III. Civil Service Reform: Progress and Prospects13
54. Civil service reform is critical to Cambodia’s development. Progress was slow during most of the decade, but initial and concrete steps have been taken during the past three years. Furthermore, the Royal Government of Cambodia (RGC) is formulating a medium-term reform strategy that it believes will improve administrative performance and make civil administration an effective development partner of the private sector.
55. This chapter reviews progress in reforming civil administration14 and military demobilization as downsizing of the defense and security sector is part of the broader public service reform agenda. The chapter then turns to prospects for improving public sector pay, regarded by the government and its external partners as a matter of urgency for improving public service. However, raising average pay without endangering macroeconomic stability—to which the government is openly committed—will be a major challenge for the authorities.
A. Public Sector Employment and Wages in Cambodia
56. Cambodia’s public sector largely reflects the consequences of the country’s recent history. Several decades of internal military confrontation swelled the defense and security ranks. For example, both the Paris Peace Accord of 1991 and the surrender of remaining Khmer Rouge soldiers in 1998 provided for the integration of large numbers of soldiers from the various factions in the Royal Cambodia Armed Forces. Although downsizing is underway, the defense and security establishments remain large employers, with a payroll covering 129,000 persons (at end-2001) and amounting to about 2 percent of GDP.15 The civil administration payroll covered 163,500 persons at end-2002, also amounting to about 2 percent of GDP in 2002.
Climboo.ia: Composition of the Wage Bill
Source: Ministry of Economy and Finance
57. At first glance, overall public sector employment (defense and security plus civil administration) in Cambodia is well within international norms. With 2.3 percent of the population employed in the public sector, Cambodia’s public sector employment is in the middle range for a large number of developing countries for which data are available.16 At about 1.3 percent of the population, the civil administration workforce is about half the unweighted average for Asian countries, and lower than most developing countries. Similarly, the overall wage bill remains low as a percent of GDP, and compares favorably against an average of 5½ percent among a sample of developing countries and 4.5 percent among developed countries.
Cambodja: Government Employment in Cambodia, 2002.
Sources: Ministry of Defense, and Ministry of Economy and Finance.
Cambodia: Low income economies: public sector Wage bill, 1995.
Source: Governmont Finance Statistics
58. At the same time, the exceedingly low level of official public sector wages—currently averaging $28 per month—is a serious impediment to development. Public sector wages that are totally out of line with pay for comparable work in the private sector create perverse incentives, reduce morale, induce widespread absenteeism, result in poor service delivery, prevent the recruitment and retention of qualified personnel and, in general, weaken governance.17 The government’s medium-term strategy comprises a two-pronged approach to improving public sector pay without endangering macroeconomic stability. The adequacy of these plans is assessed in a later section of this chapter.
B. Track Record since the Mid-1990s
Difficult initial conditions
59. Progress in reforming the civil service was relatively slow during the 1995-1999 period. The civil service and defense and security establishments were overstaffed and underpaid as the reconstruction and rehabilitation of Cambodia got underway in the early 1990s after the Paris Peace Accord of 1991. During the early part of the first PRGF (approved in May 1994), some progress was achieved: the automatic recruitment of high school graduates was ended; a probationary period was introduced for new graduates taken into the civil service; and new rules for entry into public employment were established and institutionalized.
60. In early 1995, the government committed to maintain the number of civil servants at the level determined by a preliminary census. According to a full census completed in 1995, there were about 153,000 on the civil service payroll, plus another 21,000 employed in the provincial and communal administrations. At that time, the average monthly wage was estimated to be around $20. The payroll for defense and security in late 1996 was estimated to cover roughly 250,000 persons, earning comparable pay. Under the program supported by the first PRGF, the government had committed to reduce the number of civil servants by 10 percent per year in 1996 and 1997. However, large scale hiring for political reasons during the mid-1990s put these targets for net reductions out of reach.
61. Several factors contributed to the inertia in reforming the civil service at that time. First, economic uncertainties, the absence of any formal social safety net, and slow private sector job creation meant that recycling dismissed civil servants and demobilized soldiers was not assured. Second, hiring for political reasons clearly complicated significantly the task of rationalizing the civil service. Third, more technical support and external financing were needed to help design and implement a socially acceptable retrenchment program.
62. Only limited progress was achieved in reforming the military. The number of generals was reduced from 1,876 to 147, and the number of colonels from 5,000 to 304, while the number of divisions was reduced from 28 to 12. Sizable cuts in the number of paid soldiers would not be achievable until later, however, when external financial assistance was secured.
