F Policy Research q WORKING PAPERS Public Sector Management and Private Sector Development Country Economics Department The World Bank August 1992 WPS 945 Managing the Civil Service What LDCs Can Learn from Developed Country Reforms Barbara Nunberg Centralized civil service management models provide the best starting point for most developing countries because decentral- ized agency systems require technological and human resources beyond their capabilities. Some better-endowed countries could use certain agency-type features selectively, moving toward an agency system as their institutional capabilities increase. Policy ResearchWoikingPapers disseninatethe findings of work in progress and encouragetheexchangeof ideas amongBank staffand aU others interested in developmentissues Thesepapers,distributedby theResearch Advisory Staff,carry thenamesoftheauthors,reflect only theirviewsand should be used and cited accordingly.The findings, intepeations, and conclusions are e authors'owni They should | ~~~~~~~~~not be attributed to the Wodid Bank, its Board of Direcors, its managerrent, or any of its member counltries. Policy Research Pulic Sector Management and Private Sector Development WPS 945 This paper -a product of the Public Sector Management and Private Sector Development Division, Country Economics Department - is part of a larger effort in the department to help countries develop sound practice for civil service reform. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Priscilla Infante, room N9-059, extension 37642 (August 199?, 53 pages). Nunberg examines current civil service manage- agency system as their institutional capabilities ment (CSM) practices in advanced countries to increase. provide guidance for developing country govern- ments that face the dilemma of how to recruit, o Developing countries face trade-offs in retain, and motivate appropriately skilled staff at choosing which CSM functions should be affordable costs, given a limited human resource strengthened first. Two functions - personnel base. establishment control and staff recruitment - are essential for civil service performance and Advanced country administrations are should get top priority. following two distinct paths to improving CSM. Some countries, such as the United Kingdom, Senior Executive Services have proved are engaged in sweeping "managerialist" reforms difficult to design and implement in advanced to decentralize civil service functions and make countries, but many flaws can be corrected in them more responsive to the client public. By adapting them to developing countries, where introducing complex financial reporting systems, there is often an urgent need to groom higher- managers have increased autonomy; some level staff. functions are spun off into semi-autonomous agencies operating on an increasingly commer- 4 Assuming minimal, essential levels of cial basis. personnel establishment and budgetary control, unified pay and classification could be relaxed in By contrast, other industrialized countries, developing countries, following the lead of such as Singapore, have retained more traditional, increasing numbers of advanced countries that largely centralized civil service structures, pursuing have done this. only incremental improvements in specific aspects of CSM. * Given the urgency of other CSM tasks, lower priority should be assigned to reform Nunberg speculates about what is likely to involving performance pay, the benefits of which work best in developing country administrations: have yet to be demonstrated in the public sectors of developed countries. The management Centralized civil service management requirements and costs of installing performance models provide the best starting point for most pay systems can be considerable and employee developing countries because decentralized resistance may subvert such efforts. But perfor- agency systems require technological and human mance-related promotion systems, even if resources beyond their capabilities. Some better- imperfectly implemented, can help move devel- endowed countries could use certain agency-type oping country civil service values toward features selectively. Such administrations could standards of competence and merit. establish strategic plans to move toward a fuller ThePolicy ResearchWorking Paper Series disseminates the findings of work under way in theBank. Anobjectiveofthe series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusioins in these papers do not necessarily represent official Bank policy. Produced by the Policy Research Dissemination Center TABLE OF CONTENS I. INTRODUCTION . ........... .... ........ 1 II. GLOBAL TRENDS IN CIVIL SERVICE MANAGEMENT . . . . . . . . . . . . . . 5 Rethinking Civil Service Management: Currents of Reform . . . . . . . 6 III. GOOD CIVIL SERVICE MANAGEMENT PRACTICE: ELEMENTS OF A SUCCESSFUL APPROACH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 A. Organizing Civil Service Management: Institutionalizing Establishment Control and Planning ...................... . 15 Establishment Control: Organizing Crucial Functions . . . . . . . . 17 Lessons ............................. . 19 B. Improving the Quality of Human Resource Inputs: Recruiting and Building Leadership Cadres . . . . . . . . . . . . . . . . . . . . 21 Recruiting the Best and the Brightest . . . . . . . . . . . . . . . 21 Lessons ............................. . 24 Building an Elite from Within: Lessons from the Senior Executive Service Experience ............ ............ . 26 Lessons ............................. . 29 C. Managing Incentives to Civil Service Performance -- Approaches to Classifying, Paying and Promoting Staff . . . . . . . . . . . . . . 30 How to Classify and Pay Civil Servants: Traditional Standards Vs. Flexibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Lessons ............................. . 32 Assessing Performance: Links to Pay and Promotion . . . . . . . . . 33 Lessons ............................. . 36 Performance and Career Advancement ............... . 36 Lessons ............................. . 38 IV. CONCLUSIONS . . . . . . . . . . . . .38 Suggestions for Further Work . . . . . . . . . . . . . . . . . . . . 39 BIBLIOGRAPHY ................. ............ . 41-53 Table 2 . . . . . . . . . . . . . . . . . . 4 Figurel ......... ....... . 7 Figurel(continued) . . ... ............. 8 Fiure 2 .......... ....... . 9 Figure3 .................. 10 Figure4 ................. 11 Figure5 .................. 12 I. INTRODUCIION As thb link between efficient public administration and the state's ability to play an effective role in economic development is more acutely perceived, the reform of civil services has become a major goal of World Bank lending. Pursued mainly in the context of structural adjustment, civil service reform has thus far meant programs to contain the cost and reduce the size of government under the weight of dramatic fiscal constraints. It has emphasized public employment cuts and the revision of civil servants' pay structures. These early downsizing efforts, though necessary to reverse long term trends of govermment expansion and decay, have constituted only the initial phase of the reform process. Where successful, such initiatives have resulted in a "leanere but not necessarily "meanere state; a state, that is, which can competently formulate policy, implement programs and deliver services. In the era of reform,, shrinking states will have to do more with fewer but, it is hoped, better human resources. Indeed, the need to recruit, develop, and motivate competent cadres to carry out core government functions becomes especially pressing as the public sector contracts, focusing its attention on a limited number of critical activities. Sustaining adequate government performance in these areas will depend on improvements in civil service management (CSM) - narrowly defined here as the management of staff for the core public service -- over the medium and long term. Unfortunately, the increased demand for better human resource management in the public sector comes at a time when basic personnel management in many developing country administrations is in a state of collapse. Lacking suitable rewards and sanctions, civil servants are often unmotivated and demoralized. They are frequently ill-equipped and insufficiently or inappropriately prepared to do the jobs required of them. In many countries, qualified personnel are difficult to recruit and retain because the private or overseas labor markets offer better alternatives. Mechanisms of authority and often probity have broken down. That LDC civil services might in the foreseeable future achieve efficient, much less effective, delivery of services to the client public has become an increasingly elusive goal. At present, developing country governments dispose of only limited means to confront their problems. Most third 2 world administrations have inadequate systems of establishment control, personnel information management, and manpower planniag. They cannot provide sufficient incentives to motivate staff performance. Appropriate career development and training programs for civil servants are mainly non-existent. Although Bank operations have begun to addr,,is some of these problems (30 projects have included CSM components since 1981 -- s;ee Tables 1-2), efforts to date have been ad hoc and experimental in nature, constrained. by a limited knowledge of what constitutes 'best practice'. Technologies used in these programs have tended to be drawn from a narrow range of advanced country models, in some instances applying outmoded or unsound techniques, unrepresentative of leading-edge civil service management approaches. Because rapid changes are taking place in global thining about. the management and motivation of human resources for both the private and public sectors, and because developing country programs must be able to assess the latest wisdom on these issues, it is particularlv vital to scrutinize CSM assistance strategy at this time. This paper seeks to provide some guidance for Bank borrowers on how to manage their civil services better. It does so by surveying stuccessful or promising experiences in advanced and newly industrialized countries. The objective is to identify models or elements therein which might be usefully applied and adapted to LDC administrations for improved government performance. Two caveats should be duly noted. First, a st ict construction of the notion of 'best CSM practice" is simplistic. Despite significant global commonalities, the diversity of successful personnel administration approaches, even in apparently similar environments, belies the possibility of a standardized modeL Differences in socio-cultural, economic, and political endowments significantly affect what is feasible in a given country, thus undermining attempts at uniformity. The search should be for "good" management practice suitable for particular country conditions. Second, the human resource focus of CSM used in this paper represents an attempt to narrow the scope of an already overwhelming topic. The links between personnel ptactices and other aspects of administration, especially financial management, are, of course, intricate. In this last respect, a forthcoming CECPS analysis of public financial management reforms in OECD countries complements the present study.' The paper analyzes the implications for specific CSM functions of two (ideal) types of administrative systems: a traditional model of centralized, hierarchical, administrative control and a "reform" model that, utilizing sophisticated information management mechanisms, devolves greater financial and personnel responsibilities down the administrative line and relies increasingly on market signals to orient civil service management behavior. The industrialized country experience with these two broad models is examined in the hope, if not the hypothesis, that the selective introduction of successful and/or innovative CSM 'See, Michael Stevens, "Innovations in Public Expenditure Management in OECD Countrics", forthcoming, isi2. Organrzauiu-a, structural and, for lack of a less iritating term, "proces" aspects of civil service management are also important (and too often ignored in Bank work) but could not be reasonably accommodated in the scope of this study. Ile terse reference to the political (and socio- cultural) dimensions of civil service reform is not intended to underestimate the centrality of these variables, but rather to acknowledge that these ame in themseles vast topics, the subject of extensive literatures; it was not possible to address them seriously in this review. Finally, despite the paperes human resoure focus, the mbric, "civil service management" Is used in place of the more precise 'civil servie pesonnel management" on grounds of linguistic economy. 3 TABLE I BAMK OPERATIONS WITH CIVIL SERVICE MANAGEMENT REFORM CONIEUfS, 19981-1991 (thrnwuph July 31, 1991) \a Region/Country Ln/Cr Ln/Cr Amamt Approval Type NMer (UeSS m) Oate AFRICA Angola Eco. got. Cacity Cr 227N 23.00 06/91 Carn Eco. Mgt. Ln 3110 9.00 07/89 CAR TAL cr 1150 4.00 05/81 CAR TAL It Cr 1581 8.00 04/85 CAR Eco. Mgt. Cr 1971 13.20 12/88 Gabon TAL Ln 3114 5.00 08/89 Ghan SA Instit. Support Cr 1778. 10.80 04/87 Ghana Eco. Mgt. Support Cr 2224 15.00 03/91 Guinea TAL Cr 1559 9.50 03/85 Guinea Eco. Mgt. Suoport 11 Cr 1963 14.50 11/88 Guinea Bissau 'AL 11 Cr 1935 9.70 06/88 Matawi :D Cr 2036 11.30 06/89 Mati TAL Cr 1307 10.40 12/82 Mauritania Develapoent Mgt. Cr 1865 10.00 12/87 Senegal Oevelopimtnt Mgt. Cr 1910 17.00 05/88 Sudan PE & Eco. Mgt. Cr 1789 9.00 05/87 Togo TAL III Cr 1600 6.20 05/85 Uganda TAL 11 Cr 1434 14.20 12/83 Zaire Eco. Mgt. & ID Cr 1832 12.00 06/87 AStA Bangledesh Pub. Admin. Cr 1349 12.00 04/83 Papue New Guinea Pub. Sector Training Ln 3290 20.80 01/91 EMENA Y&men Arab Republic 10 Pub. Adnin. Cr 2015 10.80 05/89 LAC Argentina Pub. Sector Reform TA Ln 3362 23.00 06/91 Botivia Eco. Mgt. Strengthen. Cr 1977 9.70 12/88 Bolivia Pub. Fin. Mgt. 11 Cr 2279 11.30 06/91 Costa Rica TAL Ln 2519 3.50 04/85 Haiti TAL Cr 1786 3.00 04/87 Jamica Pub. Admin. Reform Ln 2423 4.50 05/84 Jamice Fin. & Program Mgt. Ln 3386 11.50 06/91 Peru Pub. Sector Mgt. Ln 2204 10.20 09/82 \a Operations in which training is the only civil service reform comaonent have been excluded. -~~~~~~~~~~~~~~~~~ TABLE 2 CONPONENTS OF CIVIL SERVICE MANAGENENT REFORM OPERATIONS, 1981-1991 (through July 31. 1991) Establishment Controlt Diagnostic Nanagement Information and Satary and Grading Organizational Work Manpower Planning Systems Structure Reforu Development /a Training /b Angolo Eco. Mgt. Cap. Argentina PS Ref TAL Bangladesh PA Argent'na PS Ref TA Algeria TAL Bangladesh PA Bangoladesh PA Bolivia Eco. Ngt. Bangladesh PA Angola Eco Ngt Cap Bldg Bolivia Eco. Nglt. Bolivia Eco. Ngt. Costa Rica TAL CAR TAL If Argentina Pub. Sect. Mgt. Bolivia Pub Fin Mgt ll CAR TAL I Ghana SA Inst. Sup. Ghana SA Inst. Sup. Argentina PS Ref. TA Costa Rica TAt CAR TAL 11 Ghana Eco. Mgt. Sup. Ghona Eco. Ngt. Sup. Bangladesh PA Gabon TAL Gabon TAL Haiti TAL Jamuica PA Benin TA Ptn. & Eco. Mgt. Ghana Eco. Ngt. Sup. Ghan SA Inst. Sup. Jamaica PA Juuaica Fin. Mgt. Bolivia Pub. Fin. Mgt. It Guinea TAL Ghana Eco. Mgt. Sup. Jamaica Fin. Mgt. Mali TAL Bratil PSM Jamaica PA Jamaica Fin. Mgt. Mauritania Dev. Mgt. Mauritania Dev. Mgt. Gabon TAL auritania Dev. Mgt. Mall TAL Peru PSN Papua New Guinea Ghana SA Irst. Sup. Peru PSH Mauritania Dev. Mgt. Togo TAL III Pub. Sect. Train. Ghana Eco. Ngt. Sup. Togo TAL III Senegal Dev. Mgt. Peru PSM Guinea TA Eco. Mgt. Yemen Arab R. IW PA Uganda TAL It Senegal Dev. Mgt. Guinea Eco. Ngt. 11 Yemen Arab R. ID PA Togo TAt III Haiti Ed. & Train. IV Uganda TAt 11 Jme ica Fin. & Prog. Mgt. Yemen Arab R. ID PA Jordan Marpover Dev. Madagascar Acct. & Mgt. Malavi 10 Maif Eco. Mgt. & Train. Mauritania Dev. Mgt. Mauritanis TAL 11 Morocco PA Moaubique Eco. * Fin. Mgt. Niger Eco. & Fin. Mgt. Niger Eu. PapU New Guinea PS Train. Peru PSM Rwanda TA Pub. Fin. Mgt. Senegal Dev. Mgt. Trinidad A Tobago TAL Uganda lA. II Yemn Arab Rep. ID PA Zmbia TAL 11 /a Building an strengthening of personnel management institutions. /b Includes support for management development institutes, schools of public administration, overseas fellowships and twinning arrangements. 5 approaches could actuallv enable some LDCs to "jump-start" their own process of administrative developrient. The paper represents a fi,st cut at assessing the real prospects for such acceleration. It must rely heavily on conjecture, beca,ise persuasive cross-national comparisons of the impact of different CSM systems on govemment pt rformance are non-existent, because much of what is being carried out in the name of the "reformin,g" administrations is of extremely recent origin and has still to be properly evaluated, and because judglments about the potential viability of CSM approaches in developing countries derive more from accumulated operational wisdom rather than empirical studies of feasibility. The objective of this exercise is emphatically not the analysis of LDC experience with civil service reform. Such experience should be assessed systematically in studies with that express purpose. The assumption here is that the reader is familiar with developing country administrative concerns and is seeking sources f fresh ideas for addressing them. The paper is organized as follows: section II provides a brief overview of some global trends in advanced-country civil service management. Section EII highlights, in specific countries, successful practice that holds potential for Bank borrowers, sometimes couching options in the context of current debate. The focus is on key elements of personnel management systems. For each element, the discussion narrmws to suggestions for the possible application or adaptation of these approaches to LDCs. The final section offers general conclusions and points to directions for further work I. GLOBAL TRENDS IN CIVIL SERVICE MANAGEMENT: THE CQNTEXT OF REFORM Significant changes have been occurring since the mid-1970s in developed country public administrations. Governments ha"e sought to reshape rigid, hierarchical, nineteenth-century bureaucracies into more flexible, decentrs lized, client-responsive organizations, compatible with late twentieth century technological and economnic requirements. For some countries, reforms have been sweeping, representing programs of radicai, systemic transformation. Other countries have pursued more modest strategies of incremental improvements in specific aspects of civil service management while conserving the basic core of administrative structure and practice. The cases in this study can be situated along a continuum of change between these two poles.2 New Zealand, Australia, the U.K, and Sweden cluster at the systemic, reformist end of the scale while the Asian cases, the U.S.A., and Germany have taken more incrementalist approaches France, the Netherlands, and Canada fall in the middle. Differences in emphasis notwithstanding, some basic commonalities characterize global civil service management ir. the last two decades. To convey the contextual background for the subsequent discussion of specific CSM approaches, a brief sketch of the broad thrust of developed-country administrative reform is provided below. 2Examination of public penonnel management approaches in 11 OECD and NIC countries formed the empirical basis for the study. In addition to key informant interviews, relevant secondary materials were reviewed. 6 Rethinking Civil Service Managemen' of Reforn Disillusionment with ':0 '; ..'