2
DRAFT 2 (Not for citation)
Please send comments to:
Comparative Experience with Administrative Reform
in Ghana, Tanzania and Zambia
July 28, 2003
Prepared by:
Mike Stevens and Stefanie Teggemann
Public Sector Reform and Capacity Building Unit
Africa Region, The World Bank
2
List of Abbreviations
AO / Accounting OfficeAPL / Adaptable Program Lending
BRELA / Business Registration and Licensing Authority (Tanzania)
CSPIP / Civil Service Performance Improvement Program (Ghana)
CSR / Civil Service Reform
CSRP / Civil Service Reform Program
CSC / Client Service Charter (Tanzania)
CSD / Civil Service Department (Tanzania)
CSO / Civil Society Organization
DB / Dual budget
GDP / Gross Domestic Product
GES / Ghana Education Service
GIMPA / Ghana Institute of Public Administration
GoG / Government of Ghana
IFI / International Financial Institution
IFMIS / Integrated Financial management Information System
MDA / Ministry, Department and Agency
M&E / Monitoring & Evaluation
MMD / Movement for Multi-Party Democracy (Zambia)
MTFF / Medium Term Financial Framework
MTEF / Medium Term Expenditure Framework
MTPRS / Medium Term Pay Reform Strategy
NDC / National Democratic Congress (Ghana)
NDP / National Development Plan
NGO / Non-Governmental Organization
NIPA / National Institute of Public Administration (Zambia)
NIRP / National Institutional Reform Program (Ghana)
NPM / New Public Management
OECD / Organization of Economic Cooperation and Development
OHCS / Office of the Head of Service (Ghana)
OPRAS / Open Performance Appraisal System (Tanzania)
PAC / Public Accounts Committee
PIF / Performance Improvement Fund
PIM / Performance Improvement Model (Tanzania)
PIP / Performance Improvement Plan
PS / Permanent Secretary
PSC / Public Service Commission
PSCAP / Public Sector Capacity Building Program (Zambia)
PSIP / Public Sector Investment Program
PSR / Public Service Reform
PSRP / Public Service Reform Program
SA / Subvented Agency
SASE / Selected Accelerated Salary Enhancement
SOE / State Owned Enterprise
SSA / Sub Saharan Africa
TPSC / Tanzania Civil Service College
TEVETA / Zambia Vocational Training Institution
ZAWA / Zambia Wildlife Authority
2
TABLE OF CONTENTS
I. Introduction 2
II. Setting the Scene 2
III. First Generation Civil Service Reform Programs 5
IV. Second Generation Reform Programs 6
V. The Enabling Environment: Political and Economic Realities and the Leadership of Reform 9
a) Political Commitment 9
b) Political Context 9
c) The Economic Reality 11
d) Leadership for Reform 11
e) Implementation Arrangements 12
VI. Second Generation Reforms in Ghana, Tanzania and Zambia: Some Issues of Design and Implementation 13
a) Pay 14
b) Selected Accelerated Salary Enhancement Scheme (SASE) 15
c) State Restructuring 16
d) Performance 18
e) Control Systems and Accountability 20
Expenditure Controls 21
Accountability through Civil Society Participation and Monitoring 23
f) Training 25
VII. Conclusions 27
ANNEX: Five Building Blocks of Public Service Reform Projects in Tanzania, Ghana and Zambia 28
Bibliography 31
32
COMPARATIVE EXPERIENCE WITH ADMINISTRATIVE REFORM
IN GHANA, TANZANIA AND ZAMBIA
I. Introduction
This paper examines the public service reform experience of three Anglophone African countries, which have embarked upon what might be termed “second generation” public service reform programs. Each one is distinctive, yet all three have common threads running through them. Performance has varied with country, and while they have attempted to implement many of the same things, the results have been quite different. But the reforms in these countries are ongoing, and lessons are still being learned. This paper is an initial effort to draw out some of the most important features of these country programs, hold them up to examination, and see what conclusions, even if tentative, can be drawn.
