i

Public ExpenditureAnalysis and

Management in theRepublic of Kariba:

A Case Study

Developing More Effective Waysto Control and ManageGovernment Resources

Louis A. Langlois and Robert P. Beschel Jr.with Rick Stapenhurst

The Economic Development Instituteof the World Bank

ii

The findings, interpretations, and conclusions expressed in thisdocument are entirely those of the author(s) and should not beattributed in any manner to the World Bank, to its affiliated or-ganizations, or the members of its Board of Executive Directorsor the countries they represent.

Copyright © 1998The International Bank for Reconstruction and Development

The World Bank enjoys copyright protection under protocol 2of the Universal Copyright Convention. This material may none-theless be copied for research, educational, or scholarly purposesonly in the member countries of the World Bank. Material inthis series is subject to revision. The views and interpretations inthis document are those of the author(s) and should not be at-tributed to EDI or the World Bank. If this is reproduced or trans-lated, EDI would appreciate a copy.

iii

Contents Foreword .................................................................... v

Introduction ............................................................. 1

Background Notes ................................................... 2

The Initial Meeting of the Task Force .................... 4Priorities and Expenditure Problems ................ 7The Cash Budget .............................................. 11

The O’Toole-Michaels Report .............................. 14The Policymakers ............................................. 15The Planners ..................................................... 16The Operators ................................................... 17Budget Systems and Timelines ........................ 17The Costing of Programs ................................. 20

Coping with the Wage Bill .................................... 20

Conclusion ............................................................. 24

Questions ................................................................ 25

iv

v

Foreword One of the princical objectives of EDI’s Governanceprogram is to help expand a government’s capacity toimprove service delivery to the public. It does this byadding an emphasis on changing mindsets and de-veloping leadership to the more traditional approachesof public sector capacity building of promoting con-ventional technical tools and skills-building at thecentral government level.

The Regulatory Reform and Private Enterprise di-vision of the Economic Development Institute of theWorld Bank (EDIRP) adapts its approach to the pro-motion of good governance to client country circum-stance. Thus, for example, it has facilitated publicawareness raising workshops at both a regional andnational level in Sub-Saharan Africa (Uganda, Tanza-nia, Malawi, and Mauritius), Eastern and Central Eu-rope (Ukraine and Georgia), Latin America (Boliviaand Nicaragua), and South Asia. Elsewhere, it has de-veloped and implemented workshops to strengthenthose institutions that play a role in curbing corrup-tion: parliaments (in Ghana, Ethiopia, and Uganda),the media (in Francophone Africa, Indian Oceancountries, Uganda, and Tanzania), and supreme au-dit institutions (in Ghana and Zambia).

Sound financial management systems are power-ful instruments for preventing, discovering, or facili-tating the punishment of fraud and corruption. Bud-get reform is an important element of public sectorreform, which in turn incorporates elements of skills-building and institutional strengthening.

Governments should match policy, affordability,structures with functions, but they should undertakeonly what they can do well, given their own resourceconstraints. Empirical studies suggest that corruptionafflicts the allocation of budget resources, especially

vi

those devoted to education. A well-functioning bud-get system can help ensure the matching of govern-ment policy with affordability and, in so doing, mini-mize the possibilities of “financial leakage” and at thesame time greatly assist the work of those “oversight”institutions, such as the auditor general, the Public Ac-counts Committee, and Parliament.

EDI, in collaboration with World Bank Operationsand Ministries of Finance in client countries has fa-cilitated workshops in reforming budget processes.This case study is an outgrowth of such workshops.Set in the fictitious country of Kariba, it concentrateson how senior officials can develop more effective waysto control and manage government resources. It high-lights the principal elements of a well-functioning sys-tem of expenditure management, including strategicfocus, transparency, predictability, accountability,affordability, accuracy, timeliness, flexibility, and effi-ciency in the use of resources.

Louis A. Langlois is the principal of LAB Interna-tional consultants and former director of the BudgetOffice in Canada. Robert P. Beschel Jr., is a Good Gov-ernance Advisor for the Asian Development Bank. RickStapenhurst is a Public Sector Management Specialistat the World Bank’s Economic Development Institute.The authors gratefully acknowledge the assistance ofMalcolm Holmes, Peter Miovic, Lemma Merid, EdCampos, and Brian Ngo in providing useful feedbackin the production of this case.

Teaching notes for this case study are also available.