. -tralized control of civil service management began to pervade advanced country civil setvices in the 1970s and 1980s. Such controls were initially introduced by governments as a means to assure impartial (mainly uniform) application of personnel policies, to professionalize public service cadres, and to rout out patronage and corruption.? Having largely achieved these aims, however, governments were increasingly coniing under pressure to improve administrative efficiency and to render faster, better and more responsive services to a client public whose requirements were becoming mcre diverse. Under greater fiscal strain and facing intensified, wide-ranging citizen-customer demands, tightly controlled, uniform, centralized bureaucracies were attacked for failing to provide effective service delivery at affordable cost Countries thus embarked oin a search for ways to reduce civil service expenditures and at the same time to improve performance standards in government. Several trends in civil service management reform emerged from this quest. In distinct phases, these reforms -- variously denominated the "new public management movement" or the "new managerialism" -- moved swiftly from one country to another, manifesting a kind of global demonstration effect. The first reform phase, commonly called "cutback managen,ent," consisted of measures to curb public spending and civil service staffing in response to fiscal constraints. Downsizing exercises were implemented in a number of countries, including Canada, t'ie U.K. Australia, New Zealand, Japan, and the USA. Cutback management was essentially neutral ib"tween ieft- and right- leaning political regimes. Labor governments in Austalasia pursued fat as avidly as did staunch conservatives in the U.K and the U.S. Staff reductions were mainly achieved through recruitment slowdowns or freezes, natural attrition, early retirement, reorganization and/or privatization of government activities, and, to a lesser extent, through involuntary employment reductions. Some examples of cutback management include the United States' Reduction-in-Force program of the early 1980s, a Japanese program to reduce civil service staff by five percent between 1986 and 1992, the Dutch government's downsizing exercise reducing civil service persoinnel by two percent per annum in the mid-1980s through a privatization and efficiency audit program, and the UK's reduction of civil service staff by 22 percent between 1979 and 1989. While cutback management programs brought about reductions in public employment in a number of countries (See Figures 1, 2 and 3), government spending reductions nonetheless proved difficult to sustain, largely due to the overwhelming burden of entrenched transfer payment systems.4 Recognition of the limits of cutback management in effecting sustained cost- containment and efficiency improvements led to a broader reform approach, one which emphasized more cost-efficient utilization of "inputs" in the provision of public services. Input-focused reforms swept through a number of developed countries, including New Zealand, Austra.a, and the U.K In 3See, Willam Mosher and J. Donald Kingsley, Public Personnel Administration. Harper and Brothers: New York, 1941, for a history of early civil service reform. Successful professionalization and sanitation of the civil service in many countries in the late nineteenth ihrough mid-twentieth centuries also coincided with rising levels of educational attainment, legal and constitutional reforms, wider suffrage and the strengthening of mechanisms of parliamentary oversight and accountability. These parallel achievements must be recalled when contemplating emulative administrative reform in today's developing countries. 4Christopher Hood, Andrew Dunsire, and Lynne Tlomson, 'Rolling Back the State: Tlatcherism, Fraserism and Bureaucracy", in Governance: An International Journal of Polic and_Adrinistration Vol. 1, No. 3 (July 1988), pp. 243-270. Also, as Figures 4 and 5 suggest, civil service employment reductions only tell pan of the story. General government employment (including central, regional and local figures) also need to be analyzed. (mousanid) ~osns _ _ _ _ _ _ _ _ _ _~~~ ~~~~~~~~~~~~~~~~~~~ i ..... . ..L.... CD * a .~~~~~~~~~~~~~~.. .... I Izv~~~~~~~~~~~~~~~~~~~~~~~~~~C I~~~~~~~~~~~~~~~~~~C . ......... .-.- I,~~~~~~~~~............ I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..... . I ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. . . . . .. . . . . 4~~~~~~~~U .... .... .ands)......anCD .. .. ....... IA~~~~ . . .. ...~~~~~~~~~~......... xii Tm-Lh1 I I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ........ .. 1ZILILZI I~~~~~~~~~~~~... ... .... ....... Z ,(Mousands) (thounds) - - . . _ Ig &~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ w L. ID. .. i^ .. . .. % _ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. T ....... .............::y. .i , s ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... .. ... ... . ....................... x xxQx.... .......... ; =t ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ... . r,'...,.. , ..X :.. .... 4 [ | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.-.-..-. . .. . . ..... ... ... 00 .I. . _ . |, A > Hs . X *, _ J = , I I 0 ^ _ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. _iB-.... ... ........ .. 1..... I~~~~ g~~~~~~~~~~~~~~~~~~~~~~~~~. . . ...,.,.. | o l X ~~~~~~~~~~~~~~~~~~~~~~~~~~., .y. ... , .,, . ( d.) (, .'n . _ : . . ..ds) ......s) a___ __ 8 __a a b f a Y a~~~~~~~~~~~~~~~~~~~~~~c (ftumar4s) (VOUnrAft) ~ ~ .. .. .. ... ... .... ... . .. . .. ... .. .. .... .... ..... I. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....... ... I.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.... ..... ...... I.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ........... ..I........ I ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ..... .... . . ._........_. I ~~~~~~~~~~~~~~""..,..... '5 - 222L~~~~~~~~~~~~~~~~~~~~~~.. .... . ........... -3- ~~~~~~~~~~~~~~~~~i2IZI2~~~~~~~~. ........ _ _ _ I ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. . . . . . . I__ 77E7n1f_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....... I I~~~~~~~~~~~.. ..... I _ _ _ _ _ _ _ _ _ _ _ _ _ _~~~~~~~~~~~~~~~~~~~..... .... . .... . . . ... 9 Figure 2 PERCENTAGE CHANGES IN CIVIL SERVICE EMPLOYMENT FOR SELECTED OECD COUNTRIES, 1984-89 United States France Germany udx Sweden -2s5 Japan ąz7% Canada Australia -9.~~~~~~ .t1%1 UnRed Kingdom -a .- 7 Total change Average annual change Note: Changes in Gennany are based on 1985-89 time period. Source: CECPS 10 Figure 3 Civil Servants per 100 in the Population in Selected OEC1D CountTles 376 36 Earliest period 1989 !.26 1.24 1.0 S42uMI.I $:. So~~~ce CECPS~~~a JYnged Skates Canada France Germany Sweden 20% Australia U/nited Kingdom -1.3 m Japan * Total change Average annual change- Note: General govemrment employment encompasses pefsonnel at the central, regional and local levels. Changes in Germany refer to 1985-89 and those in Canada to 1984-88. With the exception of Australia, public enterprise employees have been excluded. Source: OECD 12 Figure 5 Civil Service Employment as a Share of General Government Employment in Selected OECD Countries, 1989 4L-1 41.3X 18.2% 173% IZJ1%3 1aL6% &8%~~.2 France Sweden United lingdom Australa United States Japan Canada Germany Note: General government employment encompasses personnel at the central, regional and local levels. With the exception of Australia, public enterprise enployees have been excluded. Share for Canada refers to 1988. Source: CECPS and OECD 13 the latter, they were embodied in the Financial Management Initiative (FMI).5 Building upon a series of "efficiency scrutinies" begun in 1979 to identify potential budgetary savings through audits of individual government departments, the FMI was officially created in 1982 as a progrnm desiged to introduce innovative, computerized financial information management systems. The program sought to make the budgetary process more transparent and thereby more accountable to higher management and political levels. Department heads were to be given responsibility for efficient use of resources. Departments were subject to cash limits and, with improved information management, were held accountable for control of running costs. FMI reforms were largely confined to better management of financial inputs through strengthening of information technology. Devolution of responsibility was limited to the top level departments. These improvements, concentrating mainly on financial inputs, proved necessary but insufficient to the task of making govemment performance more efficient and responsive. Several years into the reform process, the need to deepen reforms by increasing flexibility and devolving still more responsibility and accountability to lower managerial levels was perceived. In the U.K, these more fundamental reforms were taken up in what came to be called the Next Steps Program.' Initiated in 1988, Next Steps has sought to relax further central financial and management controls and to devolve more responsibilities to the line managerial level. In a variant on similar reforms being implemented in Australia amd, even mnre radically, in New Zealand, the Next Steps administrative reform has spun off various central government functions to newly formed executing agencies. The reform is fashioned loosely after the Swedish administrative model, in which semi-autonomous agencies have operated for centuries with minimal central controls. While policy functions are retained in central departments, agencies are charged with service delivery and, through the person of their Chief Executive (CE), are held responsible for efficient use of inputs and the achievement of performance targets. The ministers to whom they report are ultimately responsible for program outcomes. Organizational culture in agencies is supposed to shift accordingly to emphasize efficiency and service. These new administrative relationships between agencies and parent departments are embodied in what is known in the U.K. as the framework agreement.! Implicit in the agency reform movement is a challenge to the traditional notion of a unified civil service. Too large and unwieldy to be managed as one organization, govemment as envisaged by Next Steps architects and other Agency reform proponents is instead a collection of interrelated but separately functioning businesses.S 5Tle U.K cxperience is presented here to illustrate the nature of agency reform programs. While it differs in numerous respects from other agency modeLs, it also shares many traits in common with Australia, New Zealand, and Sweden, and has served as a model for particular aspects of reform in other countries, including France and Canada. 6'The analogues to the FMI-Neat Steps reform sequence in Australia were the Financial Management Improvement Program and Public Service Reforms. In New Zealand, the Financial Management Reforms, the Corporatization Program and the State Services Act took reforms even further. 7Framework agreements establish the policies and aims of the government and the agency; the respective roles of the minister, permanent secretary and the CE; the financing available to the agency, the estimates process and the degree of financal flsbility permitted; the annual operational or business plan, the annual performance agreement negotiated with the agency, indicating performance targets, indicators and monitoring arrangements, and staffing policies (e.g., creation of new posts, pay flexdbilities). Simibr agreements exist in New Zealand and Australia. Although there has been discussion of introduction of formal contractual agreements in Sweden, currently arrangements are more informal. heEtorno~mist, December 21, 1991, pp. 6667. 14 Agency reforms are stiDl at a very early stage.' Many uncertainties remain, including questions about the degree to which agencies should be held accountable politically, whether the central financial authorities will cede enough autonomy to agencies to allow the concept to work, and whether the focus on performance targets, unit costs and individual agency independence will erode the traditional public service ethos.'° While agency reforms are being pursued fully in only a few countries, many of the "new managerialist" principles driving them are informing other, less comprehensive reforms in civil service management elsewhere as well. A number of "incrementalist' administrations are also undertaking management improvements which reflect an emergent view of government as a service delivery conglomerate. These improvements draw on a range of private sector management approaches, including: organizational mission, corporate planning, customer responsiveness, performance incentives, and the use of new advanced computer technologies. To what degree should traditional administrative controls give way to managerialist/new public management approaches in managing civil services? This question is at the heart of the ongoing public administration debate in developed countries. In the meantime, developing country administrations must make choices about how best to develop their own civil service management systems. This question must be answered by looking at specific CSM functions, determining their priority in the overall task of administrative development for a given country, and assessing what practice will work best under which conditions. Without proffering definitive solutions, the following sections try to analyze appro Ales that have proved useful in advanced countries and to extrapolatt lessons for likely successes in adapting them in developing contexts. III. GOOD CIVIL SERVICE MANAGEMENT PRACTICE: ELEMENTS OF A SUCCESSFUL APPROACH This section focuses on those civil service management functions which, it is argued, are priority items on developing country administrative agendas. These functions constitute a minimal set of essential tasks that must be well executed if government personnel are to perform professionally and effectively. The discussion therefore concentrates on three principal areas of civil service management." It begins by examining the organization and management of the system itseiff including questions of how to structure institutions for crucial maintenance functions like establishment control. The focus then shifts to the quality of human inputs. This involves programs 9Actually, the Swedes are revisiting and reforming their own system in light of reforms taking place in other administrations. l0As will be discussed at various points in this paper, the jury is still very much out on the effectiveness of agency reforms in their countries of origin. The difficulties in assessing the merit of these programs are compounded by the familiar presence, informally noted by some observers, of the Hawthorne effect, whereby the introduction of m environmental change stimulates performance improvements over the short-run, but m3y,not necessarily have enduring positive impact. Some of the most fervent advocates of the agency model may be affected by this phenomenon. See, The Economist. ibid. I"The choice of CSM areas is dictated by the importance of the function to government performance in developing countries, and by the extent to which innovative or interesting approaches have been identified in developed countries. Many important CSM functions and innovations will unfortunately escape these criteria, but, in the interest of manageable scope, could not be treated here. 15 for recruitment of personnel as well as development of elite cadres through senior executive schemes.' Finally, the discussion centers on incentives and their relationship to staff performance, concentrating on pay, classification, and performance evaluation. A. ORGANIZING CIVIL SERVICE MANAGEMENT: INSTMJTUIONALIZING ESTABLISHMEENT CONTROL AND PLANNING Choices about optimal organizational structures for civil service management and the assignment of responsibilities for key personnel functions are both critically important to successful administrative development in LDCs. Civil service reforms springing from fragile, uncoordinated or politically impotent institutions typically founder in developing settings, while those with a stronger institutional base of support have a better chance of succeeding. This section analyzes some of the organizational qualities of well-functioning personnel systems in industrialized countries and then suggests which features might be most useful in establishing a solid institutional foundation for CSM activities in developing administrations. The discussion begins with the organizational arrangements for overall personnel policy and management and then concentrates on the institutional set-up specifically for establishment control and planning, a vital CSM activity. Two basic models of CSM organization predominate in developed administrations. One is a model of centralized control allowing a limited degree of deconcentration. The other is a system with increasing decentralization, flexibility and delegation of CSM authority to the line level. For the cases studied, the Asian countries, France, Germany, the Netherlands, and (to a lesser degree) the USA and Canada conform to the first model, while the U.K, Australia, New Zealand, and Sweden either now fit or are moving toward the second. In addition to a high degree of control and professional standardization, centralized systems of civil service management display a number of common traits. First, for most modem administrations, centralization implies not a single, central monolfithic organ controlling all system personnel movements, but rather some degree of shared personnel authority exercised by several entities at the central level. The arrangement is often tripartite, usually with powers divided among a policy guidance body, an oversight agency which helps ensure fair and meritorious practices, and a financial control cum monitoring organ. Such systems create checks and balances, ensuring redundancyU and fairness, and they curb excessive concentration of authority for civil service management. They work best when limited to a very few organs with clear division of labor and good mechanisms for coordination. Singapore is an illustrative case. Authority for public personnel functions in Singapore is divided among the Ministry of Finance's Budget Division (which handles manpower control), its Public Service Division (which develops policy, coordinates implementation and is responsible for pay and grading, training and productivity improvement), and the Public Service Commission, (which approves higher-level appointments and disciplinary actions). Singapore's success can be contrasted with less functional arrangements elsewhere. In the USA, where clear lines of 17'le subject of training, admittediv important to the development of civil service cadres, is not treated here in order to limit this papers breadth. 13See Martin Landau, on the need for redundancy in successful organizations "Redundancy, Rationality and the Problem of Duplication and Overlap", in Public Administration Review 29. 4 (July-August): 346-58, 1969. 16 responsibility are not drawn, and in Canada, where personnel policy organs proliferate, centralized power-sharing arrangements have frequently led to fragmentation and policy confusion."4 Successful centralized models elevate the status of the personnel function in government by makdng key CSM organs directly responsible to political authority, often with designation as principal agents for major system reforms. In Japan, for instance, the Management Coordination Agency, part of the personnel "troika' in central government, operates under the direct authority of the Prime Minister, coordinating CSM policies and ensuring their direct responsiveness to political requirements. The Japanese attach prestige to the personnel function by placing deconcentrated personnel divisions near the top of line ministry hierarchies and identifying personnel officers as "high fliers with good prospects for eventual appointment to elevated government positions." In the interest of bureaucratic flexibility and more rapid operations, some centralized systems delegate selected personnel management authority to deconcentrated levels. For example, routine personnel information management, especially for lower-level staff, might be handled by departments, subject to central policy guidelines and safeguards. Still, the hallmark of this model is the retention of close, central supervision of functions with crucial implications for the financial soundness of the service, such as establishment control and planning. In contrast to these traditional systems, several countries are developing an alternative model of decentralized civil service management. This model increases the decisional autonomy of line managers on most personnel matters, leaving only broad policy guidelines to be worked out at the center.1' Such decentralization is most advanced in New Zealand and Sweden. Sweden's central 141In the USA, powers are divided among the Office of Personnd Management, Office of Management of the Budget, and the Merit System Protection Board, with functions still being disputed among the three, calfing attention to the lack of coordination in the system and the consequent difficulty in achieing needed pesonel reforms. Critks of the Canadian system highght the fragmentation of CSM policy among five official and non-official institutios the Treasury Board, the Treasury Board Secretariat, the Public Semice Commisso, the Privy Council Office, the Auditor General's Office, and the Public Policy Forum. See, Cohn Campbell. Govemment Under Stess Political Executives and Kev Bureaucrats in Washington, London. and Ottawa, Toronto: University of Toronto Press, 1983, pp. 228-255. 