An acceleration of reform since 1999
63. Reforms were inevitably stalled during the political instability that followed internal military confrontation in 1997. The government was successful in avoiding expansions in civil service employment levels, however, and the pace of reform picked up with the restoration of peace since 1998. At the Tokyo Consultative Group (CG) meeting in 1999, the Royal Government of Cambodia (RGC) presented a draft pilot project for military demobilization and a three-stage medium-term National Program for Administrative Reform (NPAR).
64. The military demobilization program agreed at the Tokyo CG consists of four stages: registration of soldiers; discharge and disarmament; re-insertion; and re-integration. The identification and registration process to establish the size of the army was completed in October 1999, following which 15,551 nonexistent (ghost) soldiers and 163,346 nonexistent children (claimed as dependents of soldiers) were removed from the payroll.
65. Physical demobilization would be carried out in two phases: a pilot followed by a broader and bolder dismissal of soldiers. Under the pilot program, 1,500 soldiers in 4 provinces were discharged during May-July 2000. Demobilization and re-integration of an additional 30,000 soldiers was estimated to cost US$42 million, to be financed by the government (US$8 million), and substantial external assistance, including: a credit of US$18.4 million from the World Bank, and grants from Japan (US$10 million), Sweden (US$2.4 million), the Netherlands (US$2 million), and the World Food Program (US$2 million).
66. The discharge of the first 15,000 soldiers in 10 provincial centers was completed in December 2001. All of the demobilized soldiers received a cash payment of US$ 240, and food assistance was provided by the World Food Program in mid-2002. The distribution of reintegration packages (including the choice of a motorcycle, water pump, generator or sewing machine) is underway and expected to be completed by February 2003.
67. Demobilization of the next 15,000 soldiers is underway. The computerized database of military personnel created in 1999 is being updated in consultation with stakeholders to verify those eligible for discharge. The next wave of demobilization should be completed in late 2003. Further reforms to defense and security employment and pay arc currently being considered in the context of broader reforms to military service under the direction of the Ministry of Defense.
Civil administration reform: institutional changes
68. The NPAR is the RGC’s medium-term strategy consisting of three stages:(i) consolidation of public services; (ii) restructuring and redeployment; and (iii) rationalization and capacity building.18 The first stage of the NPAR was largely completed by end-2001, with the exception of a functional analysis, now planned for the first half of 2003. The principal objectives of this stage were to: (i) document and control the composition and distribution of the workforce; (ii) develop essential instruments to manage and motivate personnel; (iii) assess the needs of ministries relating to corporate services; (iv) complete preparatory work to move the administration closer to citizens; and strengthen the capacity to plan and (v) manage the reform. Achievements thus far under the first phase include:
Completion of a database and identification of 9,000 ghost employees;
Automation of the payroll;
Introduction of a new employee classification system and salary grid;
Design of Priority Mission Groups (PMG) and adoption of an implementing legal framework;
Development and introduction of a Human Resource Management Information System;
Launch of the Economic and Public Service Capacity Building Project;
Initiation of a variety of initiatives to improve service delivery (e.g., studies on approaches to de-concentration, reviews of “back-office” processes, establishment of working relations with donor working group).
69. Many of these measures have begun to bear fruit or should begin to do so in the near future. Improved information on the size and composition of the civil service has made possible improvements in management of the civil service workforce. The authorities report that, together with savings from the removal of ghost workers from the payroll, better management of the overall civil service has generated savings that enabled the government to begin to raise salaries in 2002.19
Civil administration: pay reform
70. Pay reform has only recently begun to be implemented. Wages in the civil service consist of both base salary and allowances, the former accounting for about three fifths of the average wage. There are several types of allowances: functional, pedagogical, family and risk, the first three of which are the most important. Civil service wages remained virtually unchanged during the second half of the 1990s. Beginning in 2002, average civil service pay was increased by roughly 38 percent in January from levels in October 2001, raising average monthly pay to $27. This was followed by a further increase to $28 per month, bringing the total increase to 44 percent (representing an increase of 35 percent of the total wage bill).
71. Pay increases during 2002 represent the first phase and one element of a two-pronged approach to pay reform. The first consists of gradually raising civil service pay while also decompressing the wage structure. Pay increases for civil servants are to be kept in line with increases in the means available to the government. The strategy calls for members of top management positions to receive special allowances of S200 a month. The strategy for the remaining civil servants consists of ensuring that their pay is adequate to meet living costs. Currently, the government’s medium-term wage projection—which assumes a restructured (in favor of education) civil service of the same size as at present20—targets increases in the average civil service pay to $52 per month in 2006.21 The authorities recognize that these levels of pay alone are not likely to be competitive with the private sector, notably in urban centers. In order to make more informed decisions regarding competitive pay levels, the government will undertake a labor market study during the first half of 2003, with World Bank assistance. Implementing this program, however, will pose major macroeconomic challenges to the government as it is dependent on meeting ambitious targets for fiscal revenue (see Chapter II), and may need to be adjusted in view of major uncertainties regarding pending reforms in the pay structure for defense and security services as well as significant future debt service obligations to bilateral creditors that will arise from the completion of outstanding rescheduling agreements.
