels made up of, CHPS 
sectional survey and document analysis. The cross- zones, sub-districts, district hospitals and health directorates. 
sectional survey was in 2 parts. The first utilised structured Participants were part of management involved in the daily 
questionnaires to collect data from the relevant study operations and service delivery at the facility level, thus 
participants on the discretion or authority available to them take key and strategic decisions related to the operations of 
regarding human resource and other key management the health facility. The second part of the survey involved 
functions. The second part involved an in-depth interview of conducting an in-depth interview was conducted on 1 
a key informant to elicit information on healthcare delivery participant who is the head of the regional health service with 
relevant to human resources for health and other management over 15 years of management experiences in the Ghanaian 
functions covered by the study. This was done to ensure in- healthcare sector, thus most suitable to provide in-depth and 
depth and rich data was collected from the perspective of a key insightful information. 
decision-maker on the subject. The survey study design was A multi-staged sampling procedure was adopted for the 
favoured as most appropriate for this study due to its added survey. First, for participants in the cross-sectional study, the 
strength of facilitating a broader understanding of human study used a purposive sampling technique to enlist them. 
resources practices within a health sector.30 As a supplement Only heads of health facilities were enlisted in the study 
to the empirical data, document analysis approach using because of our exclusive focus on local actors responsible 
a document content analysis method was conducted as the for the implementation of decentralised authority and 
second part of this larger study to review all relevant national, responsibility. Here, all heads of regional and district health 
regional and local health policy documents on this subject services as well as hospitals in the study area or persons acting 
matter. This serves to validate and add on to the empirical in these positions were recruited. Given their small number, 
data from the cross-sectional survey. it was practical to enrol the regional health administration, 
In this paper, drawing on insights from the cross-sectional all district health administrations and hospitals in the region. 
survey, and document analysis of this larger study, we report Second, regarding the sub districts, given the large sample 
the extent to which decentralisation practice transfers size, we sampled the population using a standardised formula 
management autonomy and discretion to subnational units, to estimate a representative sample. Stratified sampling was 
and the effect of the level of decision space on human resource thereafter used to enlist the heads of the sub-district units 
management in the health sector. and CHPS zones. Each district was considered as a stratum 
The study got approval from the Regional Director of and the number of sub-district managers and CHPS zones 
Health Services through the Research Division of the Upper were statistically determined to reflect the proportion of the 
West Regional Health Service. The study was conducted sub-districts and CHPS zones in the district. Consequently, a 
according to sound ethical principles and guidelines as simple random sampling was used to recruit participants. A 
outlined in the Helsinki Declaration for Medical Research.31 purposive sampling technique was also employed to recruit 1 
Participants were informed about the study purpose and the participant for the semi-structured interview.
voluntary nature of their participation prior to data collection. 
Anonymity of participants’ identity and data confidentiality Study Instruments, Data Collection, and Analysis 
were emphasised prior to all data collection sessions with the A structured self-administered questionnaire and interview 
participants. guide were developed based on the decision space framework 
and used to elicit responses from all study participants. The 
Study Setting questionnaire had several sections seeking to understand 
The study was conducted in the Upper West Region, 1 of the decision space available to managers. The sections included 
10 administrative regions in Ghana. The administrative and level and practice of decentralisation regarding human 
governance structure of health service in this region is also resources, planning and budgeting, finance, governance 
reflective of the other 9 regions and may share similar human and community participation, limitations to effective 
resources for health characteristics. The governance structure decentralisation and facilitators of effective decentralisation. 
for public health service in the study region is 4-tier made The semi-structured interview with the Regional Director 
up of; a regional health administration at its apex, 11 district of Health was facilitated by an interview guide to gather in-
health administrations, 65 sub-districts and 166 CHPS zones. depth information relevant to the research question.30 The 
In addition, the region is made up of 6 service organisations interview guide was also designed to cover the decision space 
comprising a regional hospital and 5 district hospitals. The elements as espoused in the survey instrument but was flexible 
region also contains private healthcare providers as well to ensure further and more in-depth understanding of these 
as health training institutions. These are not included in key aspects were elicited. As regards the document analysis, 
this study because of their multiple reporting structures a predesigned electronic form was specifically designed to 
which inhibit their direct classification under the purview serve as a guide to review policy documents and institutional 
International Journal of Health Policy and Management, 2019, 8(1), 28–39 31
Sumah and Baatiema Figure1: Organisational structure of the regional health service 
 
 GHS NATIONAL 
HEADQUARTERS 
 
 
 Regional health 
administration 
  Regional level 
Regional Hospital 
  
 
 District health 
 administration 
 District Level 
 District hospital  
 
 
 
Sub-district health 
 management team Sub-district Level 
  Health Centre 
  
 
 
Community Health Community Level  management committee 
 CHPS compound 
  
Legend 
 
Health management units responsible for overall health service at the respective 
 levels 
 Service organisation responsible for providing curative & preventive care at the 
 
 respective levels  
  
 
Figure. Organisational Structure of th e Regional  Health Service. Abbreviations: CHPS, community-based health planning and services; GHS, 
Ghana health service.
Source: Adapted from GHS organisational structure 
Source: Adapted from GHS organisational structure.