Edward Phillips barely had time to prepare for the taskforce meeting. As the new director of the Budget Of-fice in the Republic of Kariba, he had just assumed hisformal duties last week—leaving a promising careeras professor of public finance at the national univer-sity for what he hoped would be a relatively brief stintin public service. His predecessor had been summarilydismissed on the grounds that he was neither able tobring any discipline to the resource allocation processnor to provide any useful advice on effectively man-aging the machinery of government.

When the Permanent Secretary of Finance urgedPhillips to accept the position, he had argued force-fully that Phillips’ knowledge and expertise could helpreform a situation that was far from optimal. The al-location of resources over the recent past did not re-flect the government’s priorities, and individual min-isters and their ministries were, for the most part, pur-suing their own agendas.

In an attempt to begin the process of reform, thepermanent secretary had recently appointed a taskforce to advise the government on upgrading Kariba’scurrent system of expenditure management. Its mem-bership included Mia Petrovic, deputy governor of theBank of Kariba; Louis Gerrard, auditor general;Patricia Barre, permanent secretary of the Ministry ofHealth; Colin McNamara, permanent secretary of theMinistry of Education; Peter Hurst, permanent secre-tary from the Cabinet Office responsible for policy; aswell as José Eduardo, permanent secretary from theCabinet Office responsible for the management of thepublic service. Phillips himself, as director of the Bud-get Office, would act as chair.

As he prepared for the first meeting of the task force,Phillips reflected upon the serious problems that were

Introduction†

† The circumstances presented within this case have been drawn from a compos-ite of actual country experiences. The characters are fictional.

confronting his country and the critical role of theBudget Office in addressing them.

Kariba, a former British colony, became independentin the early 1960s. Before independence, Kariba had asmall but fairly professional public service, adequateinfrastructure, a well-developed system of education,and a decent health care system. It was considered amiddle-income country with abundant mineral re-sources. Fertile land and a benign climate enabled pro-duction of a great variety of agricultural products forits own use as well as for export. Hydropower bothwas and is its primary energy source, with some sur-plus available for export. All petroleum and naturalgas are imported.

For more than 30 years after independence, Karibawas governed by a single party and president, whose at-titude toward the private sector ranged from indiffer-ence to outright hostility. At various periods, Kariba hasbeen confronted with guerrilla conflicts in several bor-dering states, which necessitated spending significantsums on national defense and foreign representation. Thepublic sector grew rapidly during the 1960s and 1970sand was viewed as the engine for national development.Prior to the 1980s, Kariba prospered, in large measurebecause of world demand for its non-renewable re-sources—in particular, its large copper deposits.

Copper prices declined precipitously in the early1980s, leading to a rapid drop in government revenues.Assuming, incorrectly as it turned out, that this de-cline resulted from short-term market fluctuations andnot long-term reduction in demand due to the grow-ing use of fiber optic cable, the government borrowedheavily abroad and continued to pursue an expansion-ary fiscal and monetary policy at home. Predictably,the country’s international indebtedness grew toalarming proportions. Consequently, several bouts ofhyperinflation resulted in a series of devaluations forthe local currency, the ngo.

BackgroundNotes

Against a backdrop of increasing macroeconomicinstability, a new government was elected in 1992. Itsought to reverse many of the policies of its predeces-sor, including adopting a tight fiscal and monetarypolicy to reduce inflation and stabilize the currency,and privatizing a number of loss-making state enter-prises. This approach lowered the inflation rate andreduced several of the more prominent drains uponthe state treasury.

Kariba currently has a population of roughly 10million, twice what it was at independence. Althoughits population density is relatively low by internationalstandards, at well over 3 percent its rate of growth isone of the highest in the world. Most inhabitants livein the urban areas. Many have a low level of educa-tion, live in shantytowns, and support themselves withintermittent low-skilled work.

Kariba has seen its position drop from middle-in-come status to one of the poorest countries in theworld. At independence, it had over US$1 billion inthe treasury, but the country’s current debt now standsat almost US$5 billion. Its per capita income has de-clined steadily over the last three decades to the cur-rent level of roughly US$550. Real growth has beeneither declining or holding at the same level for mostof the last 20 years, and real income is now roughlyhalf what it was in the 1960s. In spite of its fertile landand climate, the country is often a net importer ofbasic food stuffs.