72. Priority Mission Groups (PMGs), to be launched in February 2003, comprise the other, parallel, prong of the reform strategy. PMGs are similar to existing donor-financed pay supplements that arc widespread throughout the government. PMGs will consist of small groups of civil servants tasked to selected high priority “missions” of limited duration (generally 12-18 months). Staff selected for PMGs are to receive monthly pay supplements ranging initially between CR 190,000 and CR 520,000 ($48 and $130 at an exchange rate of CR 3,925/USS). These “agents of change” arc seen by the RGC as a key axis of the overall reform agenda, and arc expected to promote higher productivity and initiative, transparency and accountability in public administration.22 The initial plan is to have 1,000 PMG members by the end of 2003, increasing by 500 in each year to reach a targeted 2,500 (1½ percent of the projected civil administration workforce) by end 2006. The 2003 budget has allocated adequate funds to pay performance allowances for the first 1,000 PMGs, and the government is working with donors to mobilize external resources to complement the government’s own funds. The annual costs of PMGs are projected in the MTWF to grow from US$2.0 million in 2003 to US$3.4 million by 2006 for a total cost of US$12.8 million.
73. PMGs are designed to address service delivery bottlenecks that arise due to inadequate pay incentives. Indeed, for high priority “missions” to be successful, appropriate incentives need to be provided to secure the dedication and commitment of civil servants assigned to those missions. At the same time, a few weaknesses of the concept need to be kept in mind. First, as short-term (18-month) assignments, PMGs do not resolve the most fundamental problem plaguing the Cambodian civil service: widespread inadequate pay for essential personnel. Second, it will be a challenge to ensure that the process of selecting members of PMGs is transparent. Third, and related to the short-term nature of the assignments, there is a risk that the extremely small group of civil servants participating in the initial wave of PMGs will become entrenched “elite”, moving from one “mission” to another. Questions about the transparency and fairness of the system may in turn become prevalent. It is therefore urgent that a strategy be devised for PMGs and the broader reform program to converge over the medium term.
C. Addressing the Inadequacy of Pay over the Medium Term
74. Improving public sector pay is clearly an urgent priority. At the same time, improving remuneration takes time, and requires a clear strategy for financing the increases. This section considers several options for improving public sector pay that are consistent with the medium-term macroeconomic framework that underpins the NPRS, Most of the options for improving public sector pay hinge on accelerating and deepening reforms needed in the critical areas of budgetary and cash management, public procurement, tax administration, and revenue performance in general.
75. Consistent with the government’s own commitments to pursue only sustainable improvements in public sector pay, the simulations reported here are grounded in the medium-term macroeconomic framework, reflected in the NPRS. In this scenario, GDP growth is projected to accelerate to reach 6½ percent by 2006. The scenario assumes that macroeconomic policies are calibrated to keep inflation below 4 percent. Fiscal policy is to remain cautious, with substantial effort focused on boosting revenue from 12.1 percent of GDP in 2002 to 14.3 percent in 2006. Current expenditure would grow more slowly than revenue, and be increasingly tilted toward social spending, interest payments, and civil administration wages. Under this “baseline” scenario, the average monthly wage in the civil administration would rise to US$50 in 2006, assuming an unchanged civil administration workforce of 163,500, consistent with the authorities’ MTWF.
76. An alternative scenario consists of asking the following question: if the government wanted to raise the share of overall wages by about one percentage points of GDP to 5 percent of GDP during 2004-06, how could such an increase be financed without undermining macroeconomic stability? In answering this question, we assume that the “additional” resources (i.e., about one percentage point of GDP) are equally applied to civil administration and defense and security, allowing for urban/rural wage differentials.23
77. In order to avoid undermining macroeconomic stability, this increase would have to be financed from either expenditure reductions and/or revenue improvements. If savings can be achieved without reducing real non-wage spending, such savings could of course be allocated to the public sector wage bill. Recent years have witnessed substantial increases in nonwage operating outlays in civil administration. These increases in large part reflect a conscious effort—including through the Priority Action Program—to improve delivery of appropriated expenditure, notably in the priority sectors (education, health, rural development and agriculture). At the same time, there remain substantial improvements to be made in budgetary execution (e.g., deconcentration of expenditure management, intensified and expanded application of procurement guidelines to reduce costs), from which reductions in spending inefficiencies could be achieved. In this regard, it is worth noting that combined outlays on operating expenditures and special programmes rose by 68 percent between 1999 and 2002, while nominal GDP grew by 14 percent. Hence, if budgetary reforms could yield a cumulative savings of 20 percent in operating outlays by 2006, average wages could be increased to reach a level equivalent to 5 percent of GDP during 2004-06.