The paper starts with a brief recapitulation of the story of the public service in Anglophone Africa. This began with independence and the adaptation of the colonially inherited administration to the political and social demands for development. Public services experienced rapid expansion during much of the first two decades, until growth was checked by economic crisis, in the course of which the mismatch of resources numbers and pay, and the informal compensation systems that arose in the wake of sharp real pay falls, brought many public services close to dysfunction. The legacy of this collapse provides the backdrop to present reforms. The first reforms came in the wake of IFI supported structural adjustment programs, and emphasized wage bill containment. These are now termed “first generation” reform programs. They were broadly successful in what they set out to achieve, but much more needed to be done to restore well performing government. The first reforms were followed by “second generation” reform programs which continued the stress on pay reform, but also emphasized the restructuring of government and the linkages with other control systems of government. They also featured specific performance improvement schemes, drawing in various ways on the New Public Management (NPM) reform experience in OECD countries. The current reform programs of Ghana, Tanzania and Zambia exemplify this approach, and the paper identifies key features of these countries’ reform programs and explores them in greater depth, and makes comparisons. The paper concludes with some general observations on what policy makers should pay attention to in the future.
II. Setting the Scene
The three countries that are the subject of this paper came to independence between the late 1950’s and the early 1960s. Ghana, the first to gain independence, had a significant number of Ghanaians in senior professional and managerial positions in the public service, which was regarded as one of the best in Africa, with a long tradition of local professionals in public positions. By contrast, both Tanzania and Zambia had very few indigenous graduates available to fill civil service positions at independence. In both countries, expatriates dominated senior positions, and, especially in Tanzania, local Asians occupied many middle level technical positions. Very few Africans had made it to the officer level in either civil service prior to independence. In all countries, the public service was much smaller that it was later to become.
In one sense the nomenclature of this paper, the identification of first and second generation public service reform programs is historically inaccurate, for there were important measures taken in the post-Independence period to change the public service. In the early years of independence, both in Ghana, Tanzania and Zambia, and in other similar countries, the focus of public service policy was localization (or “Africanization”). Much effort by governments and donors was invested in building of indigenous capacity through the creation of public service training institutions, the training of key staff overseas, and the establishment of local universities. At the same time, the rigid class structure of the colonial public service was abolished, and cadres were introduced to provide a career structure for staff advancement and training. Pay scales were adjusted to remove colonial anomalies. In some countries the powers of the Public Service Commission were deliberately reduced (in Tanzania the PSC was abolished), and Public Service Ministries were created to implement training and localization policies more vigorously. These were the true “first generation” reforms, the traces of which can be found in all Anglophone countries, and they were successfully implemented.
During the first decade of independence (and even into the second) graduates entering the public service received good pay, and were assured of a car loan and access to government pool housing at a low and fixed proportion of salary. Most of these officers felt they had taken the first step up a ladder of promotion to status, authority and an assured standard of living. Officers retiring from public service now look wistfully back to those days as they contemplate a pension far less in purchasing power than anticipated at entry. At the lower end of the pay spectrum, daily paid unskilled workers earned slightly better than a subsistence wage. While we might nowadays regard the pay scales earned by the first generation of senior civil servants as overly influenced by living standards of the departed colonial administrators, there was no doubt that the pay structures in place then were sufficient to recruit, retain and motivate the best brains in the country for public service.
Figure 1
Source: Valentine, 2002, p. 4. Based on data provided by Civil Service Department and Valentine, 1999.
There was, nevertheless, a recognition that pay in government had to be proportionate with what was happening elsewhere in the economy, and there were equity as well as efficiency issues to be concerned about. Government scales in all three countries served as the basis for a national incomes policy in the late 1960s and into the 1970s. The aim at the time was to integrate pay scales across the public sector, so that they could be a guide to the private sector, at least the modern one, and could facilitate interchange of staff between central government and parastatals, and with local government. By the early 1970s pay was still good for the majority of civil servants, but in several countries (Tanzania especially) there was a deliberate compression of scales in pursuit of social equity goals. In all three countries the principal mechanism for pay setting was a government established salaries commission. Industrial class pay grades, however, tended to be set by wages councils, and in some countries (Zambia in particular, with its strong history of trade unionism), collective bargaining played a central role.
Real salaries broadly held up until the oil price shocks of the early to mid-1970s, and somewhat longer in a country like Zambia which enjoyed an offsetting but temporary mineral boom. Pay, however, came under increasing strain in the mid to late 1970s, as economies slowed, revenues faltered, and staffing numbers continued to rise, driven by expanding social services, and a paradigm of development that envisaged government playing the dominant role in many sectors (see figure 1). And instead of decentralizing functions or contracting them out, governments opted for providing services directly and centrally.