After more than 30 years of one-party rule, the newgovernment has inherited a public service that is over-staffed, underpaid, demoralized, and unproductive.Chronic under-investment in public infrastructure hasresulted in poor quality roads, hospitals, schools, etc.The percentage of students enrolled in primary edu-cation is declining, and infant mortality is increasing.The crime rate is growing, and the ability of the po-lice to maintain law and order is being severely tested.

The mandate of the government expires later thisyear, and an election will be held in the fall. The

ruling party is expected to form the next govern-ment, as the other parties are fragmented, poorlyfinanced, and not well organized. Nevertheless, sup-port for the ruling party has been waning, and se-nior government officials are reluctant to jeopar-dize their standing among the Civil Service by en-gaging in any major downsizing at this time. Theinternational situation has improved, and there arecurrently no significant internal or external threats.However, the army remains powerful and, if pro-voked, could be a threat to the government.

Against this sobering background, Edward Phillipsconvened the first meeting of the Task Force on Ex-penditure Management. All of the task force mem-bers were dedicated professionals and the meetingbegan promptly at the appointed hour. After a briefround of introductions, Phillips outlined the na-ture of the task force’s mission: to restructureKariba’s current budget process to enhance fiscaldiscipline and to ensure that resources be allocatedto high-priority government programs.

He then asked Mia Petrovic, deputy governor ofthe Bank of Kariba, to offer some brief remarks re-garding the macroeconomic framework underwhich government fiscal policy and expenditurereform would have to take place.

Petrovic, a dynamic woman in her mid-40s, made a num-ber of points related to the consequences of the fiscaland monetary policies that had been pursued over thelast two decades. She noted that, as a consequence of thegovernment’s having relied on borrowed capital to bridgethe gap between revenues and outlays, a growing shareof revenue was now required simply to service the debt.The debt-to-GDP ratio had increased steadily over time,and by 1980 it amounted to 200 percent. It peaked atover 400 percent in the early 1990s and currently hovers

The InitialMeeting of the

Task Force

MacroeconomicFramework

around 185 percent. This, in her view, was high by anyinternational standard.

Petrovic noted that high annual deficits generallycreate two related sets of problems: an increase in thesize of the national debt, and inflation. She remindedher audience that the annual deficits from the previ-ous regime’s expansionary fiscal policy had resultedin an ever-increasing stock of government obligationsthat had to be serviced. These servicing costs repre-sented a non-discretionary demand on governmentresources. If the debt was not repaid at maturity, itmust be “rolled over” at the prevailing rates, and Karibawould be left hostage to exogenous shocks rather thanbeing able to fully control her own agenda. Excessivegovernment borrowing in the domestic market effec-tively “crowded out” other potential borrowers, therebystifling capital investment by the business sector. This,she opined, had profoundly damaged Kariba’s eco-nomic capacity and ability to compete internation-ally.

Since a significant portion of Kariba’s debt hadbeen financed through the creation of money,Petrovic reminded her listeners of the inflationaryconsequences of “monetizing” the debt. She notedthat Kariba had recently experienced several peri-

Debt-to-GDP

01980 ’85 ’87 ’89 ’90 ’91 ’92 ’93 ’94 ’95

100

200

300

400

500

%

ods of high inflation, which in turn resulted in sev-eral currency devaluations. Annual inflation ratesexceeded 50 percent for the first half of the 1990s,and they peaked at nearly 200 percent in 1992 be-fore abating. Inflation rates are currently in the 30percent range.

Petrovic noted that the initial effect of devaluation hadbeen to make exports more competitive. However, giventhat Kariba’s exporting businesses relied, in large mea-sure, upon importing equipment, the initial competitiveadvantages of devaluation quickly evaporated. Inflationhad a devastating effect on individuals whose income wasfixed. It tended to favor business and individual inves-tors in capital assets, while putting those who had in-vested in domestic financial assets at a disadvantage. Italso favored domestic debtors at the expense of domes-tic financial investors. In a brief aside, Petrovic voicedher opinion that inflation also had an implicit redistribu-tive effect on various expenditures. It tended to favor morepolitically sensitive programs—such as wages and de-fense expenditures—at the expense of less politically sen-sitive programs.