1sFrance and Sweden have also made recent efforts to elevate the personnel function to higher status. To support CSM reforms in France, the General Dimetorate for the Public Service, once an appendage of the Prime Minister's Office, became part of a free-standing Ministry of Public Service and Reform in the late 1980s. The head of the Ministry was upgraded to Cabinet Status. In Sweden, the Ministry of Public Administration was formed in 1983 to spearhead the public sector renewal progrm. 16While moving toward a more flmdble, increasingly decentralized structure of personnel management in the public sector, Australia has not moved toward full 'agency" autonomy in CSM, but rather retains departments headed by Depanment Secretaies. Within running cost ceilings, though, the Department of Finance has delegated establishment control to the departments, which, in fact, have responsibility for most operational aspects of personnel management, with the exception of the Senior Exeutive Service over which the Department of Finance retains control of numbers and levels. while the Public Service Commission exerts quality control and assures due process. In addition, responsibility for efficiency scrutinies was moved recently from a central Efficiency Scrutiny Unit, formed in the footsteps of the UK, to the departmental level, but reportedly, this has led to their demise. The UK is also operating with a decidedly hybrid system in the midst of reforms unfolding through the Next Steps Program described earlier. Al present, CSM powers are distributed much as in the centralized systems in a tripartite division of tabor among the Treasury, with responsibility for all financial aspects of pemonnel administration including pay, pensions, manpower control and labor negotiations; the Office of the Minister for the Civil Service, which, as part of the Cabinet Office under the Prime Minister, develops personnd policies and guidelines for recruitment, staff appraisal, promotions. training and career development, and controls senior management appointments; and the Civil Service Commission, an independent body attached to OMCS which oversees senior remuitment. In addition, the Prime Minister's Efficiency Unit supervises departmentally executed efficiency audits and continues as a driving force behind further CSM changes 17 personnel "troika" -- the Ministry of Public Admiristration, the National Audit Bureau, and the Agency for Administrative Development -- performs minimal oversight, sets policy guidelines, and promotes and disseminates sound and innovative practice. These oversight functions aside, and although pay guidelines are still determined centrally, responsibility for the majority of personnel tasks now rests with 100 semi-autonomous agencies and 13 small policy ministries have responsibility for the majority of personnel tasks. In New Zealand, a central agency, the State Services Commission (SSC), is assigned only a quality control and oversight role. It advises government departments on policy and standards, monitors the perfc.rmance of departments and chief executives, oversees the new Senior Executive Service and oversees the selection of candidates for appointment to chief executive positions. Most aspects of personnel management, including recruitment, promotions and dismissals, position classification, manpower control, and almost all training are now handled by executive departments which, under the purview of small headquarters departments, carry out the bulk of government functions. In principle, decentralizing CSM functions allows managers to tailor staffing to real requirements with more flexibility. The absence of central controls means these decisions can be made rapidly and in response to local market conditions. Note, however, that decentralization of personnel management functions, even where it has been taken farthest, is being implemented in the context of elaborate, technologically sophisticated systems of information and financial management, utilizing complex and expensive computer software and hardware in order to maintain tight reins on running costs. Technical training of staff throughout government has been necessary just to run these systems. In addition, good financial and personnel management skills must be present down the line in order to make such arrangements work. Managers must Lb. prepared to make a range of financial and personnel decisions, and line staff need sufficient technical capability to administer routine personnel movements. Capacity to feed appropriate information back up through the system also needs to be present at decentralized levels. The importance of these types of organizational issues is illustrated clearly in the exercise of such key functions as establishment control. Establishment Control: Organizing Crucial Functions Recent experience with public sector reforms in developing countries clearly shows that building sound establishment control and planning practice is a necessary condition for the success of other civil service management improvements." Among the industrial country cases examined for this study, the centrally organized civil services appear to perform these tasks Increasingly, departments and, as they develop, agencies are being given more flexibility for managing lower and middle-level staff, and on a limited bi sis, professional and technical staff at higher levels. While pay bargaining and the staffing of the vast majority of high level positions continue to be centralized, the prowss of decentralizing many personnel management functions is underway througb the Next Steps Program. The Framework Agreements negotiated between the departments and Treasury are starting to provide Chief Executives with personnel management flexibilities suited to the particular activities and needs of their agencies. Eventually, the civil service is expected to retain only a small policy core staff managed by the center with considerable autonomy for personnel functions at the executing agency level. 17See Barbara Nunberg and John Nellis, "Civil Service Reform and the World Bank," PRE Working Paper, 1990, and Louis de Merode, "Civil Service Pay and Employment Reform in Africa: Selected Implementation Experiences," World Bank, 1990, for confirmation of this point. 18 effectively."' They generally ensure a high level of system coordination, providing policy makers and planners with the necessa degree of information and control both to stabilize public eanployment during fiscal crises and to design manpower policies for the longer term. Centralized models usually operate tight monitoring systems through which nationally determined staffing levels are enforced via a coordinated effort of the central personnel authority and the Finance Ministry. Relying heavily on staff audits or inspection techniques, governments in Japan, the Netherlands, France, Singapore and Korea have all sought to impose government-wide Ceilings on staffing levels and to utilize forward planning to contain public employment. Japan's cenetally controlled system, for instance, has since the late 1960s facilitated government staff reductions. Working in concert, the Management Coordination Agency enforces ministerial employment ceilings mandated by law, while the National Personnel Authority adjusts civil serice recruitment targets. Since 1968, these ceilings have gradually been lowered, enabling the Japanese to move beyond personnel stabilization to net reductions."' A centralized model has also proven highly useful in Singapore's manpower control program. The Ministry of Finance's Budget Division sets staffing levels in its manpower plan, and ministries must obtain the Budget Division's approval to add staff. This enforces the government's zero-growth policy (begun in 1988), with the Budget Division and the Public Service Commission working in concert to control recruitment. To support central efforts, the capacity to plan and monitor staffing levels has also been developed at the ministerial level. With training provided centrally in staff review techniques by the Management Services Department and the Auditor General's Office, all public agencies carry out comprehensive programs of manpower scrutinies to identify superfluous functions and staff. Other highly centralized systems include Korea and France. Through tight control of department manpower levels, regulation of recruitment and periodic staff inspections, Korea has kept public employment constant since 1980 at four percent of the fabor force. France maintains establishment control through the budget, carefully setting the number of officials in each salary scale and establishing fixed percentages for the number of officials who can be promoted from one salary scale to another. The key mechanism to adjust civil service numbers is recruitment restraint. Having achieved these successes, some centralized systems of manpower control are now relaxing their reins a bit, ceding limited discretion over staffing levels to departments and line managers in the interest of greater flexibility. In Canada, for example, the Treasury Board Secretariat allocates person-years to departments and only controls the number of senior managers. Departments and agencies have gotten increased autonomy in distributing person-years among different levels of staff, and they can make special appointments amounting to 10% of their base complement of positions. I8Data which would provide definitive bases for comparison of different approaches to establishment managemcnt are not easy to come by. Ile goal of any system should be to ensure adequate stability and forward planning of wappropriate,f -- i.e. neither too high nor too low - staffing levels in order to avert emergency retrenchment programs. Sufficient time series on civil service numbers with reasonable comparability across countries (i.e. with consistent definitions of what constitutes the civil service) are not available,however, and thus comparative indices of stable establishment management are difficult to construct. 19!n Japan, multi-year forecasting Is facilitated by the hign staff retention rate and the lifetime nature of public employment. 19 Other countries have moved even more deliberately toward a distinctly decentralized manpower control model, using running cost controls as the key determinant of optimal staffing levels. These systems impose across-the-board spending limits but are indifferent as to the level of staff expenditures within recurrent spending categories. Typically, such systems require "micro- management" of information inputs, with regularized reporting at short intervals to the finance authority responsible for overali system monitoring. While the U.R's Next Steps program is moving government in this direction,' Australia and New Zealand have progressed farthest down the decentralization road. Individual departments or agencies have decisional autonomy over manpower levels, subject to central running cost monitoring. In Australia, line departments may freely substitute between staff and equipment expenditures.21 Significantly, this delegation has been accompanied by a transfer of central monitoring authority from the now defunct Public Service Board to the Finance Ministry, thus tightening central controls through improved technologies. In the Australian case, staffing levels have declined steadily since the mid-1980s, due, some claim, to consistent and firmly applied cash limits. New Zealand's establishment arrangements are similarly decentralized. Departments report staff strength to Treasury on a trimester basis. Manpower levels have been reduced by 25%. One cause of this reduction, it has been reported, is the newly acquired ability of managers to customize staffing levels to actual requirements. Lessons: The advantages of centralized civil service management organization -- enhanced control, coordination, capacity for longer-term strategic planning, and generally greater political access for the personnel function -- probably favor such arrangements in many LDCs, where civil service management lacks just these attributes, and questions of flexibility and responsiveness are important but secondary concerns. The industrialized country experience shows, however, that centralization does not imply (as frequently evolves in developing contexts) a center composed of a single, mega- ministry of personnel management or public administration, or an omnipotent Ministry of Finance. Stability and coordination is often achieved through a triumvirate of personnel organs which can check each other, allowing for sufficient system redundancy and backup of key functions. Operating through well-institutionalized coordination mechanisms, these organs share responsibilities for policy formulation, merit protection, maintenance of standards, and financial monitoring. Strong, centralized systems exhibit some degree of deconcentration or delegation to personnel units within line ministries. The degree of delegation depends on two variables: the importance of the function and the technical and institutional capacity at the deconcentrated level. In centralized advanced country CSM systems, control over most personnel functions remains at the center, but delegation of routine maintenance functions to line entities is increasing, sometimes with ongoing support from central authorities. -0The present establishment control arrangements in the U.K. continue to give the central Treasury control over the number of staff in the top three grades. But Treasury now only monitors departmental staffing levels for the other grades to ensure these fall within running cost expenditure plans. With the assistance of Treasury inspectors, departmental staff inspection teams carry out functional reviews to detect overstaffing and grade drift. Since 1979, the Prime Minister's Efficiency Unit has worked with the departments to organize efriciency audits which identify surplus staff as part of broader reviews of departmental operations. 2The conversion factor used to effect this substitution in the Australian case, however, is not based on absolute parity between staff and equipment: KM.25 staff dollars equals $1.00 equipment dollars, reflecting overhead costs such as superannuation, etc., which have the effect of inflating total staffing costs. (John Baker. personal communication, 1/92.) 20 As a general principle, centralized organizational models probably represent the prudent course for developing country administrations, typically characterized by serious skill shortages below high bureaucratic levels. Although a few LDCs with greater personnel management potential might reasonably choose sel- tive deconcentration as do some advanced countries, even the most limited delegation would req re serious assessment of both human and infrastructural requirements and existing capacity throughout the system, as well as careful planning for future training and investment. For crucial functions such as establishment control and planning, the case for LDC centralization is that much stronger. Ubiquitous concerns about fiscal restraint and the poor track record of so many LDCs in restraining civil sevice expansion underscore the need to keep civil service numbers in tight check through central, uniform controls.' While decentralized CSM models emerging from Agency reform experiences may offer the future possibility of enhanced flexibility and line manager accountability, they also increase system susceptibility to fraud, patronage and corruption - clear and present dangers to many developing country administrative establishments. And implementing decentralized establishment management arrangements requires tremendous management capabilities and technical capacity. In most developed countries, staff have needed significant training to come up to speed. Indeed, to the extent that such capacity does not exist down the line, the tendency of central organs to "micro-manage' through excessive information requests can actually result in an even greater burden on the center than in an initially centralized system. It goes without saying that the skdll cum technical gaps in most LDCs are likely to be that much greater. Management of establishment levels appears to have been particularly successful- meriting close scrutiny and emulation by LDCs-- in the Asian countries. They have installed 'smart' programs of forward manpower planning that enable governments to avert overstaffing crises before they happen. The relatively low and constant number of civil servants in Japan, as compared with other OECD countries in recent years, attests (with the aforementioned data caveat) to the effectiveness of this system (Figure 3). Planning has been carried out on a govemment-wide basis but with considerable input from intensively trained line ministry personnel. Such training (in the case of Singapore, training administered by central Ministry staff for specific tasks) can offset the potential risks of an overly rigid, centralized system, providing more balanced participation from the periphery. Increasing the skills of line staff may also set the stage for agency-type reforms which could come later, at least in countries where informatic infrastructure and overall skill levels are relatively high. 22This does not mean that some aspects of establishment control such as staff inspection or reporting cannot be shared by decentralized units, however. 21 B. IMPROV-ING THE QUALrTY OF HUMAN RESODURCE INPUTS: RECRUrTIG AMD BUILDING-LEADERSHIP CADRE Recruiting- the Best and the Brightest The recruitment and initial placement of civil servants are crucial determinants of the enduring quality of the civil service.' The importance of recruitment is weOl understood in successful industrial country administrations; in contrast, relatively little attention has been given to this critical area in the Bank's own civil service operations in developing countries. Cultural traditions, labor market behavior, and educational capacity all obviously condition the kind of recruitment system whit h evolves in a particular country, but some common practices and some promising innovations have emerged in industrial civil services. As above, the aim of the following discussion is to cull advanced country experiences for lessons applicable to developing civil services. Recruitment systems in industrial country civil services break down into two principal ideal types, with real cases falling between them: 1) closed entry, rigid hierarchical systems with limited inter-class (and often inter-departmental) mobility and highly selective entry requirements; and, 2) more open systems with lateral entry, greater vertical and sometimes horizontal mobility, and more flexible entry mechanisms. Japan, Korea, Singapore, Germany, the Netherlands, and France fall roughly into the first category. The U.K. is in a transitional position. The U.S. (to a degree) and, increasingly, those countries subjecting recruitment to market considerations, including Australia, New Zealand, and Sweden, conform generally to the second. Closed systems (sometimes called "mandarin") rely heavily on selective recruitment to maintain high performance standards. They perpetuate the notion of a "chosen elite" in government service. Competitive recruitment takes place in a centralized context and is usually ensured by rigorous entrance examinations or high scores on general educational tests. Examinations are often coupled with other selection mechanisms, such as intensive interviews; assessment centers or academic degree or certificate requirements. Frequently, recruits are pre-selected through single "feeder' institutions in the higher educational or training system. Often these highly competitive selection criteria are used to lock qualified recruits into a closed system to which admission is only possible at the entry level. Entrants go directly into a particular class of service, usually on a "fast track.' Entrance into lower classes is based on less rigorous but still standardized examination or educational criteria, and movement among classes is very restricted. Elite cadre entrants are generalists, not technical specialists, and they are inducted into a corps (often an inter-ministerial personnel group serving as the basic managerial cadre for the civil service) or career stream, not hired for a specific job.24 They may be easily rotated, if not throughout government as a whole, then at least within the particular Ministry directly hiring them. Throughout the recruitment and training process, elite cadres undergo intense socialization to internalize core civil service values. The two countries most nearly embodying the mandarin ideal are Japan and France. The backbone of the Japanese system is an extremely competitive examination with supporting 23Recruitment has traditionally had important consequences for public sectors where separation of non-performers has been difficult if not impossible. To the actent that some governments may be moving toward impermanent contractual employment arrangements, recruitment mistakes may be correctable and therefore less daunting. 