 
 
records relevant to the 3 key domains of the decision space frequently used to ensure clearer and further information was 
analytical framework. The tool aided review and analysis collected. Overall, the interview lasted for 50 minutes.
of policy outputs and the extent to which existing policy The quantitative data was analysed using a statistical 
documents recognise and recommend actions to improve software analysis package, SPSS version 17.0. Data were 
decision space at the sub-national levels. analysed descriptively and presented mainly in tables and 
The questionnaire and interview guide were pilot-tested in 2 graphs. Data from the interview was analysed through open 
districts prior to the actual data collection. The choice of these coding by the first author. This process involved reading and 
districts was purposefully sampled based on convenience. rereading of interview transcript to code information relevant 
Consequently, the instruments were revised with minor to the study question. Relevant codes from the interview were 
changes made before utilised in the actual data collection used to support or clarify some essential points or underscore 
process. Results from this pilot exercise did not form part of some of the results from the cross-sectional survey.
the results reported in this paper. Data for the document analysis were content analysed. For 
Regarding the questionnaire administration, the data each of the human resource functional areas reported in the 
collection process proceeded as follows. Field research study, we analysed the existing relevant policy documents 
assistants were recruited and trained about the study and the in the service. The document analysis targeted at policy 
questionnaires administration procedure. Data collection documents relevant to human resource management. In all, 
followed a self-administration approach and so participants 8 key policy documents were consulted as shown in Table 
were handed out questionnaires for completion and returned S1 (see Supplementary file 1). We also consulted the annual 
to the research assistants. Verbal consent was sought from reports from the regional level and national, program of 
all participants before data collection. Participants were works, including the legal provision (Act 525) that triggered 
also made aware about the voluntary nature of the study. the decentralisation reforms agenda in the health service.
All sampled respondents completed the survey resulting 
in a 100% response rate. On the other hand, a face-to face Results 
interview was conducted in English by the first author using Basic Characteristic of Respondents
a semi-structured interview guide with the key informant, The study enlisted 164 participants comprising a regional 
in this case the Regional Director of Health Service. The director, 11 district directors, 6 medical directors/
interview process was audio-recorded to ensure all essential superintendents, 50 sub-district in-charges and 96 
information was captured without distortion. Prompts were community health officers. Of these, whilst 1 participant 
32 International Journal of Health Policy and Management, 2019, 8(1), 28–39
Sumah and Baatiema
(Regional Director) was interviewed using an interview to appoint a category of staff termed category D & E (staff 
guide, the remaining 163 participants made up of district grades for which secondary, post-secondary certificate or 
directors, medical directors/superintendents, sub-district basic education is the entry requirements). To further explain 
in-charges and community health officers were surveyed by the limitation to appoint or dismiss staff at the level of the 
means of a questionnaire. Of the 164 participants, 89 (54%) regional health administration, the respondent had this to say; 
were females. Participants were most likely to be community “it is just a functional authority and not an initiating authority” 
health nurses (58%), younger than 25 years (66%), and would (key informant interview). This informant further explains 
have held their current management position for less than that in the exercise of the discretion to appoint category D & 
5 years (73%). Of the overall participants recruited for the E staff as stated above, it must be with financial clearance or 
survey, 28 (17%) were registered nurses or midwives, medical approval from the ministry of finance and economic planning 
doctors comprised 5 (3%), physician/medical assistants (MoFEP). A review of the appointments policy of the GHS 
were 9 (5%) and community health nurses were 96 (59%) corroborates the respondent’s views. Depending on the 
respectively. The remaining 26% represented other categories category of staff to be appointed or dismissed, the authority 
including enrolled nurses. Table 2 shows the distribution of ranges from the head of state, through the GHS council, to the 
study respondents. director responsible for human resource development of the 
GHS except for categories D & E indicated above for which 
Decision Space the regional director exercises some level of autonomy.32 This 
Employment Management policy does not however make provisions for appointing 
The responses regarding the decision space which allows for the authority beyond the level of regional director such as the 
appointment, dismissal and determination of remuneration district, sub-district and CHPS zones level.
across 3 management levels (CHPS zones, sub-district, and Similar levels of autonomy was reported across the various 
district) within the regional health service is reported in Table management levels within the regional health service 
3. The responses pointed to a limited authority to manage regarding salary determination as indicated in Table 3. While 
these human resource functional areas. However, the regional the CHPS zones and the sub-district levels report a lack of 
director in the interview indicated that he had the prerogative authority regarding this function, 12 respondents representing 
70.6% of district and hospital managers indicate that they play 
Table 2. Basic Characteristics of Respondents a role such as recommending suspension of salary. 
Evidence from the review of policy and programmatic 
Characteristic Total (N = 164) %
documents also revealed that the regional administration plays 
Age (y)
a critical role in pay determination through pay processing 
<25 109 66
for final approval by the MoFEP.32 However, regarding the pay 
25-50 16 10 levels, the existing pay policy which took effect from January 
>50 39 24 2010 vested the authority to determine pay levels in the Fair 
Gender Wages and Salaries Commission of Ghana which is centrally 
Female 89 54 managed at the national level.33 Because of this central 
Male 75 46 determination of the pay levels exercised by this commission, 
Category of managers none of the management levels within the region exercises 
Regional director 1 1 discretion in that regard. Table 3 below provides details on 
District/municipal director 11 7 this subject. 