Cambodia: Composition of Current Expenditure
78. An alternative measure would be to acknowledge that given the government’s low revenue mobilization when compared to other low-income countries, there might be a need to scale back the size of the civil service.24 Reducing the number of civil servants by 500 (0.31 percent of civil service workforce), distributed between urban and rural staffs in proportion to the relative size of each area, beginning in 2004 could achieve roughly the same results as the previous simulation.
79. Naturally, the most promising option for sustainably raising civil servants’ salaries is to improve revenue performance, including through both policy measures and administrative reforms. Although revenue performance has improved substantially since 1999, the revenue base continues to be eroded by tax policies that have dubious impacts on private sector employment but obvious implications for government’s spending options, including on wages. For instance, it is estimated that half of potential revenue from profit tax in 2002 was lost to enterprises qualifying for new licenses granted under the Law on Investment. This amounts to close to ½ percent of GDP, enough to finance half the increase in the wage bill to 5 percent of GDP. In addition, a recent government study of the budgetary impact of tax policy affecting the motorcycle assembly sector found that the various tax advantages granted in favor of motorcycle parts cost the National Treasury another 0.2 percent of GDP.
80. Civil service reform of course consists of more than simply raising pay, and the government is rightly also concentrating on systemic reforms to civil service. Together with pay reform, there is a need to instill an “esprit de corps” among civil servants, in part by providing clear career paths traveled via performance-based promotions, Pay reform alone will have little impact on the quantity and quality of public services without managerial improvements to ensure the full-time presence of civil servants. Pay reform would also have to be buttressed by appropriate legal reforms and/or codes of behavior aimed at preventing corrupt behavior.
81. The RGC has made progress in setting the stage for improving the civil service over the coming years. It has been rightly cautious until needed information was available. During the first half of 2003, the government will, in collaboration with the World Bank and other partners, undertake a number of key studies for taking the reforms forward. These include targeted operational and functional reviews, labor market analysis, pay and employment analysis, and departure policy analysis. Ultimately, however, the RGC will have to redouble its efforts in a wide range of reforms presently underway—on budget execution and deconcentration, decentralization, improving cash management, reducing revenue leakages from weak enforcement and broadening the revenue base through administrative and policy measures, and generally improving the efficiency of spending through better procurement policies. The challenges are indeed daunting; but the possibilities are promising.
Council for Administrative Reform“The PMG Programme—Progress, Strategy and the Next Steps,”Discussion Paper2002.
_____“Strategy to Rationalize the Civil Service,”Discussion PaperOctober2002.
Schiavo-CampoSalvatore et al.“An International Statistical Survey of Government Employment and Wages”,Policy Research Working Paper1806World Bank,August1997.
World Bank“The State in a Changing World,”World Development Report1997Washington,D.C.
______“Integrated Fiduciary Assessment and Public Expenditure Review: Civil Service Reform,”World Bank,forthcoming,2003.
This chapter was prepared by Robert P. Hagemann and Sopanha Sa.
A more thorough assessment of civil service reform is being undertaken in the context of the World Bank’s IFAPER.
The number of personnel in the military is not available for end-2002, and this figure therefore does not reflect demobilization.
Low public sector pay, especially when it is exceedingly low, is known to contribute to poor governance and corruption. See World Development Report 1997, (World Bank).
See “Strategy to Rationalize the Civil Service,” Council of Ministers Discussion Paper, October 2002.
Even though the government has indicated that approximately 9,000 ghost workers were removed from the payroll, this change has not been fully reconciled between the Council for Administrative Reform and the Ministry of Economy and Finance.
With education accounting for close to half of the civil service, there will have to be substantial reductions elsewhere.
The Council for Administrative Reform has developed its own medium-term wage framework (MTWF) on which the wage targets reported earlier are based. The existing MTWF was developed prior to the government’s Medium-Term Expenditure Framework (MTEF) or the macro framework of the National Poverty Reduction Strategy (NPRS). The targets implied by the MTWF still need to be transparently integrated into both the MTEF and the NPRS.
The PMG Programme—Progress, Strategy and the Next Steps, Discussion Paper, Council for Administrative Reform, 2002.
One official from the CAR suggested that a 4 to 1 wage differential between urban and rural areas might be reasonable in present day circumstances, with a monthly income of about $150 per month being an average needed in Phnom Penh to meet daily subsistence needs.
Politically, this perspective might understandably be difficult to accept. This may account for the official government view, as expressed in the NPRS: “Current simulations show that to increase the minimum wage in the civil service to levels in the garment industry, for instance, Cambodia’s GDP would need to increase considerably if the increase was to be financed through government revenues alone.”