24To the ectent that the French corps system may also be related to professional specialization, it deviates from this ideal type 22 interviews to select staff for each of three career tracks: higher, middle and lower civil services. With virtually no exceptions, recruits enter a track for life; no mobility operates among bureaucratic classes. Because of the high social status traditionally afforded to government service as a profession, large numbers of applicants compete for relatively few places.' Civil servants are recruited directly into individual ministries, where they remain throughout their careers. One feeder educational institution -- Tokyo University, established in the nineteenth century expressly for public service training -- produces the lion's share of higher civil service recruits. Most recruits are law-generalists, selected for their potential for highzr management posts.' The assumption is that supplementary technical training can be provided later, if needed, by the recruiting ministry. Personal as well as professional and academic attributes are recruitment criteria. The French also rely on competitive examinations to determine entrance to the exclusive feeder training institutions for elite civil service cadres, the Ecole National d'Administration (ENA) and other specialized 'grandes ecoles." ENA students complete a one-year professional internship followed by 15 months of course work Although the government is starting to open ENA admission to functionaries already in service at lower grades with less elitist backgrounds, their chances of succeeding on the competitive entrance exams are, in practice, low. The system's exclusivity remains intact. Some countries have modified this model, intensifying pre-service preparation through practical internships. For example, Singapore recruits elite cadres into a rigid class hierarchy but relies on general educational credentials and interviews. It also uses two-year, probationary, pre- service, fast-track, "cadetships" with rotating ministerial assignments to train and socialize the mandarin class. The German practicum is another internship program combining on-the-job training and course work for potential elite recruits before admission to service. With the expansion of private sector opportunities and the comparative erosion of public sector salaries in some countries, top graduates need to be wooed more aggressively to join elite government corps. Several countries use scholarship bonds. These are training or educational benefits provided in exchange for promises of time in service. In Singapore, bonds are used to recruit scarce technical skills. In France, entrants to "grandes ecoles" are also bonded in return for an enforceable (i.e., penalties for evaders) commitment to a specified period of government service. In addition to these recruitment mechanisms, governments use other instruments to attract people with scarce skills. Market supplements or recruitment bonuses are used for 250n average only 1 in 19 examinees pass the level I exam (for the higher civil service), I in 9 the level 11 exam, and I in 7.5 the Level II! exam. Each year about 1300 applicants pass the Level I exam, but only about 700 (54%) are appointed by the Ministries. The number of applicants taking the entrance exams is reported to have fallen in recent years due to the gnrwth of the Japanese economy and competition with the private sector, as well as eroding prestige of the public service. 2&1The predominance of Tokyo University graduates becomes even more pronounced at the highest levels of government. Ctitics maintain that the "clique' of high-ranking civil servants with Tokyo University degrees hae used their influence to block promotions of graduates from other schools. A few ministries have recently moved to promote some graduates from other universities to top management positions, but it is too early to tell whether this represents a major policy shift. In addition, new recruits increasingly come from more diversified backgrounds with derees, for eample, in engineering and public administration. See, Paul S. Kim, Japan's Civil S.Vvice System: Its Structure, Persnonel and Politics Westport: Greenwood Press, 1988. 23 specializations in short supply. In Japan, bonuses are now offered to recruits in remote geographical areas as well as for hard-to-recruit professions. In addition, more flexible working and leave conditions are offered to scientific personnel on entry. In contrast to the mandarin systems, some countries favor an alternative, more open model of civil service recruitment. They emphasize flexible, decentralized, sometimes market-oriented practices. The U.S. system stands in stark relief from elite models, for example, permitting horizontal entry without age restrictions and allowing considerable inter-class mobility, though inter-agency mobility is rare.' The U.S. also has moved toward more flexible hiring practices, replacing its centralized, competitive entrance examination for entry level staff with profession-specific exams and in some cases non-competitive procedures affording individual managers considerably more discretion in hiring.3 Decentralized, flexible recruitment is gaining ground in the agency reform administrations as well. Although the U.K retains its highly centralized recruitment system for the higher civil service, inducting elite-trained generalists into its fast-stream Administrative Trainee Program, some flexibilities are being introduced on a limited basis at lower levels, thus giving greater discretion to Department and Agency heads.2' Many departments have begun recruiting directly from the market, using pay flexibilities within central guidelines.' "The non-hierarchical qualities of the U.S. system contriouted, along with steady erosion of real pay since the 1960s, to a civil service leadership vacuum, motivating the creation of the Senior Executive Service program, discussed later in this paper. 38More flexible hiring practices were introduced in the U.S. through the Civil Service Reform Act of 1978 which aimed to balance the need for greater managerial flexibility in personnel decisions with protection of the merit principle. Some observers worry, however, that by decentralizing many aspects of personnel management, CSRA may have encouraged more frequent use of the buddy system and nepotism in recruitment. This potential for abuse is exacerbated by the suspension of the Professional and Administrative Career Exam (PACE) program and the widespread use of "Schedule B" recruitment procedures which allow managers to dispense with competitive exams when hiring college graduates or through "name requests" which allow managem to place a known candidate's name on a vacancy short-list, if he meets minimal qualification requirements. When surveved, managers preferred Schedule B arrangements. as it gave them more control from central intervention in hiring. But results of a Merit System Protection Board survey in 1989 revealed that 1/3 of rank-in-file federal employees reported that they had been denied a job because someone else had benefitted from personal connections. See Carolyn Ban and Harry C. Redd, "he State of the Merit System: Perceptions of Abuse in the Federal Civil Service", Review of Public Personnel Administration. 1991. 2'Recruitment for high-level positions and the higher civil service's fast stream - the Administrative Trainee Programme (ATP) -- remains under the control of the OMCS and its CSC. The process for recruiting Administrative Trainees (ATs) is very selective and involves eaaminations-only 10-20% of the applicants pass this stage--and interviews and group work exercises at an assessment center. This recruitment process still seeks out elite-trained generalists. Members of professional and specialist groups are not admitted to the ATP fast stream, and as a result, they are largety prevented from rising to the top civil service management positions. ATs are give rotational assignments during a two-year probationary period. which is followed bv a 16-week course at the Civil Service College. The AT stream provides the policy core for the UK civil service. But as Next Steps reforms progress in the UK. this policy core is likely to shrink in relation to the executing agency staff. 3In accordance with centrally determined guidelines, departments are free to recruit through their own personnel units, request the services of the cenLral civil service commission or use outside contractors. The government is currently creating a Recruitment Agency out of the CSC which will market its services to the departments and agencies. At the top of the agencies. there are as yet no fixed rules for the recruitment of agencv Chief Executives but generally these positions are open to candidates outside the civil service and are advertised nationally. (The fact remains that most of the appointees still come from the career service, but this reform is stiu quite recent.) Chief Executives are selected on the basis of management experience and are appointed to fixed terms. In some cases, the government will pay them salaries above civil service scales to capture highlv attractive candidates in the market. 24 Recruitment reforms in Australia have also sought to decentralize the hiring process and make selection fairer and more openly competitive. Departments rather than central authorities now handle recruitment of university graduates for higher level positions. Some departments, such as Treasury, run their own internship programs as a recruitment tool. Although many higher level staff are selected on a service-wide basis through the Senior Executive Service, heads of Departments are now appointed on a time-limited, contract basis. Many high level positions are publicly advertised, and departments are legally obliged to weigh various selection criteria, including professional qualifications, experience, and personal qualities. Panel interviews and reference checks are used to approximate private sector recruitment procedures. Devolution of recruitment authority has progressed farthest in Sweden and New Zealand. In the former, agencies develop their own procedures in conformity with national guidelines. Most use examinations, interviews and professional job descriptions. The central government personnel agency runs a head-hunting unit that assists ministries and agencies in the recruitment of high-level managers. In addition, market salary supplements are paid to key personnel in hard-to-recruit professions. In New Zealand, recruitment has become increasingly decentralized. Department cLhief executives select and appoint their own staff, subject, once again, to central guidelines. Agency staff positions are advertised, and market salaries may be offered. To ensure the political neutrality of chief executive appointments, the State Services Commission assembles interview panels with representatives of the business sector and other elements of the non-government community to select CEs. CE job descriptions are cleared with the individual Minister and the Cabinet. CE positions are advertised nationally within and outside of public service, but so far most CEs have come from the Senior Executive Service, the only remaining cross-departmental civil service core recruited centrally rather than by individual departments. Lessons: What are the lessons of advanced country recruit ent experience for building professionalized civil services in developing countries? First, developing countries should pay at least as much attention to recruitment as do advanced administrations. The induction of good staff can compensate for numerous other system deficiencies; conversely, the absence of qualified cadres is difficult to overcome later. Developing country CSM reforms, including those sponsored by the Bank, have not adequately addressed the need to improve the quality of recruitment practices (except with regard to hiring freezes to stem public employment expansion). Concentrated efforts in this area may yield higher payoffs than some other types of CSM assistance. Second, while staffing constraints in developing countries affect all levels of the bureaucracy, many countries correctly perceive the most urgent task to be that of filling higher level management and policy positions with qualified professionals. This need justifies adopting at least some features of the mandarin model, which emphasizes the selection and grooming of elite-trained cadres for higher positions. The mandarin model is attractive in spite of its disadvantages, including the approach's perceived violation of egalitarian values; low credibility of elitist credentials (due to inferior quality training and educational institutions); and the absence of qualified subordinate cadres, whose support is important to successful performance of the elite administrative class. These obstacles can be formidable, as witnessed by the numerous unsuccessful efforts to replicate the French ENA ,perience around the world. 25 Still, certain elements of the elite cadre approach have real advantages for developing countries with limited human resource endowments, where selective rather than comprehensive efforts may be warranted. Creating a prestigious image for a small, well-chosen mandarin class can have favorable spillover effects in motivating other parts of the civil service. Moreover, a flexible, eclectic strategy, using only the most positive aspects of the mandarin model, and combining them with other approaches, may avert some of the aforementioned pitfalls. For example, the exclusive aspects of the mandarin model can be retained while the exclusionary ones mitigated by (1) allowing entry into the elite class or the feeder institutions at multiple points on a career path, (2) by ensuring equitable access from all parts of society into the elite cadre, and (3) by strictly enforcing meritocratic entrance criteria.3 Such policies should be coordinated with establishment planning mechanisms to avert excessive entrants into feeder institution. This is necessary to minimize pressure for civil service employment expansion from swelling numbers of feeder institution graduates claiming entitlement to government employment.32 Other recruitment mect. nisms used in advanced systems could also prove adaptable to developing country needs. Bonded scholarships are already in use in some LDCs. Difficulties faced by governments in raising civil service remuneration ievels can be offset by such sweeteners designed to attract high caliber candidates to the public sector. Such programs could be expanded and targeted to specific skill gaps, as they are in Singapore. Needless to say, enforcement difficulties may be greater in developing countries, and conditions of service upon return may largely determine the success of this type of instrument. Deficiencies in developing country education systems can be compensated by supplementary pre-service education programs, which, as the Asian experience shows, can be a highly efficient means to target civil servants with customized training. Pre-service internships or "cadetships" may serve the same purpose. The additional "probationary" mechanism used in the German system may provide an effective way of weeding out unsuitable employees before they become legally entrenched. The mandarin preference for generalist recruits who can be molded to fit precise civil service needs can also be complemented in developing countries by adjunct programs to recruit specialists with necessary technical skills at various bureaucratic levels. Relaxation and diversification of remuneration norms on a limited and carefully monitored basis may be necessary and desirable in LDCs where such technical expertise is in short supply. Some of the more flexible, market-oriented approaches emerging from the agency reform countries might be instructive here. New Zealand and the U.K, for example, are emphasizing the importance of technical and management skills to supplement the existing policy orientation of civil servants. These reforms also stress the importance of going to the market to find such talents and of providing recruiters with the necessary discretion to compete in that market with regard to remuneration and benefits. These reforms also underscore the need to determine on a decentralized basis which skills are necessary and for how long, a flexibility which could be helpful in LDC settings. 31Perhaps such criteria could be expanded to include practical experience as well as academic achievement. 32 The distonion of the French ENA system in several African countries legally entitled university graduates to civil service jobs upon receiving their degree, thus creating tremendous pressure to expand government employment bevond reasonable fiscal limits. 26 Introducing these types of market competition and challenging traditions of lifetime entitlements may indeed spur civil service performance, but prudent operationalization is required. Full adherence to decentralized, market recruitment models also poses risks in developing contexts, especially when the "market" consists mainly of international donors bidding up the price of scarce professional talent, often resulting in a relentless brain drain from the public to the private - international sector. The introduction of recruitment flexibilities in developing contexts also provides opportunities for favoritism and corruption, which are perhaps less likely when uniform standards are enforced. This highlights issues related to the retention of a government-wide corps inculcated with a public service ethos, discussed at greater length shortly. The tensions between the development of such a cadre and the thrust toward decentralization and relaxation of central recruitment norms have still to be resolved in agency reform countries. The dilemma looms even larger in LDCs where the supply of professional skills is so limited.' Finally, the emphasis on elite recruitment should not lead to neglect of higher recruitment standards for middle and lower level civil servants. While recognizing that improving r:ie quality of lower level staff is in many ways dependent on improving the capacity of the basic education system - usually a long term proposition, -- tightening admissions criteria for all civil service entrants is something that can begin to be addressed now. A fundamental lesson, therefore, to be drawn from the industrial country cases is the importance of developing rigorous (but appropriately conditioned by individual country circumstances) standards for entrance examinations to improve recruit quality. Such testing attains even higher importance where educational credentials are difficult to rely upon, unfortunately the case in many LDCS2' Building an Elite from Within: Lessons from the Senior Executive Service Experience Recruiting new entrants into a prestigious higher civil service class through elitist, pre- service feeder institutions offers one strategy to meet the urgent need to build higher level management and policy capacity in developing countries. Some LDCs are investigating another. building an elite, professionalized, service-wide corps through the creation of Senior Executive Service (SES) programs. The SES experience is important to examine, because some civil service reform programs have installed such schemes without a full understanding of the intricate problems they involve. This section looks at the SES experience in several developed countries, and considers the merits and feasibility of adapting similar arrangements in developing administrations. A number of developed country governments have taken steps to isolate the management of senior personnel, providing more attractive incentives and conditions of service than 33An extreme example of this pervasive di'emma is found in Mozambique where short supply of skilled professionals coincides with low government pay and artificially high salaries paid by international donor projects. See, Peter Fallon and Luis Pereira da Silva, 'Public Pay and Employment Issues in Mozambique," World Bank. 1991. 