Medical director/superintendent 6 4
Sub-district in-charge 50 30 Personnel Administration
Community health officers 96 58 The finding regarding decision space in personnel 
Educational status administration is mixed as indicated in Table 4. Across the 6 
Postgraduate 17 10 functions that constitute personnel administration, the CHPS 
Bachelors/graduate 1 1 zones report a lack of autonomy and discretion to carry out 
Diploma 28 17 these functions. Though same can be reported for the sub-
Certificate 118 72 district managers regarding promotion, demotion, transfers/
Basic professional category reassignment and sanctions, higher levels of autonomy and 
Medical doctor 5 3 discretion have been reported at this level for personnel 
Registered general nurse or midwife 28 17 administration functions such as supervision (92%) and 
Medical/physician assistant 9 5 performance appraisal (88%). At the district and hospital 
Community health nurse 96 59 levels, the autonomy and discretion for sanctions, supervision 
Enrolled nurse 13 8 and performance appraisal of staff mainly resides with the 
Other 13 8 district directors and medical superintendents as displayed 
Years of experience in management position in Table 4. Again, high levels of authority for promotions 
Less than 5 years 119 73 and transfers/reassignments have been reported for these 
5 years or more 45 27 categories of health managers in the region. On the contrary, 
International Journal of Health Policy and Management, 2019, 8(1), 28–39 33
Sumah and Baatiema
Table 3. Respondents View on Employment Management Functions
CHOs Sub-district In-charges District Directors/Medical Overall TotalEmployment Superintendents
Management 
Functions Yes No Total Yes No Total Yes No Total Yes No
No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%)
Appointment 0 96 96 0 50 50 0 17 17 0 163 (0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (0.0) (100.0)
Dismissals/sack 0 96 96 0 50 50 0 17 17 0 163(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (0.0) (100.0)
Salary determination 0 96 96 0 50 50 12 5 17 12 151(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (70.6) (29.4) (100.0) (7.4) (92.6) 
Abbreviation: CHOs, Community Health Officer. 
Table 4.  Respondents Views on Personnel Administration Function
District Directors/Medical 
Personnel CHOs Sub-district In-charges Superintendents Overall Total
Administrative Yes No Total Yes No Total Yes No Total Yes No
Function No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%)
Promotions 0 96 96 0 50 50 14 3 17 14 149(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (82.4) (17.6) (100.0) (8.6) (91.4)
Demotions 0 96 96 0 50 50 2 15 17 2 161(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (11.8) (88.2) (100.0) (1.2) (98.8)
Transfers/ 0 96 96 0 50 50 15 2 17 15 148
Reassignments (0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (88.2) (11.8) (100.0) (9.2) (90.8)
Sanctions 0 96 96 0 50 50 17 0 17 17 146(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (100.0) (0.0) (100.0) (10.4) (89.6)
Supervision 0 96 96 46 4 50 17 0 17 63 100(0.0) (100.0) (100.0) (92.0) (8.0) (100.0) (100.0) (0.0) (100.0) (38.7) (61.3)
Performance appraisal 0 96 96 44 6 50 17 0 17 61 102(0.0) (100.0) (100.0) (88.0) (12.0) (100.0) (100.0) (0.0) (100.0) (37.4) (62.6)
Abbreviation: CHOs, Community Health Officer.
only 11% of the respondents are of the view that they exercise has been specified to include assessing staff due for promotion 
autonomy and discretion over demotions. The regional and recommending such staff for the attention of the regional 
health administration has equally reported mixed levels of director. In respect of the regional director, he has been vested 
discretion. In response to the level of authority and discretion with authority to; promote category D&E in line with the 
exercised at the regional health administration regarding prevailing promotion guidelines, assess category B & C staff 
personnel administration the response included; below their last grades and submit the assessment reports of 
‘...it depends on the HR function and the level. For such staff to the HRDD of the GHS for further action.
instance, while the authority to promote or demote category In respect of postings which also includes transfers and 
D & E staff in the region resides with the regional director, reassignment within the regional health service, the policy 
same cannot be said of other categories as that will have to be and guidelines on postings stipulates among other things 
referred to the GHS HRDD [Human Resource Development that, the regional director shall issue posting letters to the 
Directorate] for consideration.…in terms of postings, respective district directors who will in turn place such staff 
supervision and performance appraisal of staff within the appropriately into needy health facilities within the district.35 
region, the regional director is the final authority. However, Per the policy, the role of health facilities regarding these 
regarding sanctions, the regional director has limitations. For postings is to notify the posting authority of the assumption 
instance, action to dismiss a senior staff in the region can only of duty or otherwise of the staff.35,36
be initiated at the regional level through recommendation to Again, per the policy direction of the GHS, all staff are 
the director general [of GHS]. As such, the regional director required to be supervised and appraised by their immediate 
does not have absolute discretion and authority on a matter superiors or managers.37 Responses from the survey as 
like this” [key informant interview]. indicated in figure 3 largely confirms compliance of practice 
A review of the GHS policy and guidelines on promotion with policy except at the CHPS zones where respondents 
provides some insight into the responses reported above. Per expressed a lack of authority or discretion for supervision 
the said policy, no specific role or authority has been assigned and performance appraisal. Similarly, per existing policy, the 
to the CHOs and sub-district in-charges.34 On the other hand, GHS vests the authority for sanctions in the GHS council, the 
the role of the district directors and medical superintendents director general, directors at the various divisions, regions, 
34 International Journal of Health Policy and Management, 2019, 8(1), 28–39
Sumah and Baatiema
districts as well as institutional heads38 Table 4 provides of the HRDD of the GHS at the national level. The policy 
response on the role the respondents play regarding this also corroborates the reports of the key informant regarding 
subject. approving authority for non-degree awarding post-basic 
training and fellowships. Regarding CPDs, both the study 
Staff Development leave policy and the service employee handbook40 are not 
The decision space that have been reported from the survey explicit on the roles of the various management levels of the 
in respect of in-service training, continuous professional service on the subject matter.