Data of public service pay, particularly for professional and senior administrative staff, and total public service employment, are notoriously difficult to compare across countries on a common basis.[1] But when this is done and shown on the same graph, for almost all countries (Botswana and South Africa are the only exceptions, in Nigeria the full impact came later) show a sharply divergent trend, from the early 1970s to the mid 1980s, with total numbers, driven by policies of expanding health care and education and the automatic employment of graduates from tertiary training institutions, continuing to grow each year at between 5-8% in many countries. Real pay was eroded by high inflation, stagnant revenues and infrequent pay awards, which, when given, were biased towards the lower scales. There was some cushioning of the pay decline initially, as overvalued exchange rates protected staff from the true erosion of purchasing power, but once the structural adjustment programs of the 1980s had begun to liberalize trade regimes, abolish credit rationing and price controls, and bring in more realistic exchange rates (all of which were necessary if economic growth was to resume), the full impact of these divergent trends was felt on official salaries. Increments within salary scales, available supposedly to reward performance, became too small to mean anything. To some extent governments were able to offset these falls by selective allowances. In real terms pay in many countries fell to about one tenth of its original levels for professional, senior technical and administrative grades. At the level of junior and daily paid staff, what had once been an acceptable minimum wage became insufficient to feed a family[2].
This collapse of pay destroyed the employment contract between government and its public service employees. Although the public service outwardly maintained many of its formal routines, informality mushroomed, and performance plummeted. Ordinary civil servants in all three countries (and elsewhere) adjusted for the collapse of the formal reward structure by a series of survival strategies. These ranged from taking second jobs, within and outside office hours, manipulation of travel allowances and per diems, extracting private fees for public services, commissions on office procurement contracts, to outright theft of public assets. Professionals with scarce skills left to work in the private sector or even overseas. Donors seeking to maintain operational effectiveness of their projects, paid supplements, often large ones. These had the effect of drawing skilled staff out of departmental structures into project units, further undermining core administrative capacity. Aid dependence increased and the management of aid programs moved in the hands of donor agencies. Training budgets evaporated, training institutions atrophied and trainers lost skills and relevance. Systemic corruption became ingrained. The formal rules became in many public services a shell in which huge informality was practiced. Public interest and the merit principle were relegated to second place.[3]
These deleterious effects were compounded by the pressure on the non-wage components of budgets. With reduced operations and maintenance allocations, civil servants were not equipped with the materials to be effective, an effect compounded by the unpredictability of the cash release system for budget funds.
III. First Generation Civil Service Reform Programs
While many of the structural adjustment programs countries adopted at this time contained programs for the rationalization of the public enterprise sector, through closures and privatization, it was not until the mid-1980’s that countries and external donors realized than a crucial governance component has been left out of economic adjustment programs, and a start was made to fill this gap. The first CSRPs supported by the Bank were in Gambia, Ghana and Guinea. Ghana began its CSR in the mid-1980s. Tanzania followed suit in the early 1990s, and Zambia a little later.
Many nowadays criticize these early CSRPs as overly technocratic, dominated by the IMF’s demand to control the wage bill, and paying too little attention to the broader governance environment. This is an unfair charge bearing in mind that the international community generally had failed to appreciate the extent to which the institutions of governance had declined in poor countries, and governments themselves were in no position to articulate more comprehensive public service reform programs. Yet undeniably they were trying to address a central cause of the decline in public services, through pay and employment reform. Thus the main thrust of these early programs was wage bill containment, through functional reviews, public service censuses to identify “ghosts” (i.e. payroll fraud), the restoration of payroll control systems, and the separation of staff through voluntary departure and mandatory layoffs. The automatic acceptance of graduates into public employment was ended, recruitment was frozen, and over-age employees were dismissed. To mitigate the social hardship of layoffs, there were special training programs, and resettlement schemes, and financial compensation was paid. Tracer studies were undertaken to ascertain how retrenchees (to the extent to which governments were actually able to lay off actual as opposed to fictitious staff) fared after departure from government employment (some went farming, many eschewed special training programs and settlement schemes and continued to live in the same location as when they were in government employment, essentially making their second job now the primary one or taking public sector jobs outside the civil service proper). Efforts were made to curtail the proliferation of allowances, incorporate them into pay, and raise, to the extent possible, total remuneration. And attempts were made to strengthen establishment and payroll controls, to prevent the annual accretion of staff, and the proliferation of payroll fraud.