34Attempts to apply uniform standards to civil servants have been made through some Bank supported programs, such as the Economic Management project in Guinea which introduced competency testing to establish minimal requirements for civil servants. The implementation of this program proved difficult as it met with considerable resistance from entrenched employees. This experience underscores the need to apply rigorous selection criteria at the outset through the recruitment process, rather than relying solely on programs to raise standards at mid-career points. 27 those available to lower level staff. Senior Executive Services were designed, for the most part, to combat a common set of problems plaguing many higher civi; services in recent years. These include: 1) wage erosion in comparison with the private sector and wage compression in relation to lower civil service echelons; 2) increasing marginalization of higher civil service cadres from policy and political decision makers because of perceived lack of the former's perceived lack of responsiveness to political agendas; 3) the absence of a public-spirited, interagency, service-wide corps or elite cadre; and 4) difficulty in attracting highly qualified professionals away from the private market. Governments have thus commonly sought to establish interdepartmental executive corps, sometimes through strengthening or perhaps diversifying existing elite cadres (as in France) and sometimes by trying to instill a public service ethos to counterbalance the atomistic tendencies of agency reforms (as in New Zealand). Japan and France have made efforts to strengthen the elite corps of their civil services by broadening the experience and vision of corps members through increased interdepartmental mobility. Both countries have stepped up rotational policies to encourage movement of the top administrative classes among agencies. Both have sought to diversify the composition of the elite (through widening the net of recruitment beyond Tokyo University in Japan and through opening access to ENA beyond a small select group in France). 'n its efforts to strengthen its elite administrative class, Singapore introduced a superscale pay category above the general salary schedule and a separate career scheme for the top grades. Perhaps the most intensively studied program is the U.S. Senior Executive Service, created in 1979 and administered by the Office of Personnel Management. Uniquely, the U.S. SES includes both career civil servants (90%) and political appointees (10%). The SES was created to improve the pay and status of senior officials; to create a service-wide senior personnel system with increased opportunities for mobility, management training and career development; and to reduce tensions between political appointees and career officials. Enhanced remuneration packages linked to performance evaluation criteria were to be traded off against secure civil service status, foregone upon entering the program. The U.S. concept differs from closed,,elite models in that it is possible to rise through the civil service ranks to attain SES status or to enter laterally from outside government. These objectives have frequently conflicted, and on most progress has been slow. Efforts to depoliticize the cadre have been resisted, with the number of posts filled by career SESers declining in relation to political appointees since inception. A service-wide senior personnel system with increased mobility opportunities has not yet taken hold, and the federal bureaucracy remains segmented into powerful agencies with separate cultures and loyal permanent officials. No formal mechanism for inter-agency assignments has been established, and management developmen; and training for senior personnel has occurred only on a limited basis. Progress on the pay front also lagged. Inflation steadily eroded real pay between 1979 and 1990, but in 1991 legislation raised average SES salaries by 18 to 25 percent. Despite these failings, SES critics have sought to strengthen rather than eliminate the corps, assailing the government's lack of sustained commitment as the root of its troubles. The 1989 Volker Commission, charging that the SES's problems stemmed mainly from underfunding and politicization rather than a flawed program concept, urged restoration of 1969 pay levels, assignment of greater responsibility and the provision of training to career officials, and reduction in the number of political appointees, 28 Canada has also tried to develop a new interdepartmental, corporate identity for senior managers and to increase their mobility throughout government. To build this corps, government consolidated all senior managers into the six-tier Management Category. While the number of staff in this category is controlled by the central personnel authority, departments can decide whom to promote and recruit into this group. So far, the inertial power of the department- as-fiefdom system has thwarted mobility, but recently introduced management training and rotational assignment mechanisms have raised hopes of achieving program goals. The desire to maintain a service-wide elite corps is strong, even in agency reform countries with an apparent commitment to decentralized staffing approaches. In this effort, some reforming administrations are wrestling with the inherent contradictions between private sector models and public service values. In the U.K, for example, questions have been raised about the role of the traditional corps of elite administrators now occupying key positions in gradually shrinking central departments vis a vis the new executing agency staff. The latter are hired increasingly on fixed-term contracts (mainly higher level professionals) and, it is intended, more and more at market- determined salaries. The central elite corps, on the other hand, has traditionally been revered for its policy skills; it continues to be dominated by recruits with Oxbridge humanities degrees. If the social, administrative, and functional cleavages between this core and the new staff in executing agencies is not resolved, policy and implementation could become dichotomized and uncoordinated. In Sweden, the higher civil service faces the same tensions. While higher civil servants continue to be paid in compressed pay scales, senior managers in agencies are converting to fixed contracts, with individually determined pay. Thus heads of large agencies are sometimes paid more than highly ranked civil servants in central core ministries with agency oversight responsibilities. New Zealand and Australia face similar quandaries. With the introduction of the "market economy model of management" in New Zealand, departments are increasingly hiring on fixed contracts at market competitive salaries. Conforming to private sector norms, staff loyalties increasingly lie with individual departments rather than the public service as a whole. In an attempt to "have it both ways," the State Services Act sought to combat the threat of departmental atomism and the potential erosion of the public service ethos through the creation of the Senior Executive Service. A select, cross-department corps appointed by the individual CEs, but with State Services Commission (SSC) approval, SESers are retained on fixed-term contracts with centrally determined salaries for up to five years. SES members receive common training from the SSC. The result is that SESers are sometimes earning less than other managers in the departments to which they are assigned. The SES concept, in consequence, is viewed with frustration and confusion. Australia's Senior Executive Service was created in the midst of wide pay and grading reforms, isolating the top levels of the former higher civil service from lower echelons in order to decompress salaries and restore purchasing power to higher management levels through the introduction of a separate category of senior specialists. Salaries for department heads --not SES members-- are supposedly set according to market criteria (as opposed to rank or grade level) but cannot exceed ceilings set centrally for the SES. Although many of the SES experiences are too recent to evaluate fully, so far efforts to build strong interdepartmental senior executive corps with appropriate remuneration and the capacity for independent responsiveness to high-level policy makers have had only modest results. Preventing wage compression and erosion in the context of fiscal constraints has proven difficult, and 29 the inherent tensions between market-oriented managerialism and a public-spirited service-wide cadre have yet to be resolved in reformist administrations. Still, in the main SES experiments have succeeded minimally in establishing self-contained corps with superscale pay categories, introducing separate performance criteria and cross-agency training programs, and laying the basis for inter-agency rotation. Significantly, governments with quite distinct approaches to civil service management appear to share a commitment to continuing the struggle to develop elite corps. Lessons: Should developing countries create senior executive services to build and insulate an elite, professional cadre? While it is difficult to recommend the adoption by LDCs of an instrument which has met with such mixed results in developed contexts, the dramatic dearth of qualified higher civil servants in many developing administrations argues powerfully for the adoption of some mechanism to groom and motivate a strong, cross-service executive class. In this case, LDC civil services might profit more from the mistakes than the successes of advanced countries in designing their own SES programs. For example, in order to keep professional standards high and to avoid abuses, SES programs should not include political appointees.35 Conditions of service may be either with permanent civil service tenure or on a rfxed-term basis. If they are the latter, terms should be long enough to ensure continuity, and the remunerative and career rewards must justify the extra risk assumed by SES prospects in giving up civil service security. If SES members retain civil service permanency, salaries should be kept high through periodic review, and, if necessary, insulated from generalized erosion down the civil service ranks, since underfunding has been an Achilles' heel of SES programs in developed settings. But SES programs must aspire to grander objectives than maintaining or restoring higher civil service salary levels or decompressing wages. Unfortunafely, deteriorated pay conditions have kept some developing country SES experiences narrowly focused on compensation issues. Other attributes of elite cadre development should not be ignored. SES programs should be comprehensive, offering targeted training, greater career opportunities, and high prestige, in addition to better remuneration. In addition, while it has admittedly proved difficult in advanced countries to achieve interdepartmental mobility for SES members, such mobility remains an important goal for LDCs. 'he objective is to develop broad policymaking skills and to prevent uneven distribution of scarce expertise in government, leading typically to strong professional enclaves in high profile ministries such as finance and conversely weak ones in less prestigious sectors such as education. Finally, developed country experience emphasizes the difficulties in "getting it right" on the first (or even the second) try. SES design is an ongoing process, and it should build in possibilities for reassessment and reformulation. 35Keeping the SES pure of political patronage and corruption is a particularlv vexing problem in Latin America. for example, where politicized rather than professionalized civil services are the norm. Attempts to install a (vcry modified) SES arrangement in Pem with World Bank financing, for example. resulted in charges that the superscale salaries amounted to a slush fund for powerfully connected individuals. 30 C. MANAGING INCENTIES TO CIVIL SERVICE PERFORMANCE -- APPROACHES TO CLASSIEYING. PAYING AND PROMOTING STAFF Reward structures are central to performance motivation for civil servants. This section analyzes several advanced country approaches to key aspects of civil service incentive systems. It poses the following questions: What are some successful and/or innovative approaches to classifying and paying civil servants in developed countries? How should staff performance be assessed and linked to pay and promotion? What are the most constructive ways to assess staff performance? What strategies should developing countries follow in light of advanced country experience? How to Classify and Pay Civil Servants: Traditional Standards Vs. Flexibility Developed government systems for classifying and paying public service employees range from traditional approaches, which emphasize centrally determined standards with little room for individual deviation, to increasingly flexible and decentralized approaches. The following discussion briefly reviews the established conventions, looks at attempts to introduce flexibilities into pay and classification systems, and then considers the implications for civil service requirements in developing countries. Traditional civil services have relied mainly on unified pay scales with automatic, fixed step increments based on seniority and determined on a service-wide basis. Sometimes public pay scales were derived from private sector comparator surveys; sometimes they were formed in the context of collective bargaining with trade unions. The basic elements of this traditional system still operate for the majority of civil servants in many advanced countries, including Japan, France, the Netherlands, Canada, the U.S., Germany, and Korea. These systems classify positions according to differences in job duties and responsibilities and rank these positions into occupational groups. Each group has a unified salary schedule determined by the central personnel authoe tv, often on the basis of cost-of-living analysis. Recommendations for service-wide adjustments are usually made on an annual basis. Each grade has a flat salary amount and contains a number of incremental steps. For each year of service civil servants receive one step increase. Emplovees typically have guaranteed job tenure and progress through increments automatically. Occasional variations in this system stem from the role of trade union negotiations in setting wage levels for lower level staff. These systems pose a number of problems. They have usually been designed by technical experts in central personnel departments, not by managers who understand real line job requirements. Because traditional classification systems make it difficult for workers to exceed their job descriptions and develop new skills, they can discourage creativity and initiative. And because pay cannot be raised without promotion to the next grade, a "grade creep' phenomenon occurs. Many countries are introducing modifications in traditional classification systems. Alterations represent attempts to moderate the rigid standardization of conditions of service to reflect more accurately actual differentiation among jobs and employees. This is being done through a variety of mechanisms. The increasingly popular practice of "broad-banding," for example, encourages flexibility. Broad-banding means recruits enter the service in a salary range rather than at a fixed point, as was traditionally the case. Moving up the ladder, increments are neither automatic (i.e. through seniority) nor fixed. Staff can therefore be rewarded more flexibly. In the US an experiment 31 with broad-banding is being applied to laboratory scientists through the Navy Demonstration Project, though efforts to extend the approach to the general civil service have met with some obstacles. As a rule, broad-banding is accompanied by classification simplification and a reduction in the number of grades and job categories. In Australia, where broad-banding is used selectively, the 1987 pay agreement with government unions replaced over 100 white-collar office classification structures with a single, broad-banded, 8-level system. This was a significant political accomplishment, as broad- banding is often opposed by unions fearing the erosion of standardized norms. In general, broad- banding is thus more easily implemented at senior levels.3' Broad-bands are being used in Australia's new Senior Executive Service and in the Swedish higher civil service. New Zealand's civil service reform program has succeeded in instituting broad-banding for all civil service employees.37 Pay and classification systems have been made simpler and more flexible through a range of other mechanisms as well. "Multi-skilling", for example, is being used in Australia to increase mobility and staff deployment flexibility for lower and middle-level jobs. Through departmental workplace reviews, tasks are redesigned to include a much wider range of skills requirements, and staff are given a broader array of training opportunities. Undertaken with extensive union consultation, multi-skilling is reputed to have increased job satisfaction and worker productivity significantly. State governments in Australia have taken this flexible pay and work condition approach further. In New South Wales, innovations include greater use of permanent, part-time employment, and longer daily shifts traded off against shorter work weeks and job-sharing. These steps have in turn afforded managers more flexibility in paying staff. New South Wales is also offering flexibilities through "cafeteria-style super-annuation benefits which can be combined with and substituted for base pay to create individually customized remuneration schemes up to a pre-set ceiling. Supplementing basic salary packages with non-wage allowances represents another move toward pay flexibility. In Japan, an elaborate set of allowances can comprise as much as 50 percent of an employee's total compensation package.3' The merits of complicated non-wage allowance systems in third world countries are often justifiably assailed for their lack of transparency. But in developed administrations, where public wage and employment policies are generally under control, allowances can be used as instruments for flexibility; that is, they can be allocated or raised in prosperous times and lowered or retracted in lean years. Salary policy in Singapore, for example, intentionally retains such allowances over the basic wage as a built-in adjustment mechanism which accommodates budgetary uncertainties. 3'Broad-banding also brings with it, however, the need to introduce appraisal-feedback systems in order to evaluate and clasify staff property. 37Chief Executives and Senior Executive Service staff pay is set centrally on an individual contract basis. In place of the 40 to 50 raed salary scales that existed prior to reforms, there are now only frve salary ranges for lower level staff and two for professional and advisory staff. 