development, post-basic training and fellowships across the Overall, the decision space allowed for the 12 human 
various management levels in the study area are reported in resource functions across the 4 management levels in the 
Table 5. The survey indicates a lack of autonomy and discretion regional health service have been summarised into a matrix as 
to manage these human resource functions at the CHPS zones shown in Table 6. This analysis was based on insights drawn 
and sub-district level. However, at the district level, 88% of from the questionnaires, the in-depth interview and document 
the respondents reported that they exercise authority for analysis. In the matrix, decision space has been categorised for 
post-basic training and continuous professional development each human resource function as narrow, moderate or wide. 
while 52% indicated same for in-service training. Regarding For a human resource function to be categorised narrow, it 
fellowships, the general decision space reported point to a lack means that at that management level, there is no exercise of 
of or limited discretion and authority regarding this human any discretion or authority regarding that function. Where 
resource function as shown in Table 5. the particular management level exercises some discretion 
The key informant interview reported similar levels of albeit limited, the decision space is considered moderate. 
decision space regarding human resource development Finally, where absolute discretion or full authority is exercised 
functions. over the function, the decision space at that management level 
“The ultimate reference point in matters of fellowships is considered wide. To further illustrate this, in response to 
and in-service training is the national level. But where the a question such as what role or authority you exercise over 
in-service training originates from the region, the regional promotion of staff? Responses that indicate the exercise of 
director remains the officer in-charge…, for fellowship, my no role or authority are rated narrow. Where the responses 
role is often to nominate or recommend eligible candidates indicate that staff are either nominated, assessed and or 
for consideration at the national level. For post-basic recommended to a senior management level to be promoted, 
training, recommendation is made to the HRDD for degree such responses are rated moderate. Finally, where the 
or higher qualification programs. However, with regard to responses indicate that the promotion of staff is ultimately or 
other lower qualifications, the regional director has authority absolutely carried out at a particular management level, the 
to grant these albeit this must be in line with the study leave discretion at this management level is rated wide.
policy in force at the time…in respect of CPDs [continuous 
professional development], they are profession-based and Discussion
depends on the individual to be able to finance them or In this study, we sought to examine the level of decision space 
secure financing from their respective institutions to be able available in practice to subnational health authorities and 
to attend.’ [key informant interview]. the implications of such decision space on the management 
Per the GHS 2016 study leave policy and guidelines of human resource for health in Ghana. Whiles in the 
regarding the level of responsibility for managing the various international literature decentralisation is expected to lead 
personnel development programs,39 fellowships (both local to the transfer of responsibility for the management of 
and external) require national level approval although human resources to subnational units,41,42 the findings from 
nomination or recommendation may be initiated from the this study suggest that most human resource functions are 
regional level. Equally, the policy confers the approving rather centralised. The study further revealed that discretion, 
authority for post- basic degree programs on the director limited as it may be, diminishes along the management 
Table 5. Respondents Views on Staff Development Functions
CHOs Sub-district In-charges District Directors/Medical Overall Total
Staff Development Superintendents
Functions Yes No Total Yes No Total Yes No Total Yes No
No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%) No. (%)
In-service training 0 96 96 0 50 50 9 8 17 9 154(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (52.9) (47.1) (100.0) (5.5) (94.5)
Continuous professional 0 96 96 0 50 50 15 2 17 15 148
development (0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (88.2) (11.8) (100.0) (9.2) (90.8)
Post-basic training 0 96 96 0 50 50 15 2 17 15 148(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (88.2) (11.8) (100.0) (9.2) (90.8)
Fellowships 0 96 96 0 50 50 16 1 17 16 147(0.0) (100.0) (100.0) (0.0) (100.0) (100.0) (94.1) (5.9) (100.0) (9.8) (90.2)
Abbreviation: CHOs, Community Health Officer.
International Journal of Health Policy and Management, 2019, 8(1), 28–39 35
Sumah and Baatiema
Table 6. Decision Space Across the Various Management Levels
Range of Decision Space Across Various Management Levels
Regional Level District Level Sub-district level Community Level
HR Functions
Regional Health District Health 
Administration Administration Hospitals Sub-district CHPS Zone (Compound)
Employment management
Appointments Narrow Narrow Narrow Narrow Narrow
Dismissals/firing Narrow Narrow Narrow Narrow Narrow
Remuneration determination Narrow Narrow Narrow Narrow Narrow
Personnel administration
Promotions/demotions Moderate Moderate Moderate Narrow Narrow
Transfers/reassignment Wide Moderate Narrow Narrow Narrow
Sanctions Wide Moderate Moderate Narrow Narrow
Supervision Wide Wide Wide Moderate Narrow
Performance appraisals Wide Wide Wide Moderate Narrow
Staff training and development
In-service training Moderate Moderate Moderate Narrow Narrow
Continuous professional development Moderate Moderate Moderate Narrow Narrow
Post-basic training Moderate Moderate Moderate Narrow Narrow
Fellowships Moderate Narrow Narrow Narrow Narrow
Abbreviations: HR, human resource; CHPS, community-based health planning and services.
Note: the following operational definitions were applied in contextualising the decision space analysis.
Narrow: There exist no autonomy and discretion over this HR function at the respective management level.
Moderate: Autonomy and discretion exercised for the HR function is partial/limited but not absolute, such as recommending action for the consideration of 
the next superior management level.
Wide: Ultimate authority for the HR function resides at this level of management.
continuum from the national to the community level. of recruitment and appointment functions within the GHS 
Decisions regarding hiring, firing or remuneration are highly could be attributable to the choice decentralisation adopted. 
centralised at the national level. This finding is in consonance That is, the choice of deconcentration over devolution or 
with another study which conducted a comparative analysis delegation as prescribed by ACT 525. As observed in the 
of decision space between 4 countries –  Ghana, Uganda, literature, deconcentration is unlikely to lead to the transfer of 
Zambia, and Philippines.5 Among the findings of the study, it recruitment and appointment responsibility to the lower levels, 
was reported that Ghana has a unified hierarchical personnel rather devolution and delegation are more likely.44 As such, 
structure in which decisions on contracting, hiring, firing and deconcentration may only lead to transfer of administrative 
civil service benefit is centralised. The findings also resonate responsibilities in line with national level directives for service 
with a similar study of decentralisation and decision space delivery at the sub-national or lower levels of management.45 
in the Suva subdivision in Fiji which sought to examine In accordance with this definition and as reported in this 
how different functions under decentralised reforms have study, the level of decision space available in practice at the 
been decentralised.15 The study reported that decisions various management levels of the GHS is determined by 
regarding established post, selection, salary determination, national policy directives as evidenced by the various human 
appointment, training, promotion and discipline are vested resource policies reviewed in this study. These directives 
in a centralised public service commission. The consequence prescribe the autonomy and discretionary levels regarding 
of the above could include staff shortage and adverse each human resource function at every management level 
implications on productivity and the attainment of health which has informed current practice. Again, the justification 
sector objectives because it has been reported in other studies for the central control of employment in the health sector 
that, greater flexibility in hiring and firing at the local levels are is not far-fetched. According to the MoH,46 57.3% of total 
associated with increased efficiency and quality of services.16 health spending and 90% of the recurrent budget was used 
Furthermore, efficiency is likely to be adversely affected in the for personnel compensation for the year 2013. Therefore, 
GHS because there is evidence to suggest that, the ability of an attempt to maintain control of the health sector budget 
local managers to hire, fire and provide specific incentives to requires maintaining control over the human resource 
employees improves efficiency.43 budgets. A study by Bach47 equally cited the desire to maintain 
On the other hand, some benefits could be envisaged tight control of public sector wage bill as the rational for the 
from the centrally controlled hiring, firing and salary reluctance of government to delegate significant autonomy 
determination responsibilities. These could include providing for pay determination to lower organisation levels. 
the opportunity to centrally manage the public sector wage Further, in relation to personnel administration, the study 
bill, ensure equity in wage determination and equitable reported mixed levels of decision space and this could present 
distribution of workforce. significant and varying implications on human resource 
It is however worth indicating that, the centralised nature management within Ghana’s health sector. The evidence that 
36 International Journal of Health Policy and Management, 2019, 8(1), 28–39
Sumah and Baatiema
the regional level exhibits wide discretion and autonomy in their professional knowledge, skills, values and practices.53 
all aspects of personnel administration whilst, the district The level of discretion exercised at the local level on staff 
level only exercises wide discretion regarding performance capacity development is, therefore, an opportunity to tailor 
appraisal and supervision suggests an emphasis on output and personnel development to meet local preferences and the 
personnel control without equal attention on other personnel changing health needs of the communities. Therefore, the 
functions such as motivation and staff development. On the limited discretion regarding this functional area reported in 
other hand, given that part of the personnel administration this study could affect the performance of the health sector 
function is to manage the performance of staff through the and how responsive training and staff development programs 
provision of technical supervision and performance appraisal, could be to the respective local communities. As revealed 
the wide range of discretion exercised at this level as per this by this study, regional, as well as district health service, are 
study could be considered appropriate and likely to improve limited to nominating or recommending candidates for such 
the performance of the health sector. Several studies in opportunities. An important aspect of this moderate decision 
other jurisdictions support this position because, it has been space at both the regional and district level is that it ensures 
observed that, by simplifying the personnel administration that the inputs of the lower levels are considered in granting 
function through the reduction in bureaucracy, and less approval for all forms of personnel development programs. 
overload and congestion in the channels of communication It goes without saying however that, opportunities which are 
and administration, improvement in the quality and developed centrally may not be with recourse to local level 
quantity of public services is often realised.42,44,48-50 It is needs and preferences. The bureaucratic procedures involved 
worth noting however that, some studies have reported that could equally have an impact on the timeliness of such 
decentralisation in many situations confuses supervisory personnel development programs and how useful they may 
responsibilities, diminishes technical supervision capacity be to the health sector. 
and reduces supervisory visits.51,52 This study does not entirely We situate our discussion within some limitations of 
agree with this position as findings from the present study the study. Decentralisation encompasses 3 dimensions viz 
present a contrasting picture. The hierarchical structure of the decision space, individual and institutional capacities and 
GHS as depicted in Figure equally defines accountability and local accountabilities. This study however focussed on 
responsibility relationships which to a large extent prevents only decision space notwithstanding the fact that studies 
the confusing roles in supervision and performance appraisal have underscored the utility of exploring the 3 dimensions 
as reported in other studies. together as a way of understanding the interaction among 
While the findings generally point to a wide discretion in the these dimensions.54,55 The implications of decentralisation on 
performance assessment such staff appraisal, same cannot be local capacities and accountability as a whole has therefore not 
said of the reward system. Ultimate authority for promotions been addressed by this study. The study is also descriptive and 
and demotion are mainly centralised. The regional and district as such does not provide a mechanism to measure the direct 
health services, as well as the hospitals’ roles in promotion, impact of the findings on human resource management. 