35TIese include an an-ay of supplements which mirror those in the Japanese private sector. "dty" cost-of-living allowances, 'end-of-yeae allowances (the 13th month salary) and "diligence' allowances 32 Accommodating cost-of-livingdifferences amonggeographical regions through locality pay" is another form of differentiation. Locality pay is used on a limited basis in the U.K. (for selected agencies and through the London-weighting system), and in Sweden and Japan. Pay flexibilities to attract hard-to-recruit skills have also been tried on a limited basis in the U.S." Methods of pay determination also affect the degree of flexibility governments enjoy in setting civil service wage levels. In many countries centralized mechanisms &et public sector pay according to the "fair comparison" principle; iLe., surveys of private sector pay inform service-wide wage settlements, usually negotiated with organized labor. Lately, governments have been attempting to build more maneuvering room into these pay determination arrangements. In the Netherlands, for example, the Ministry of Home Affairs continues to negotiate national pay settlements with the unions, but pay is no longer tightly linked to private sector wage movements through the previous system of semi-annual adjustments. France has established a new formula for improvements in salary structure when economic growth exceeds three percent in a given year. Similarly, Singapore now administers a "fled-wage" system, adjusting pay with the economy's performance. In good years, the government pays middle and end-of-year bonuses to all civil servants. It also pays one-off incentive bonuses when economic growth exceeds threshold levels. The locus of public pay determination also has implications for the correspondence between civil servants' salaries and local market conditions. Increasingly, civil service pay is set at more decentralized levels, resulting in more heterogeneous pay outcomes. In some cases, this decentralization reflects the increased autonomy of agencies to hire at market rates. Sweden, for instance, has undertaken local market pay experiments in a few agencies. In other instances, decentralization signifies the erosion of labor's ability to engage government in centralized, across-the- board wage settlements; thus such settlements are negotiated at more disaggregated, decentralized levels. While countries such as the U.K. and Australia are moving gradually toward decentralized pay determination, New Zealand has moved even more decisively down this road. With the State Services Commission's assistance, lower level staff salaries are now negotiated between Chief Executives and relevant unions at the departmental level. Managers' salaries are negotiated individually, with guidelines (and, for SESers, wage ceilings) provided centrally. Some anomalies have occurred: chief executive salaries and fringe benefits are on average higher than those of ministcrs. They do not yet match private sector comparators, however. Lessons: The following considerations emerge from this discussion of LDC pay and classification systems. First, for many developing countries a powerful rationale for traditional, unified, fixed-increment classification and pay systems is that they appear to require less administrative intensity to install and maintain. Administrative simplicity is an important, perhaps overwhehning consideration for developing country public sectors with extremely severe human resource constraints. Nonetheless, the reality in most LDCs is that unified classification and pay structures are undermined 39The new 1991 Federal Pay Law authorizs agencies to set up "triple.R bonus programs for the recruitment, relocation, or retention of key personnel. Within OPM guidelines, agendces can establish their own procedures for the ailocation of bonuses which can amount to as much as 25% of base pay. TripleR recipients must agree in writing to stay on the job for at least one year. 33 and distorted by a range of informal circumventions (including allowances, grade creep, double employment, etc.) reflecting the naturally occurring differentiation in government bureaucracies. Moreover, innovations such as broad-banding can actually reduce the number of job classifications and simplify the grading structure in addition to providing needed flexibility in rewarding difficult-to- attract professional skills -- a major problem in many developing countries. Similarly, multi-skilling may use scarce skills more efficiently by allowing broadly trained staff to move easily among a range of tasks. In countries with highly mobilized labor movements, the relaxation of standardized norms and the decentralization of pay determination may meet poliical resistance. But Australia's experience suggests that close consultation with relevant trade unions can overcome this difficulty. The costs of nationally negotiated pay settlements to financially strapped LDCs make this experience worth examining more closely. As the New Zealand case suggests, decentralized classification and pay arrangements may also offer potential financial efficiencies and promote better line management. In developing countries, mechanisms such as locality pay may facilitate a more efficient distribution of civil service employees to far-flung regions which normally have difficulty attracting personnel. But relaxation of centralized controls poses h *ightened risks in developing countries where the fiscal balance is delicate, and political patronage abuses are commonplace.' Thus, flexibility and decentralization should be promoted in direct relation to the strength of personnel institutions at the center, the degree of central control over establishment management, and the quality and enforceability of central policy guidelines. Finally, developed country wage policy models linking civil service pay and allowances with shifts in the macro-economy could be highly instructive to LDC administrations that have traditionally used bonuses and allowances to compensate for deficiencies in the civil service remuneration package, or as political patronage or social welfare mechanisms. Singapore's strategic and conservative utilization of these instruments has enabled its government to defuse civil servants' salary expectations and has provided the necessary leverage to avert future financial strain. Assessing Performance: Links to Pay and Promotion A topical issue in advanced country CSM reforms is how best to relate rewards to performance in motivating civil servants. This discussion focuses on the role of performance in paying and promoting civil servants. To what extent are performance based pay and promotion conducive to improving civil service staff productivity? How are advanced countries tackling these questions, and what approach makes sense for LDCs? Although most advanced civil services were theoretically built on a merit system, remuneration policies traditionally have operated on the basis of automatic, seniority related increases rather than performance criteria. Despite increasing emphasis on performance-related pay and performance appraisal as an important tool in both public and private human resource management 'The Bolivian experience with decentralization of public sector pay suggests that systemic inequities and fiscal irresponsibility can result when central personnel organs relax control of government pay determination. In Bolivia, the chaotic, uncoordinated institutional set-up at the national level contributed significantly to the systemic disorder. See Barbara Nunberg, 'Bolma: A Review of Public Pay and Employment Issues", LAC divisional working paper. World Bank. 1990. 34 circles, it is noteworthy that some quite successful administrations do not stress performance in determining civil servants' pay. In Korea, France, and Japan, for the most part, pay is very little affected by individual performance. In Korea, for example, civil service pay is mainly determined by seniority progression. Although pay does vary among individuals through the allocation of allowances, these are not awarded on the basis of performance criteria. Despite formal adherence to performance pay, informal social conventions undermine the recognition of individual as opposed to group excellence. Thus, an annual merit bonus, theoretically given to the top 15 percent of performers, is in reality distributed on a annuai rotational basis to each department staff member, regardiess of performance. Each employee will thus receive a bonus at least ucice over a several year time-frame.41 Several governments have introduced some performance-based features into basically traditional pay systems. Singapore's salary scale, for example, consists of two elements: fixed annual increments based on seniority and minimaLly satisfactory performance; and "efficiency bars" officials must pass by examination to ascend the salary ladder. The Netherlands has also shifted to a modified fixed-increment system. Annual step increments can now be withheld from poor performers and a modified performance-pay system (introduced in 1989) covers all civil service employees. One-off bonuses for special achievements are also awarded. Such one-off performance bonuses have also been used, to mixed reviews, for the U.S. Senior Executive Service. Bonuses may be useful in that they reward specific accomplishments in a targeted way without permanently increasing the wage bill through pay raises. Critics assert, however, that bonuses can ultimately have a deleterious effect on staff morale. Not getting a bonus, they argue, has negative effects outweighing the positive effects on those who do receive a bonus. A few countries have moved to install full performance-pay systems in their civil services. This process in most cases is gradual, with coverage extending beyond a small elite core in only a few instances. The U.S. Performance Management and Recognition System, introduced in 1984 for middle and higher-level bureaucrats, represents one of the more fully developed systems. PMRS has suffered from inadequate funding, and studies reveal that although employees perceive increased links between better performance and higher pay, no correlation has been established between PMRS awards and subsequent individual or agency performance.42 The U.K is gradually moving toward a performance-pay system for all civil servants, although so far only higher civil servants as well as professional, scientific, engineering and other specialized personnel are eligible. For these officials, the annual within-grade increments depend on the ratings they receive in objectives-based performance appraisals. With diffusion of the agency model, some Framework Agreements have been designed by departments and agencies to develop 4tThMiS is possible, of course, mainly because, with cmreely low employment mobility, employees are confident they will be in the same place in twenty years and that the system can be trusted to deal with them fairly. Informants interviewed for this study claimed that high- tier performers probably receied such bonuses only slightly more often than their peers, but certainly not commensurately with their true performance. Although above average staff are identified by their superiors and groomed for choice assignments, this recopition is not explicitly linked to monetary incentives. 42See James Peny, "Lining Pay to Performance: The Controversy Continues", in Carolyn Ban and Norma Riccucci, eds., Public Personnel Management: Current Concens -- Future Challenges Longmanw New York, 1991. 35 their own performance-pay arrangements. Performance-related pay is an integral part of the more radical agency reform countries. New Zealand, for example, uses it for all staff levels. Performance ratings and position in the salary range determine performance pay increases.' A few examples of attempts to modify performance-pay schemes to reward group rather than individual performance are worth noting. The Agricultural Research Department in New Zealand allocates annual pay bonuses to divisions for distribution among competing work teams. Group awards are being used in limited parts of the French and Swedish administrations, and in the U.K. some Framework Agreements permit Next Step agencies to introduce alternative group performanwe-pay systems. The Vehicle Inspectorate Agency, for example, is currently experimenting with group bonuses. In the U.S., the OPM-managed Pacer-Share Demonstration Project is using a productivity gainsharing system to distribute a portion of productivity savings to employees of McClellan Air Force Base. The use of non-wage incentives to motivate civil servants is also emerging as a variant on performancc-pay approaches. Cash awards to employees offering suggestions to improve the organization or enable savings are one such mechanism. In the U.K, for example, departments provide financial incentives to encourage staff to make helpful suggestions. When a suggestion is implemented, staff receive as much as 10% of the savings obtained during the first year.4 These variants aside, the deliberate movement in many developed countries toward individual performance-related pay systems is proceeding despite unresolved questions about the costs and benefits of this approach. Though widely used in the private sector, the concept still represents a significant departure for most civil services. There are numerous claims for performance pay: it underpins the concept of accountability; it provides alternative mechanisms for staff rewards when promotion prospects are limited; it affords the public sector flexibility in competing with private markets for scarce skills; it builds civil servants' morale in an age of public sector bashing; it provides a rational means for allocating scarce personnel expenditures; and it lirovides a means, short of major public service pay adjustments, to redress private-public pay differentials and internal government wage compression. Critics of performance pay point to a range of flaws. The most serious critiques claim that performance pay over-emphasizes short-term performance over long-term strategic objectives; produces more demotivated losers than motivated winners; is plagued by measurement problems, biasing reward unrealistically toward quantitative and financial measures; encourages standardized norms that inadequately capture the differences in civil service jobs; and is management-intensive and costly. 43Awards for administrative and support staff range from 2% to 10% of base salary, awards for advisory and professional staff range from 2% to 20% of base salarv. There is no increment for poor performems. The Chief Executive's performance is reviewed annually by the State Services Commission against the terms of agreement that were negotiated during the previous year with the Minister. Lke other higher civil servants, CEs are on fixed-term contracts which may be terminated for unsatisfactory performance. In practice, however, contract termination happens rarely if at all. 44According to one survey, in 1981, these schemes resulted in average savings of over L760,000 in large departments with approximately 40,000 employees. 36 Lessons: Despite the inconclusive debate cited above, it is difficult to reject the notion of remuneration structures that relate rewards to performance. But without clear evidence, governments may end up basing decisions to implement performance related pay systems purely on social preferences. For the Bank, the question still looms as to whether this should be a major component of technical assistance in civil service management. The hedged response is that it is probably worth pursuing in the medium or longer run, but is not likely to be a top priority currently. The practical argument against performance pay stems from its costliness and management-intensiveness. It could also plausibly be argued that several industrial country civil service management systems perform very weli without performance pay, and that even its adoption in reforming administrations has been recent and gradual in nature, utilized for only small numbers of civil servants. Moreover, given limited resources, attention to performance pay may be seen as a trade-off against improvements in other parts of developing country CSM systems. For example, nurturing a rigorous recruitment system or building a solid establishment control system may have greater medium-term returns than installing performance-pay mechanisms. Fmally, socio-cultural resistance to individual rewards undermines performance-pay systems in several advanced countries and could well do the same in some developing contexts. This is less an argument for cultural relativism than a suggestion that recent efforts to shift rewards toward the kinds of group incentives described above may offer greater prospects for success and should be considered as an alternative instrument. Performance and Career Advancement The difficulties in introducing performance-based pay, notwithstanding, many advanced countries are making increasingly frequent use of some form of performance appraisal as a basis for career advancement and promotion. This experience may be worth examining for developing country adoption. Most performance appraisal-based promotion mechanisms are imperfect, suffering from many of the same measurement problems as performance linked pay mechanisms. But in many LDCs, the alternative bases for upward mobility, i.e., nepotism, seniority, and cronyism, are probably worse. Some approaches to performance-based promotion and some suggestions for developing country applications follow. Although all CSM systems examined here had some formal performance-promotion link, the degree of consistency between formal systems and the informal practices of career advancement varied. Tr'. .ional CSM systems exhibited a greater gap between formal and informal practice. Korea, Japan, and France were in this group. In Korea, a semi-annual official performance review rates employees on task performance, abilities, attitudes and integrity. The resulting performance rating contributes to but does not fully determine promotion decisions, which still consider seniority and training accomplishments as advancement criteria. In Japan, a formal system 37 of confidential performance reviews operates annually, but real practice mixes a complex set of considerations based on seniority, social relationships and merit.' - 4 , By contrast, other developed administrations are emphasizing the centrality of truly operational performance appraisal mechanisms to their advancement systems. In the U.K., a relatively participative, objectives-based system has been in use for a few years. All civil servants are evaluated annually in Job Appraisal Reviews which become part of an Annual Staff Report. Supervisors interview employees to discuss progress made against previously agreed upon work objectives, and to define new ones for the coming year. They also identify training needed to improve job performance. Career development prospects are also discussed. An 'open" section records the comments of both parties. Staff Reports are definitely used in promotion decisions in the U.K, as are interviews with candidates' former superiors. Although merit has long been the primary criterion for civil service appointments, this had traditionally been construed as some informal and subjectively determined combination of personal attributes and the quality of policy advice. Recent reforms have attempted to systematize evaluation and consequent promotion, increasing the importance of management skills to the success formula. Singapore takes one of the most innovative approaches to performance evaluation and promotion. Promotion is largely merit based, with considerable staff mobility among different ministries over a career. The government has moved from a confidential evaluation system to a participatory performance appraisal system developed by the Shell Oil Company. The performance appraisal has two parts. One reviews and assesses past job performance; the other assesses the management and leadership potential --the so called "helicopter capacity"-- of " - subject in an effort to identify and subsequently groom executive talent within the civil service. While not directly linked to specific promotions, the latter assessment predicts an individual's career trajectory. Though it runs the risk of becoming a self-fulfilling prophecy, this approach brings a helpful, forward-looking, strategic perspective to the career development and succession planning functions. New Zealand has forged the tightest links between performance appraisal and promotion. Using private sector models, the government has introduced and applied to all employee levels a participative, objectives-based system. Performance is measured against objectives and standards agreed upon at the beginning of the review cycle. Individual objectives flow from the corporate mission statement and agency business plan which form part of the Framework Agreement between Agency and Minister in New Zealand. The performance appraisal constitutes the building block of performance agreements between departments and ministers and is credited with strengthening accountability throughout the system. 