are largely advisory and subject to the approval of the central Additionally, capturing and categorising the exact range of 
authority. The seeming gap in the discretion and authority to choice available at the local management level is an inherent 
assess performance versus providing rewards (promotion/ limitation of the decision space approach. The study is also 
demotion) can adversely affect workforce performance. Based cross-sectional, and the results point to the period of research 
on this, it is reasoned that, once health personnel do not see in 1 region which could vary over time and across regions. 
a direct link between their performance and the consequent Further, this study was conducted in 1 of 10 regional health 
reward, they may not be motivated to improve productivity services and thus results should be interpreted and applied 
in the hope of a reward. Suffice it to state that, the discretion with caution as this does not represent the overall situation 
exercised at the regional and district levels regarding in Ghana. Nonetheless, we contend that the results from this 
promotion is significant because it forms the basis upon study offer useful insights towards healthcare sector reforms 
which decisions at the central national level regarding these to ensure greater decision space for managers to operate. We 
HR functions are made. It must also be indicated that, the level acknowledge this present study presents some strengths despite 
of decision space in practice for personnel administration the limitations. Foremost, it is the first of its kind to examine 
largely reflects national policy directives except for CHPS the effect of decentralisation reform on a specific functional 
zones where there is an indication of limited autonomy and area such as human resource management in the Ghanaian 
discretion for staff supervision and appraisal contrary to context. Secondly, the study disaggregates the human resource 
national level policy directives. Additionally, it is also evident function into specific functional areas and thereby provides a 
that, the decision space for personnel administration is a well-focussed analysis of the implications of decentralisation 
derivative of deconcentration which limits administrative on these specific human resource functional areas and the 
authority to national level directives. implications thereof on the entire health sector. Through a 
In the health sector, personnel development is an comprehensive document review, the study also provides 
important function because it provides an opportunity for relevant context to responses obtained from the survey and 
health workers to provide quality healthcare to meet the thereby informs an in-depth analysis and discussion on the 
communities’ changing healthcare needs through updating subject. As such, the findings and conclusions provide further 
International Journal of Health Policy and Management, 2019, 8(1), 28–39 37
Sumah and Baatiema
insights on the effect of decentralisation reform in the health and from the various district and medical directors/superintendents of health 
sector for both policy purposes and the research community. before the study was conducted.
Competing interests 
Conclusion The authors declare they have no competing interest. 
This study concludes that though decentralisation presupposes 
the transfer of responsibility and authority for the management Authors’ contributions 
The study was designed by both authors. AMS analysed, interpreted and wrote 
of human functions from the central national level to the the first draft. LB contributed to the interpretation and writing of the manuscript. 
respective sub-national management levels, findings from Both authors reviewed, read, and approved the final manuscript.
the study point to the contrary as the decentralisation of 
the human resource function is more apparent than real. Authors’ affiliations
1
Management authority regarding core human resource Ghana Health Service, Upper West Regional Health Directorate, Wa, Ghana. 2Regional Institute for Population Studies, University of Ghana, Legon-Accra, 
functions are centralised at the GHS national level and as Ghana. 3School of Allied and Public Health, Faculty of Health Sciences, 
such, provides very limited discretion and autonomy at the Australian Catholic University, Sydney, NSW, Australia.
subnational levels. This situation is understood to be a direct 
outcome of the form of decentralisation (deconcentration) Supplementary files
Supplementary file 1 contains Table S1.
adopted in the health sector in Ghana. Deconcentration as 
a form of administrative decentralisation relies on national References 
policy directives to determine the decision space available 1. Liu X, Martineau T, Chen L, Zhan S, Tang S. Does decentralisation 
at subnational levels which is largely the practice in Ghana’s improve human resource management in the health sector? A case study from China. Soc Sci Med. 2006;63(7):1836-1845. 
health sector. We also conclude that, management discretion doi:10.1016/j.socscimed.2006.05.011
for the management of human resources diminishes along the 2. Sohag A, Wajidi FA, Miankhel AK. Impact of decentralization on 
management continuum from the national to the community the effectiveness of human resource management in health sector. 
level. Further, the study also revealed an apparent disjoint Gomal University Journal of Research. 2013;29(1):56-61.
3. Pokharel B. Decentralization of health services. New Delhi: World 
between the wide decision space for performance assessment Health Organization-Regional Office for South East Asia; 2000.
through technical supervision and performance appraisal 4. Bossert TJ, Larranaga O, Giedion U, Arbelaez JJ, Bowser DM. 
and a rather limited discretion for providing incentives or Decentralization and equity of resource allocation: evidence from 
rewards. Decentralisation was however found to have led to Colombia and Chile. Bull World Health Organ. 2003;81(2):95-100.