4`The intricacies of the Japanese system show how formal performance-based promotion systems are adapted to or distorted by dominant systems of social preference, resulting, in this case. in a rather functional arrangement. Members of a recruit cohort progress more or less en masse up through the ranks. Distinctions among them are identified on the basis of informal evaluations agreed upon by various of their superiors, and these become the basis for decisions about assignments made at their grade level. This informal ranidng does not lead to promotion above the cohort. however, which would signal a loss of face all around. Indeed, as each cohort amves ensemble at the highest civil service grade, one cohort member is designated to the vice-ministerial level --the highest civil servicc post. By common consent and tradition, the entire group then resigns to assume positions in local government, run for the Diet. or move to pre- arranged sinecures in the private sector. Although mandatory retirement age is 60, this mass *esignation takes place in the middle-fifties. This mechanism has two functions: it saves face bv avoiding the embarrassment of one cohort member supervising another and, through a 'dignified' early retirement process. it provides mobility channels for more junior staff, preserving a youthful and vigorous civil service profile. 38 A voluminous literature documents the difficulties in designing performance appraisal systems that accurately and fairly measure individual performance. Thus, questions about the utility of this instrrent as the basis for decisions about career advancement abound. In particular, it is frequently asserted that performance appraisal cannot serve the dual purpose of staff assessment and career development, and that it is generally more appropriate to the latter. Fmally, some critics see performance appraisal as a return to 'Taylorism", breaking down human performance into artificially disaggregated, but ultimately immeasurable parts.4' Lessons: The above critiques, however, constitute fine-tuning - a luxury not always available to developing countries. In LDC administrations where standards have broken down, performance- based promotion systems can minimally offer countries broad guidelines for rewarding and promoting competence. This is useful even if informal behavior and preference patterns come to distort the formal system. As the Japanese case shows, a reasonable compromise between the formal and informal can yield a system which at least considers performance as one criterion for advancement. The feasibility of performance-based promotion arrangements may depend on the conditions under which they are installed. Experience suggests that support from executive leadership, employee participation in the design of the system, and the timeliness of supervisory feedback to employees are key to the establishment of successful systems. IV. CONCLUSIONS This paper has reviewed established practice and emerging innovations in civil service management in developed countries in order to glean lessons for developing ones. In some instances, discussion about altemative approaches revealed ongoing debates, leaving uncertain the optimal reform strategy. The position assumed here, though, is that even with imperfect information developing countries must choose some reform path to improve the management of their civil services. Rather than definitive conclusions, the guidance offered here constitutes at most hypotheses and at least "hunches" about what is likely to work under LDC conditions. Focusing on three principal areas of CSM, -- system organization and management, quality of human inputs, and staff incentives and performance -- the paper's main findings are summarized below: -- As a general reform principle, traditional, mainly centralized civil service management models provide the best starting point for many of the least developed country adminiEtrations. Agency approaches, requiring technological and human resource skills beyond the present capacity of many LDCs, may, however, provide models toward which some countries -- especially the better endowed -- can strive. Certain selected elements of agency reform might even now be successfully emulated by some developing administrations. But any countries contemplating comprehensive agency type reforms need to assess seriously the degree to which the requisite human 46For views on this issue, see, Robert Glen, "Performance Appraisal: An Unneiving Yet Useful Process," Public Peonnel Management. Vol. 19, No. 1. Spring, 1990, and. Charles Fox, "Employee Performance Appraisal: The Keystone Made of Clay," in Carolyn Ban and Norma Ricucci, eds., Public Personnel Manatement: Current Concerns - Future ChallengesA. Longman: New York. 1991. 39 and technical requirements can be met and to think strategically about the necessary steps to acquire them. -- Strong centralized institutions are usually necessary to support fledgling cil service management functions in developing countries. Shared power and good coordination among a limited number of organs at the cenier ensures necessary checks and balances. To the degree that capacity can be developed in line entities, some deconcentration is desirable however, for responsive, flexible personnel management. -- Certain CSM functions should be given higher priority than others. This wisdom is even more powerful in developing countries, where resources are scarce and starker choices about allocation need to be made. Two functions emerged as central: recruitment and establishment control. -- Establishment control provides the foundation for all other CSM functions. In those developing countries where this function has historically been poorly managed, centralized models are advisable. Developed country experience with strategic, forward manpower planning could be enormously instructive to many LDCs. -- Programs to improve the rigor of recruitment standards should receive high priority in developing countries. Despite serious pitfalls, some aspects of "mandarin" systems offer many advantages to LDCs building a professionalized civil service. They compensate for other system failings. Governments should take measures to moderate the elitist characteristics of the mandarin approach and, as appropriate, judiciously introduce market flexibilities to recruit scarce skills. -- Senior Executive Services, or similar programs to develop professional elite cadres, have proved difficult to design and implement in developed countries, but many flaws can be corrected in adapting them to LDC contexts where there is often an urgent need to groom higher level civil servants. Such services may complement recruitment rigor. These programs must be carefully designed to include career mobility and training as well as remuneration features, and exclusion of political appointments is essential to their credibility. -- Assuming necessary minimal, levels of establishment and budgetary control, unified pay and classification in developing countries could arguably be relaxed, following the lead of increasing numbers of advanced countries. The common tendency toward circumvention of existing unified norms in many developing civil services, along with the need to target scarce skills more directly with appropriate remuneration packages, reinforces this view. -- The benefits of performance pay remain to be demonstrated in developed country public sectors, and the management requirements and costs of installing such systems can be considerable. Resistance may subvert efforts to effect performance pay arrangements. Given these difficulties and the urgency of other CSM tasks for LDC administrations, lower priority should be given to this reform at present. Performance-related promotion systems, on the other hand, even if imperfectly implemented, can help move civil service values toward standards of competence and merit. 40 Suggestions for Further Work These conclusions represent best guesses for LDC policy directions based on an initial, broad review of administrative experience in some advanced industrial countries. The intention was not to carry out rigorous research but to begin to elevate the quality of Bank operational advice on civil service management by flagging topical issues in worldwide administrative reform. Naturally, this work has raised a series of questions that require systematic investigation in order to move beyond the realm of speculation. The following are a few suggestions for further inquiry: 0 Rigorous comparisons of civil service performance across countries do not exist due to the absence of suitable methods for measuring civil service performance, the lack of comparable cross-national data on most questions, and the inherent difficulties in controlling for a range of intervening variables such as politics and culture. Research to develop a methodology for measuring civil service performance would be a valuable contribution in trying to determine which systems and approaches work best under which conditions. Initial comparative work might begin by developing performance measures for a single, relatively simple function carried out by most central govermments. * Research on the relationship between pay and performance in the public sector is not well advanced, and often policies are adopted on shaky empirical bases. Further work on this issue, particularly in developing country contexts, would be useful. * The effect of private-public pay differentials on civil service management functions could not be determined in this project and would require significant data generation and collection. Comparative research might be worthwhile in this area. * The relative effectiveness and efficiency of agency reform systems versus traditional civil services need to be studied rigorously before credible policy prescriptions can be extracted for developing countries. Given the recent origin of many of the reforms, such research might be appropriately undertaken after a suitable period elapses. 0 Given the urgent need in many developing countries to build governmental capacity, a more active form of applied research on civil service management topics might be an appropriate parallel approach to those suggested above. Pilot experiments could be carried out in LDC administrative "laboratories" to examine the relative merits of say, different recruitment, pay and classification, or performance appraisal systems on a comparative basis. While the actual 'experiments' in one or another civil service management improvement in a particular country could be financed as a lending operation, the exercise could simultaneously be set up as comparative, cross-national research by undertaking baseline studies and establishing performance criteria for ex- post evaluation. 41 BIBLIOGRAPHY Adamolekun, Ladipo. Issues in Development Management in Sub-Saharan Africa. EDI Policy Seminar Report. Washington, D.C.: World Bank, 1989. -. Public Administration: A Nigerian and Comparative Perspective. London: Longman, 1983. Altshuler, Alan, and Marc Zegans. "Innovation and Creativity: Comparisons between Public Management and Private Enterprise." Cte February 1990, pp. 16-24. Aspegren, Lennart. "Responsiveness in the Swedish Public Sector." Stockholm: Ministry of Public Administration, October 1987. Aucoin, Peter. "Administrative Reform in Public Management: Paradigms, Principles, Paradoxes and Pendulums." Govemance: An International Journal of Policy and Administration 3, 2 (April 1990): 115-137. Baker, John. "Management Improvement: the Human Dimension." Canberra Bulletin of Public Administration 59 (August 1989): 54-61. -. Some Private Sector Models for New Directions in Public Sector Human Resource Management: An Australian Public Service Perspective. Senior Executive Staffing Unit Occasional Paper No. 2. Canberra: Public Service Commission, March 1989. -----. "Improving Public Sector Performance: A World View." Canberra, 1990. -----. "From Management to Leadership: A Comparative Perspective on Leadership in the Australian Public Service." Australian Joumal of Public Administration 48, 3 (September 1989): 249-64. Ban, Carolyn. "The Realities of the Merit System." In Public Personnel Management: Current Concerns. Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, pp. 17-28. -----. "The Navy Demonstration Project: An 'Experiment in Experimentation.'" In Public Personnel Management: Current Concerns. Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, pp. 31-41. --, and Norma M. Riccucci. "Personnel Syste.ms, Labor Relations, and Government Performance." Paper prepared for the Commission on the American State and Local Public Service, Washington, D.C., September 1991. -----, and Sue Faerman, and Norma M. Riccucci. "Productivity and the Personnel Process." In Public Productivity Handbook. Ed. Marc Holzer. San Francisco: Jossey-Bass, forthcoming, 1992. 42 - and Harry C. Redd Im. "The State of the Merit System: Perceptions of Abuse in the Federal Civil Service." Review of Public Personnel Administration, forthcoming, 1991. -----, Edie Goldenberg, and Tony Marzotto. "Controlling the U.S. Federal Bureaucracy: Will the SES Make a Difference?" In Strategies for Administrative Reform. Ed. Gerald E. Caiden and Heinrich Siedentopf. Lexington: D.C. Heath and Company, 1982. Bargas, Didier. "Principales Initiatives Recentes de la Direction G6ndrale de l'Administration et de la Fonction Publique dans la Domaine de la Gestion des Ressources Humaines." Paris: Ministere de la Fonction Publique et des Reformes Administratives, October 30, 1989. Barzelay, Michael, and Babak J. Armajani. "Managing State Government Operations: Changing Visions of Staff Agencies." Journal of Policy Analysis and Management 9, 3 (1990): 307-338. Binder, Alan S., ed. Paying for Productivity. Washington, D.C.: The Brookings Institution, 1990. Bogdanor, Vernon. "Civil Service Reform: The Problem of Making Whitehall Accountable." Financial Times March 8, 1990, p. 21. Boston, Jonathan. "Some Reflections on the Performance Agreements of Chief Executives and the Accountability of the Executive in New Zealand." Canberra Bulletin of Public Administration 61 (July 1990): 13-25. -----. "The Future Structure and Role of Central Agencies." Public Sector (New Zealand) 12, 4 (1989). Bridge, Jane. "Human Resources Development in the 90's: New South Wales Govemment Strategies." Paper presented at the International Sympiosium on Human Resources Development in the 90's, Beijing, November 6-8, 1990. Campbell, Colin. Government Under Stress: Political Executives and aey Bureaucrats in Washin. London, and Ottawa. Toronto: University of Toronto Press, 1983. -, and B. Guy Peters, ed. Organizing Governance -- Goverming Organizations. Pittsburgh: University of Pittsburgh Press, 1988. Carnevale, David. "Florida's Senior Management System: Performance in a Decentralized Administrative Context." Public Personnel Management 18, 1 (Spring 1989): 65-74. Causey, Mike. "The Federal Diary." Washington Post. Daily column on developments in the U.S. Civil Service, September 1990 - December 1991. Chapman, Richard A. "'The Next Steps': A Review." Public Policy and Administration 3, 3 (Winter 1988): 3-10. Codd, M. H. "Federal Public Sector Management Reform in the 1980s: Recent History and Current Priorities." Canberra, November 1990. 43 Coopers & Lybrand Consultants. "Evaluation of the Effectiveness of the New South Wales Senior Executive Service." Stage I Issues Paper. October 1990. Daley, Dennis M. "The Civil Service Reform Act and Performance Appraisal: A Research Note on Federal Employee Perceptions." Public ersonne Management 19, 3 (Fal 1990): 245-251. ------ "Performance Appraisal in North Carolina: A Profile of Municipal Personnel Practices." Paper prepared for the Annual Meeting of the American Political Science Association, San Francisco, August 30 - September 2, 1990. Davies, Christopher. "Administrative Staff Control Systems in the British Civil Service." International Joumal of Public Sector Management 2, 2 (1989): 5-15. de Cotret, Robert. Public Service 2000: Status Report. Ottawa: Treasury Board of Canada, April 30, 1990. de Merode, Louis. "Civil Service Pay and Employment Reform in Africa: Selected Implementation Experiences." Washington D.C.: World Bank, Africa Technical Department, June 1991. Department of Finance (Australia). Australian Public Service Statistical Bulletin B: 1989-90. Canberra, 1990. Department of Industrial Relations (Australia). Second Report on Administrative Service Officer Structure: Progress Achieved in Agencies on Work Organization and Job Redesign. Canberra, February 1990. Dixon, Daryl. Senior Executive Service: Getting Most Advantage from Your Remuneration Package. Sidney: Premier's Department, New South Wales Superannuation Office, 1990. Duffy, Hazel. "Executive Agencies: Bringing Civility to the Civil Service." Financial Times, May 9, 1990, p. 18. Ecole Nationale d'Administration. Presentation Generale des Etudes Paris, January 1991. -----. "Enseignments de gestion publique: module de gestion des ressources humaines dans l'administration." July 1990. The Economist. December 21, 1991, pp. 66-67. Efficiency Unit. Making the Most of the Next Steps: The Management of Ministers' Departments and Their Executive Agencies. Report to the Prime Minister. London: HMSO, May 1991. -----. Managing Scrutinies: A Guide for Ministers and Managers. London: Efficiency Unit, no date given. -----. How to Do a Scrutiny: A Guide for Scrutineers and Their Teams. London, Efficiency Unit, no date given. 44 Ehrenworth, Sheldon. "Excellence in Government: A Key to Competitiveness and Its Implications for Human Resources Management." Paper prepared for International Symposium on Human Resources Development in the 90's, Beijing, November 6-8, 1990. "Extra 1/4-Mth Bonus for Public Sector." The Sunday Times (Singapore), November 11, 1990, p. 1. Falcon, David. The United Kingdom Experience." Paper prepared for the Intemational Symposium on Human Resources Development in the 90's, Beijing, November 6-8, 1990. "Life after Thatcher. Trends in U.K Public Administration and Lessons for Developing Countries." Notes on lecture sponsored by Public Sector Management and Private Sector Development Division, World Bank, Washington, D.C, March 28, 1991. Fallon, Peter, and Luis Pereira da Silva. "Public Pay and Employment Issues in Mozambique." Washington D.C.: World Bank, 1991. Finkle, Arthur L, Herbert Hall, and Sophia S. Min. "Senior Executive Service: The State of the Art" Public Personnel Management Journal 10 (Fall 1981): 299-312. Flynn, Andrew, Andrew Grey, and William Jenkins. "Taking the Next Steps: The Changing Management of Govermnent." Parliamentary Affairs 43, 2 (April 1990): 159-178. Fox, Charles. "Employee Performance Appraisal: The Keystone Made of Clay." In Public Personnel Management: Current Concerns, Future Challenges. Ed. Carolyn Ban and Norma Riccucca New York: Longman, 1991, pp. 58-72. Framework Documents. Various executive agencies in the United Kingdom. London: HMSO. Fry, Geoffrey. 'The Thatcher Government, the Financial Management Initiative and the 'New Civil Service.' Public Administration (UK) 66 (Spring 1988): 1-20. ----, and others. "Symposium on Improving Management in Government." Public Administration (U.K) 66 (Winter 1988): 429-445. Fudge, Colin, and Lennart Gustafsson, "Administrative Reform and Public Management in Sweden and the United Kingdom." Public Money and Management, Summer 1989: 29-34. Garrett, John. "Managerial Needs Make Possible Reforms Whitehall Dismissed as Impractical: The Accountable Civil Servant." Management Today November 1984, pp. 75-79. Gilbert, G. Ronald, and Ardel E. Nelson. "The Pacer Share Demonstration Project: Implications for Organizational Management and Performance Evaluation." Public Personnel Management 18, 2 (Summer 1989): 209-25. Glen, Rotxrt. "Performance Appraisal: An Unnerving Yet Useful Process." Public Personnel Management 19, 1 (Spring 1990): 1-10. Gormley Jr., W. Taming the Bureaucracy. Princeton: Princeton University Press, 1989. 45 Greenwood, John R., and David J. Wilson. Public Administration in Britain. London: George Allen & Unwin, 1984. Gustafsson, Lennart. "Renewal of the Public Sector in Sweden.' Public Administration (UK) 65 (Summer 1987): 179-191. Haggard, Stephan, Byung-Kook Kim and Chung-in Moon. 'The Transition to Export-Led Growth in South Korea, 1954-66." PRE Working Paper 546. Washington, D.C.: World Bank, 1990. Hale, Sandra J. and Mary M. Williams, eds. Managing Change - A Guide to Producing Innovation from Within. Washington, D.C., The Urban Institute Press, 1989. Heady, Ferrel. Public Administration: A Comparative Perspective. 3rd. rev. ed. New York: Marcel Dekker, 1984. Hennessy, Peter. 'Agency Chiefs Get the Chance to Grasp Torch of Freedom." The Independent. May 21, 1991. ----- '"Mandarins Maintain an Immobile Profession." The Independent January 15, 1990, p. 7. -. Whitehall. New York: The Free Press, 1989. Her Majesty's Treasury. Civil Service Statistics. 