5. Bossert TJ, Beauvais JC. Decentralization of health systems 
greater autonomy and discretion in performance assessment in Ghana, Zambia, Uganda and the Philippines: a comparative 
and personnel administration in the GHS. It was also found analysis of decision space. Health Policy Plan. 2002;17(1):14-31.
to provide the potential for improved salary administration 6. Atkinson S, Haran D. Back to basics: does decentralization improve 
and equity in human resource distribution. To realise the health system performance? Evidence from Ceara in north-east 
benefits of decentralisation in respect of human resource Brazil. Bull World Health Organ. 2004;82(11):822-827.7. Jeppsson A, Okuonzi SA. Vertical or holistic decentralization of the 
management for health, there is the need to give further health sector? Experiences from Zambia and Uganda. Int J Health 
meaning to decentralisation in the GHS by increasing the Plann Manage. 2000;15(4):273-289. doi:10.1002/hpm.597
decision space in practice at the subnational level. This will 8. Sumah AM, Baatiema L, Abimbola S. The impacts of 
require a revision of ACT 525 which provides the legal basis decentralisation on health-related equity: A systematic review of the evidence. Health Policy. 2016;120(10):1183-1192. doi:10.1016/j.
for the form of decentralisation practiced in the health sector. healthpol.2016.09.003
Human resource policy revision is also required in the GHS 9. Bossert T, Chitah MB, Bowser D. Decentralization in Zambia: 
to empower local levels with the relevant decision-making resource allocation and district performance. Health Policy Plan. 
autonomy and discretion as well as capacity to manage 2003;18(4):357-369.10. Diana A, Hollingworth SA, Marks GC. Effects of decentralisation 
human resource for health. We conclude by recommending and health system reform on health workforce and quality-of-care in 
that further studies consider a framework for measuring the Indonesia, 1993-2007. Int J Health Plann Manage. 2015;30(1):E16-
direct effects of decentralisation and should include other 30. doi:10.1002/hpm.2255
functional areas beyond the human resource management 11. Heywood PF, Harahap NP. Human resources for health at the district level in Indonesia: the smoke and mirrors of decentralization. 
function. Most importantly, we recommend that further Hum Resour Health. 2009;7:6. doi:10.1186/1478-4491-7-6
studies should assess all 3 dimensions of decentralisation 12. Pierantoni CR, Garcia AC. Human resources for health and 
namely decision space, individual and institutional capacities decentralization policy in the Brazilian health system. Hum Resour 
and local accountability together to provide an overall insight Health. 2011;9:12. doi:10.1186/1478-4491-9-12
13. Sakyi EK, Awoonor-Williams JK, Adzei FA. Barriers to implementing 
into the implications of decentralisation on human resource health sector administrative decentralisation in Ghana: a study of 
as well as other management functional areas. the Nkwanta district health management team. J Health Organ 
Manag. 2011;25(4):400-419. doi:10.1108/14777261111155038
14. Abimbola S, Olanipekun T, Igbokwe U, et al. How decentralisation 
Acknowledgements influences the retention of primary health care workers in rural 
We acknowledge Dr. Insah Baba, School of Business, Wa Nigeria. Glob Health Action. 2015;8:26616. doi:10.3402/gha.
Polytechnic, Wa, Ghana for his significant inputs into the v8.26616
study design. 15. Mohammed J, North N, Ashton T. Decentralisation of Health 
Services in Fiji: A Decision Space Analysis. Int J Health Policy 
Manag. 2015;5(3):173-181. doi:10.15171/ijhpm.2015.199
Ethical issues 16. Bossert T. Analyzing the decentralization of health systems in 
Written approval was sought from Upper West Regional Health Administration 
38 International Journal of Health Policy and Management, 2019, 8(1), 28–39
Sumah and Baatiema
developing countries: decision space, innovation and performance. 37. Ghana Health Service. Policy and Guidelines on Counselling. Accra, 
Soc Sci Med. 1998;47(10):1513-1527. Ghana: Ghana Health Service, Directorate HRD; 2007D.
17. Holtzappel CJG, Ramstedt M. Decentralization and Regional 38. Ghana Health Service. Code of Conduct and Disciplinary 
Autonomy in Indonesia: Implementation and Challenges. Institute of Procedures. Accra, Ghana: GHS; 2003.
Southeast Asian Studies; 2009. 39. Denkyira A. Guidelines for the Award of Study Leave in 2016. Accra, 
18. Hunter J. World Bank, World Development Report 1993: Investing in Ghana: Ghana Health Service; 2016.
Health. Annals-Association of American Geographers. 1995;85:774- 40. Ghana Health Service. Human Resource Development Directorate, 
775. Employee Handbook. Accra: 2006.
19. Sumah AM, Bowan PA, Insah B. Decentralization in the Ghana 41. Braathen E, Chaligha A, Fjeldstad OH. Local governance, finances 
Health Service: A Study of the Upper West Region. Developing and service delivery in Tanzania [Workshop]. Dar es Salaam; March 
Country Studies. 2014;4(12):45-52. 2004.
20. Rondinelli DA, Nellis JR, Cheema GS. Decentralization in developing 42. Gilson L, Kilima P, Tanner M. Local government decentralization 
countries: a review of recent experience. World Bank Staff Working and the health sector in Tanzania. Public Administration and 
Papers. 1983:581. Development. 1994;14(5):451-477. doi:10.1002/pad.4230140503
21. Hutchinson PL, LaFond AK. Monitoring and evaluation of 43. Chia YM. The Interaction Effect of Information Asymmetry and 
decentralization reforms in developing country health sectors. Decentralization on Managers’ Job Satisfaction: A Research Note. 