1988-89. London: HMSO, 1990. ----- -"Next Steps Agencies: Setting Targets and Measuring Output and Performance -- A Technical Note." Draft, London, January 1990. -----. "Pay and Management Flexibilities." London, August 1989. -----. An Introduction to Staff Inspection: Its Origins. Aims and Practice. London: Staff Inspection and Evaluation Division, June 1983. Holzer, Marc. "Attracting the Best and Brightest to Government Service." In Public Personnel Management: Current Concerns, Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, pp. 3-16. Hood, Christopher. "De-Sir Humphreyfying the Westminster Model of Bureaucracy: A New Style of Governance?" Governance: An International Journal of Policy and Administration 3, 2 (April 1990): 205-214. -----, Andrew Dunsire, and Lynne Thomson. "Rolling Back the State: Thatcherism, Fraserism and Bureaucracy." Governance: An Internationial Journal of Policy and Administration 1, 3 (July 1988): 243-70. House of Commons. Progress in the Next Steps Initiative. Eighth Report of the Treasury and Civil Service Committee. London: HMSO, 1990. 46 -. Developments in the Next Steps Programme. Fifth Report of the Treasury and Civil Service Committee. London: HMSO, 1989. Civil Service Management Reform: The Next Steps. Eighth Report of the Treasury and Civil Service Committee. London: HMSO, 1988. Huddleston, Mark W. The Senior Executive Service: Problems and Prospects for Reform." In Public Personnel Management: Current Concerns. Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, pp. 175-89. Hughes, Colin. "Whitehall Learns to Take a Business-Like Approach." The Independent October 31, 1990, p. 4. Hunn, D. K "New Zealand Senior Executive Service: Contract or Career Service." Can1mrra Bulletin of Public Administration 61 (July 1990): 158-60. Imbert, Magali. "Statistiques de l'Emploi Public dans les Pays d'Europe de l'Ouest.' Paper prepared by the International Labour Organization's Bureau of Statistics and presented at the Third General Assembly of the International Association of European Civil Service Meetings, Foix, France, September 19-20, 1991. Improving Management in Government -- The Next Steps Agencies: Review 1990. Report to Parliament by the Prime Minister and Minister for the Civil Service, and the Minister of State, Privy Council Office. London: HMSO, October 1990. International Labour Office. World Labour Report 1989. Vol. IV. Geneva, 1989. General Report of the Joint Committee on the Public Service. Report I. Geneva, 1988. Ives, Dennis. "International Reforms in Personnel Management -- The Australian Experience." Address to the International Personnel Management Association Training Conference. Philadelphia, October 1991. Jabes, Jak, and David Zussman. "Motivation, Rewards, and Satisfaction in the Canadian Federal Public Service." Canadian Public Administration 31, 2 (Summer 1988): 204-225. Jenkins, Kate, Karen Caines, and Andrew Jackson. Improving Manamement in Government: The Next Steps. Report of the Efficiency Unit to the Prime Minister ("Ibbs Report"). London: HMSO, 1988. Johnson, Nevil. "Change in the Civil Service: Retrospect and Prospects." Public Administration (Um 63 (Winter 1985): 415-433. Kearney. Richard, and Kathy S. Morgan. "Longevity Pay in the States: Echo from the Past or Sound of the Future." Public Personnel Management 19, 2 (Summer 1990): 191-200. Kemp, Peter. "Next Steps for the British Civil Service." Govemance: An International Journal of Policy and Administration 3, 2 (April 1990): 186-196. 47 Kim, Joong-Yang. "Civil Service Reform in Korea." Public Personnel Manament 14 (Pail 1985): 253-258. Kim, Paul. Japan's Civil Service Svstem: Its Structure. Personnel and Politics. Westport: Greenwood Press, 1988. Kirkpatrick, David. "It's Simply Not Working." Fortune. November 19, 1990, pp. 179-96. Kdnig, Klaus, Hans Joachim von Oertzen. &3d Frido Wagener, ed. Public Administration in the Federal Republic of Germany. Th- Netherlands: Kluwer-Deventer, 1983. Krauss, Ellis S., and Michio Muramatsu. "he Japanese Administrative Elite." Paper presented at the Annual Meeting of the American Political Science Association, San Francisco, August 30 - September 2, 1990. La Fonction Publihue de I'Etat en 1987: Rapgort Annuel du Premier Ministre. Paris: La Documentation Francaise, 1988. La Formation au Management des Ressources Humaines. Rapport du groupe de travail, Paris, Minist6re de la Fonction Publique et des Reformes Administratives, April 1990. Landau, Martin. "Redundancy, Rationality and the Problem of Duplication and Overlap." Publk Administration Review 29, 4 (July-August 1969): 346-58. La R6novation de la Grille de la Fonction Publique: Accord du 9 FEvrier 1990. Paris: Service d'Information et de Diffusion du Premier Ministre, 1990. Latimer, George, and Linda K. Carlisle. 'Personnel Modernization in Saint Paul: A Model for Participation and Consensus Building." National Civic Review 74 (September 1985): 365-371. Lazar, Avrin. "Experiential Learning: A New Emphasis in Human Resource Development." Paper prepared for International Symposium on Human Resources Development in the 90's, Beijing, November 6-8, 1990. "Les Fonctions Publiques dans l'Europe des Douze. Revue Franoaise d'Administration Publigue 55 (July/September 1990). Special issue on EEC country civil services. Levine, Charles, ed. The Unfinished Agenda for Civil Service Reform: Implitations of the Grace Commission Reportt Washington, D.C.: The Brookings Institution, 1935. Linde, Claes, and Peter Ehn. "The Swedish Administrative Elite." Paper prepared for the Annual Meeting of the American Political Science Association, San Francisco, August 30 - September 2 1990. Major, John. The Citizen's Charter. London, July 1991. Management and Coordination Agency (Japan). Annual Reort 1990. Chapters I-IV. Trans. Institute of Administrative Management. Tokyo: LAM, March 1991. 48 Martin, John. "Remaking the State Services." In TheF:hL&bw J1Le.WnU (2d. ed.). Ed M. Holland and Jonathan Boston. Auckland: OVP, 1990, pp. 123-39. Mascarenhas, R. C. "Reform of the Public Service in Australia and New Zealande" Wellington: Victoria University, October 1988. Megaw, Sir John. Inauiry into Civil Service Pay. London: HMSO, 1982. Middlemass, Miles. "Human Resources Management in the State Sector in New Zealand." Paper presented at the International Symposium on Human Resources Development in the 90's, Beijing, November 6&8, 1990. Ministy of Fmance (Singapore). Public Sector Wage Reform. Circular Minute No. 9/88. Singapore, May 6, 1988. Ministry of Finance (Sweden). "Renewing the PubUic Sector." Excerpts from the 1988 Revised Budget StatemenL Ministry of Government Administration (Korea). The Civil Service of the Republic of Korea Seoul, October 1988. Ministry of Public Administration (Sweden). "Brief on National Personnel Policy Bill (1984/85)." National Personnel Authority (Japan). Annual REpo 1989. Tokyo, 1990. -. Handbook on Japan's Civil Service: Statistical iew Tokyo, 1990. Newcomer, Kathryn, and others. "'Me Presidential Management Internship Program: Looldng Backward and Moving Forward. Public Administration Review 49,4 (July/August 1989). Newland, Chester. The American Senior Executive Service: Old Ideals and New Realities." International Review of Administrative Sciences 54 (1988): 625.60. "Next Steps Makes Royal Progress." The Independent November 1, 1990, p. & Nunberg, Barbara. Public Sector Pa_y and Empljmnt Rfm: A Reve of Worl Bank Experience. World Bank Discussion Paper 68. Washington, D.C.: World Bank, 1989. -----. Bolivia: A Review of Public Pay and Employment Issues." Washington, D.C.: World Bank, Country Economics Department, 1990. ----, and John Nellis. "Civil Service Reform and the World Bank." PRE Working Paper 422. Washington, D.C.: World Bank, Country Economics Department, 1990. OECD. Economic Outlook: Historical Statistics. 19601989. Pads, 1991. OECD in Figures. Paris, June/July 1991. 49 -... Labour Force Statistics: 1968-1988. Paris, 1990. -....Flemble Personnel Management in the Public Service. Paris, 1990. - . Public Management Developments: Survey 1990. Paris, 1990. -. "Remuneration of Senior Civil Servants in Sweden." Paper prepared for Meeting on Remuneration of Senior Civil Servants, Paris, December 11-12, 1989. -. Recent Trends in Performance Appraisal and Performance-Related Pay Schemes in the Public Service. Public Management Studies No. 4. Paris, 1988. -. Serving the Economy Better. Occasional Papers on Public Management, Paris, 1991. Office of the Minister for the Civil Service. "Succession and Career Planning for Senior Appointments in the Home Civil Service." London, July 1990. -. "Appraisal of Performance and Potential in the Senior Open Structure." London, April 1990. -----. "Management of the Senior Open Structure of the Home Civil Service." London, 1990. Osborne, David. "Ten Ways to Turn D.C. Around." Washington Post Magazine. December 9, 1990, pp. 1943. Oxcley, Howard, and others. "The Public Sector: Issues for the 1990s." Working Paper No. 90. Paris: OECD, Department of Economics and Statistics, December 1990. Ozgediz, Selcuk. Managing the Public Service in Developing Countnies: Issues and Prospects. World Bank Staff Working Paper No. 583. Washington, D.C.: World Bank, 1983. Paul, Samuel. Training for Public Administration and Management in Developing Countries: A Review. World Bank Staff Working Paper No. 584. Washington, D.C.: World Bank, 1983. -----, John Iclds and Jacob Levitsky. "Educating Managers for Business and Government." PPR Worldng Paper 126. Washington D.C.: World Bank, Country Economics Department, 1988. "Performance-Related Pay Schemes for Public Sector Managers: Country Report of the Netherlands," no date given. Perry, James L. "Linking Pay to Performance: The Controversy Continues." In Public Person_el Management: Current Concerns. Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, pp. 73-86. -----, Beth Ann Petrakis, and Theodore K. Miller. "Federal Merit Pay, Round II: An Analysis of the Performance Management and Recognition System." Public Administration Review (January- February 1989): 29-37. 50 Plowden, William. "What Prospects for the Civil Service." Public Administration (M 63 (Winter 1985): 393-414. Pougnaud, P. ";es Systbmes de Rdmun6ration au Rendement pour les Cadres de la Fonction Publique: R\apport de l'Expert Franqais." Paper prepared for OECD Meeting on Performance Pay Systems, February 1991. -. "Situation des Remunerations des Dirigeants et Cadres Sup&rieurs de l'Administration d'etat." Fiche d'6tude, October 1989. Priest, Dana. "Higher Starting Pay Hastens U.S. Hiring, GAO Fmds," Washington Post November 15, 1991. "Panels Urge Flexibility in Pay System." Washington Post November 27, 1991. -. "More Government Workers are Staying Put," Washington Post. November 27, 1991. Prokopenko, Joseph. Improving Public Service Productivity: Problems and Challenges. International Labour Office, Management Development Branch. Geneva, 1989. Public Service Commission (Australia). Government Response to the Report of the Senate Standing Committee on Finance and Public Administration on the Development of the Senior Ekecutive Service. October 1991. Public Service Commission (Singapore). Annual Report 1989. Singapore, 1990. Quah, Jon S. T. "Administrative Reform: Singapore Style." Intemational Review of Administrative Sciences 57 (1991): 85-100. -. "Toward Productivity and Excellence: A Comparative Analysis of the Public Personnel Systems in the ASEAN Countries." In Delivery of Public Services in Asian Countries: Cases in Development Admintration. Ed. Suchitra Punyaratabandhu-Bhakdi, and others. Bangkok: Thammasat University Press, 1986, pp. 238-272. Rabin, Jack, and others. Personnel: Managing Human Resources in the Public Sector. New York: Harcourt Brace Jovanovich, 1985. Raksasataya, Amara, and Heinrich Siedentopf, ed. Asian Civil Serces: Developments and Trends Kuala Lumpur: Asian and Pacific Development Administration Centre (APDAC), 1980. Rapoport, Carla. "Who Runs Japan?" Fortune August 28, 1989, pp. 113-114. Rehfuss, John, and Debra Furtado. "Executive Management Reform: The Case of Califomnia." State Government 55, 2 (1982): 43-49. Report of the Task Force on Public Sector Wage Reform. Singapore, December 23, 1987. 51 Reuveny, J. Civil Service Systems in Six States in Asia: Comparative Dimensins. Research Report No. 3. Tel-Aviv: David Horowitz Institute for the Research of Developing Countries, Tel- Aviv University, 1974. Richards, Sue. 'Management in Central Government: The Next Steps." Public Money and Maggment (Spring/Summer 1988): 87-90. Ridley, F., and J. BlondeL Public Administration in France. 2d. ed. New York: Bames & Noble, 1969. Roberts, Deborah D. 'A Personnel Chameleon Blending the Political Appointee and Careerist Traditions: Exempt Managers in State Government." In Public Personnel Management: Current Concerns. Future Challenges. Ed. Carolyn Ban and Norma Riccucci. New York: Longman, 1991, gp. 190-204. Robillard, Paul G. "A Macroscopic View of the Canadian Pubic Service in Tunes of Restraint." Paper presented at World Bank Seminar on the Changing Role of the State, Washington, D.C., March 23, 1989. Rocard, M[icheL "Circulaire du 23 fevrier 1989 relative au renouveau du service public." Journal Officiel de ,a Republique Francaise. February 24, 1990, pp. 2526-2529. Rosen, Bernard. "Revitalizing the Federal Civil Service." Public Administration Review 49, 5 (September/October 1989): 501-506. Scott, Graham, Peter Bushnell, and Nikitin SaHee. "Reform of the Core Public Sector: New Zealand Experience." Governance: An International Journal of Policy and Administration 3, 2 (April 199): 138-67. "SES: A New Approach to Pay and Incentives." Executive Bulletin (New South Wales) 1, 3 (Summer 1989): 3. Sherwood, Frank P., and Lee J. Breyer. "Fxecutive Personnel Systems in the States." Public Administration Review 47, 5 (September/October 1987): 410-416. Siedentopf, Heinrich, ed. Public Persoe anagen.ent in Asian Civil Services. Mainz: v. Hase & Koehler Verlag, 1983. Smith, Bruce, ed. The Higher Civil Service in Europe and Canada. Lessons for the United States. Washington, D.C.: Brookings, 1984. State Services Commission (New Zealand). "New Zealand: Changing the Public Service Culture: A Radical Approach." Wellington: Structures and Systems Division, 1990. -. Report of the State Services Commission. Wellington, July 1989. Statistics Sweden (SCB). Statistical Yearbook. Stockholm, 1988 and 1991. 52 Stevens, MichaeL "Innovations in Public Expenditure Management in OECD Countries." Washington, D.C.: World Bank, Country Economics Department, forthcoming, 1992. Sung-Yeul, Lee. "Administrative System of Korea." In Strategy and Management of National Development: The 3rd Executive Development Program for EROPA Countries. Seoul: Central Officials Training Institute, 1988, pp. 39-51. Synnerstrom. Staffan. "Personnel Management in Sweden." Paper presented at the International Conference on Human Resources Development in the 90's, Beijing, November 6-8, 1990. Tavemier, Gerard. "Sure Signs of Success." International Management. January 1979, pp. 36-39. Tellier, Paul M. "Public Service 2000: The Renewal of the Public Service." Canadian Public Administration 33, 2 (Summer 1990). Toole, Barry J. "'The Next Steps' and Control of the Civil Service: A Historical Perspective." Public Policy and Administration 4, 1 (Spring 1989): 41-52. Tsuji, Kiyoaki, ed. Public Administration in Japan. Tokyo: University of Tokyo Press, 1984. U.S. Office of Personnel Management. Annual Report on the Status of the Senior Executive Service. Washington, D.C., December 1988. -. The Senior Executive Service. Washington, D.C., February 1988. 'Trend of Federal Civilian Employment by Branch and Selected Agency, 1939-1990." Washington, D.C., 1991. Veilleux, GCrard, and Donald J. Savoie. "Kaflca's Castle: The Treasury Board of Canada Revisited." Canadian Public Administration 31, 4 (Winter 1988): 517-538. Volcker, Paul. Leadership for America: Rebuilding the Public Service. The Report of the National Commission on the Public Service. Washington, D.C., 1989. Wallis, Malcolm. Bureaucracy: Its Role in Third World Development. London: Macmillan, 1989. Whitcombe, Judy. "The New Zealand Senior Executive Service." Canberra Bulletin of Public Administration 61 (July 1990): 153-57. Winetrobe, Barry K. "'Next Steps' Executive Agencies." Background Paper No. 239. London: House of Commons Library Research Division, January 1990. Wise, Lois Recascino. "Dimensions of Public Sector Pay Policies in the United States and Sweden." Review of Public Personnel Administration 8, 3 (Summer 1988): 61-83. V orld Bank. The Reform of Public Sector Management: Lessons from Experience. Policy and Research Series 18. Washington D.C.: World Bank, Country Economics Department, 1991. 53 Wright, Deil S., and Yasuyoshi SauLrai Administrative Reform in Japan: Politics, Policy and Public Administration in a Deliberative Society." Public Administration Review 47 (March/April 1987): 121-33. Yanak, Francis V. 'Technology Utilization: The China Lake Demonstration Project." Paper presented at the International Symposium on Human Resources Development in the 90's, Beijing, November 68, 1990. Policy Research Working Paper Series Contact Tltie Author Date for paper WPS920 Economic Incentives and Point Raymond J. Kopp June 192 C. Jones Source Emissions: Choice of 37754 Modeling Fiatform WPS921 Road Infrastructure and Economic Cesar Queiroz June 1992 M. Laygo Development: Some Diagnostic Surhid Gautam 31261 Indicators WPS922 Central America at a Crossroads Sylvia Saborio June 1992 M. Colinet Constantine Michalopoulos 37044 WPS923 Listening to Firms: How to Use Firm- Andrew H. W. Stone Juna 1992 P. Infante Level Surveys to Assess Constraints 37642 on Private Sector Development WPS924 How Reduced Demand for Children Rafael Rofman June 1992 0. Nadora and Access to Family Planning 31091 Accelerated the Fertility Decline in Colombia WPS925 A General-Equilibrium-Based Social Ngee-Choon Chia June 1992 A. Cox Policy Model for C6te dlvoire Sadek Wahba 34778 John Whalley WPS926 Options for Reshaping the Railway Neil E. Moyer June 1992 B. Gregory Louis S. Thompson 33744 WPS927 General Equilibrium Effects of Andrew Feltenstein June 1991 C. Jones Investment Incentives in Mexico Anwar Shah 37699 WPS928 Pesticide Externalities, Comparative Nalin M. Kishor July 1992 D. Ballantyne Advantage, and Commodity Trade: 38004 Cotton in Andhra Pradesh, India WPS929 Managing Pollution Control in Brazil: Antonio Estache July 1992 A. Estache The Potential Use of Taxes and Kangbin Zheng 81442 Fines by Federal and State Governments WPS930 Participatory Development: Myths Robert Picciotto July 1992 A'Muhtasib and Dilemmas 84573 WPS931 How Much to Commit to an Exchange Alex Cukierman July 1992 R. Luz Rate Rule: Balancing Credibility and Miguel A. Kiguel 34303 Flexibility Nissan Liviatan WPS932 Interest Rates, Official Lending, Asli Demirgug-Kunt July 1992 K. Waelti and the Debt Crisis: A Reassessment Enrica Detragiache 37664 WPS933 Developing Country Capital Asli DemirguO-Kunt July 1992 K. Waelti Structures and Emerging Stock Markets 37664 Policy Research Working Paper Series Contact Title Author Date for paper WPS934 Public Hospital Costs and Quality Maureen A. Lewis July 1992 P. Trapani in the Dominican Republic Margaret B. Sulvetta 31947 Gerard M. LaForgia WPS935 The Precautionary Demand for Boum-Jong Choe July 1992 S. Lipscomb Commodity Stocks 33718 WPS936 Taxation, Information Asymmetries, Andrew Lyon July 1992 C. Jones and a Firm's Financing Choices 37699 WPS937 How Soff is the Budget Constraint Evan Kraft July 1992 CECSE for Yugoslav Firms? Milan Vodopivec 37178 WPS938 Health, Government, and the Poor: Nancy Birdsall July 1992 S. Rothschild The Case for the Private Sector Estelle James 37460 WPS939 How Macroeconomic Policies Affect Daniel Kaufmann July 1992 D. Kaufmann Project Performance in the Social Yan Wang 37305 Sectors WPS940 Private Sector Approaches to Karen G. Foreit August 1992 0. Nadora Effective Family Planning 31091 WPS941 Projecting the Demographic Impact Rodolfo A. Bulatao August 1992 0. Nadora of AIDS Eduard Bos 31091 WPS942 Efficient Environmental Regulation: Arik Levinson August 1992 WDR Case Studies of Urban Air Pollution Sudhir Shetty 31393 (Los Angeles, Mexico City, Cubatao, and Ankara) WPS943 Burden-sharing among Official and Asli Demirgug-Kunt August 1992 K. Waelti Private Creditors Eduardo FernAndez-Arias 37664 WPS944 How Public Sector Pay and Gail Stevenson August 1992 PHREE Employment Affect Labor Markets: 33680 Research Issues WPS945 Managing the Civil Service: What Barbara Nunberg August 1992 P. Infante LDCs Can Learn from Developed 37642 Country Reforms