Bethesda, MD: The Partners for Health Reformplus Project, Abt Hum Relat. 1995;48(6):609-624. doi:10.1177/001872679504800601
Associates Inc; 2004. 44. Wang Y, Collins C, Tang S, Martineau T. Health systems 
22. Egbenya GRK. The effectiveness of decentralization policy in decentralization and human resources management in low and 
Ghana: A case study of Komenda-Edina-Eguafo-Abrim (KEEA) and middle income countries. Public Adm Dev. 2002;22(5):439-453. 
Abura-Asebu-Kwamankese (AAK) districts in Ghana. Afr J Pol Sci doi:10.1002/pad.246
Int Relat. 2010;4(1):13-28. 45. Falleti TG. A Sequential Theory of Decentralization: Latin 
23. Vrangbaek K. Towards a typology for decentralization in health care. American Cases in Comparative Perspective. Am Polit Sci Rev. 
In: Saltman RB, Bankauskaite V, Vrangbaek K, eds. Decentralization 2005;99(3):327-346. doi:10.1017/S0003055405051695
in health care: strategies and outcomes. New York: McGraw-Hill; 46. Ministry of Health. Holistic assessment of the health sector program 
2007:44-60. of work 2013. Accra, Ghana: Ministry of Health; 2014.
24. Kwamie A, van Dijk H, Ansah EK, Agyepong IA. The path 47. Bach S. Decentralization and privatization in municipal services: 
dependence of district manager decision-space in Ghana. Health The case of health services. Sectoral Activities Programme Working 
Policy Plan. 2016;31(3):356-366. doi:10.1093/heapol/czv069 Paper; 2000:164.
25. Bossert TJ, Mitchell AD. Health sector decentralization and 48. Munga MA, Songstad NG, Blystad A, Maestad O. The 
local decision-making: Decision space, institutional capacities decentralisation-centralisation dilemma: recruitment and distribution 
and accountability in Pakistan. Soc Sci Med. 2011;72(1):39-48. of health workers in remote districts of Tanzania. BMC Int Health 
doi:10.1016/j.socscimed.2010.10.019 Hum Rights. 2009;9:9. doi:10.1186/1472-698x-9-9
26. Dussault G. Working in health: financing and managing the public 49. Mills A. Decentralization and accountability in the health sector from 
sector health workforce. Bull World Health Organ. 2010;88(5):400. an international perspective: What are the choices? Public Adm 
doi:10.2471/BLT.10.076190 Dev. 1994;14(3):281-292. doi:10.1002/pad.4230140305
27. Doyle L, Brady AM, Byrne G. An overview of mixed methods research. 50. Gilson L, Travis P. Health systems decentralisation in Africa: an 
J Res Nurs. 2009;14(2):175-185.   doi:10.1177/1744987108093962 overview of experiences in 8 countries. Background document 
28. Creswell JW, Tashakkori A. Developing Publishable Mixed Methods prepared for regional meeting on decentralisation in the context of 
Manuscripts. Los Angeles, CA: Sage Publications Sage CA; 2007. health sector reform in Africa. Geneva: World Health Organization; 
29. Greene JC. Is Mixed Methods Social Inquiry a Distinctive Methodology? 1997.
J Mix Methods Res. 2008;2(1):7-22. doi:10.1177/1558689807309969 51. Saide MA, Stewart DE. Decentralization and human resource 
30. Bowling A. Research Methods in Health: Investigating Health and management in the health sector: a case study (1996-1998) from 
Health Services. 4th ed. England: McGraw-Hill Education; 2014:538. Nampula province, Mozambique. Int J Health Plann Manage. 
31. World Medical Association Declaration of Helsinki: ethical 2001;16(2):155-168. doi:10.1002/hpm.620
principles for medical research involving human subjects. JAMA. 52. Campos-Outcalt D, Kewa K, Thomason J. Decentralization of health 
2013;310(20):2191-2194. doi:10.1001/jama.2013.281053 services in Western Highlands Province, Papua New Guinea: an 
32. Ghana Health Service. Policy and Guidelines on Appointments. attempt to administer health service at the subdistrict level. Soc Sci 
Accra, Ghana: Ghana Health Service, Directorate HRD; 2007A. Med. 1995;40(8):1091-1098.
33. Republic of Ghana, Fair Wages and Salaries Commission, Act 207 53. Giri K, Frankel N, Tulenko K, Puckett A, Bailey R, Ross H. Keeping 
(ACT 737). Accra, Ghana: The Parliament of the Republic of Ghana; up to date: continuing professional development for health workers 
2007. in developing countries. IntraHealth International; 2012.
34. Ghana Health Service. Policy and Guidelines on Promotions. Accra, 54. Bossert TJ, Mitchell AD, Janjua MA. Improving Health System 
Ghana: Ghana Health Service, Directorate HRD; 2007B. Performance in a Decentralized Health System: Capacity Building 
35. Ghana Health Service. Policy and Guidelines on Postings. Accra: in Pakistan. Health Syst Reform. 2015;1(4):276-284. doi:10.1080/2
Ghana Health Service, Directorate HRD; 2007C. 3288604.2015.1056330
36. Kwamie A, Asiamah M, Schaaf M, Agyepong IA. Postings and 55. Bossert TJ. Empirical Studies of an Approach to Decentralization: 
transfers in the Ghanaian health system: a study of health workforce “Decision Space” in Decentralized Health Systems. Public Adm 
governance. Int J Equity Health. 2017;16(1):85. doi:10.1186/ Dev. 2014;26(4):303-315.
s12939-017-0583-1
International Journal of Health Policy and Management, 2019, 8(1), 28–39 39