Petrovic concluded by underscoring the need forstrict fiscal and monetary discipline. She noted thatthe nation’s economic difficulties had resulted in a

Annual Inflation Rates

01990 1991 1992 1993 1994 1995

50

100

150

200

250

%

pronounced decline in the resources available to thepublic sector. The ratio of public expenditures to GDPdropped from about 40 percent in the 1980s to lessthan 20 percent today—a real decline of about 50 per-cent. In addition, a greater share of total expenditureswas required simply to service the debt. Although somerecent developments on the revenue side held prom-ise, it was unlikely that there would be any real in-crease in the level of aggregate government expendi-tures over the next two years. Ministries and agencieswould have to toe the line, improve their productiv-ity, and learn to get more from less.

After Petrovic finished, Peter Hurst, Permanent Sec-retary from the Cabinet Office responsible for policy,noted that evidently in the current circumstances thegovernment can no longer continue to fully fund allof its obligations. Priorities would have to be set. TheCabinet agreed, in principle, that scarce governmentresources would, after interest payments, be directedto the social, economic, and law and order sectors. Ex-penditures on programs that did not contribute di-rectly to either the economic or the social benefit ofthe country would be given a lower priority.

Patricia Barre—the tough and outspoken Perma-nent Secretary for Health—interrupted Hurst. Shetold him that the government had been espousingthis policy for the last three years, yet results hadunfolded quite differently. During this period, theshare of expenditures devoted to the high prioritysectors declined rather than increased. Expendituresdevoted to the social sector declined from 34 per-cent in 1988 to only 19 percent by 1995. Similarly,the share of expenditures going to the economicsector declined from 19 percent in 1986 to only 10percent in 1995. Expenditures in the law and ordersector declined from 6 percent in 1983 to only 3percent by 1995. Conversely, the share of expendi-tures devoted to defense and foreign affairs re-

Priorities andExpenditure

Patterns

Program Expenditures by Sector(percentage of total)

mained virtually unchanged over the 10-year pe-riod ending in 1995, while the cost of running thegovernment had more than doubled.

Although initially caught off guard by the perma-nent secretary’s intervention, Hurst quickly recovered.He acknowledged that there had occasionally beensome deviation between Cabinet policy and the ac-tual distribution of expenditures. However, he argued,the ministries and operating departments themselvesbore considerable responsibility for these deviations.The distribution of resources within sectors, for ex-ample, was often inconsistent with the government’sstated priorities.

Turning directly toward Barre, he noted that—inthe case of expenditures on health—the government’sstated objective was for universal access to affordablehealth care of good quality, close to the family. How-

tbeDecivres cimonocE laicoS

dnawaLredro

ngieroFsriaffa esnefeD

yrenihcaM.t’vogfo latoT

3891 71 61 33 6 2 11 51 001

4891 41 61 43 6 2 41 41 001

5891 81 51 33 5 1 51 31 001

6891 42 91 42 3 2 21 61 001

7891 81 11 92 4 4 61 81 001

8891 41 81 23 4 3 71 21 001

9891 7 81 32 3 2 61 13 001

0991 7 71 92 4 3 51 52 001

1991 31 21 22 3 2 41 43 001

2991 21 21 32 3 2 41 43 001

3991 41 21 32 3 3 41 13 001

4991 11 01 71 3 2 21 54 001

5991 51 01 91 3 3 11 93 001

ever, over the past decade, spending on primary healthcare had declined by almost a third. Moreover, an ever-increasing share of the health care budget had beendevoted to administration—mainly in the form ofwages—with the ratio of funds going to the largerhospitals remaining fairly constant. By comparison,the share of the budget directed to local institutionshad declined from 18 percent to only 5 percent.

Thus, while the demand for health care had beenrising—due to such factors as the rapid populationgrowth and increased public awareness—the qualityof health care had been declining, especially in localinstitutions and at the primary level.

Hurst maintained that health wass not unique inthis regard. Glancing at the Per manent Secretary foreducation, he noted that the government’s stated ob-jective was to improve both access to education as wellas the quality of primary education. Yet, in spite ofthe government’s stated objective, the portion of theeducation budget devoted to primary education con-tinued to decline from 52 percent to only 40 percentover a five-year period. By comparison, the portion ofthe budget devoted to secondary education increasedfrom 26 percent to 34 percent. Similarly, the budgetfor university education increased from 22 percent to25 percent of the education budget. Hurst also noted

09–0891egareva 1991 2991 3991 4991 5991

noitartsinimdA 8 01 11 41 71 12

seilppuS 01 41 71 61 41 9

snoitutitsnilacoL 31 81 61 21 9 5

slatipsohrojaM 92 52 42 52 82 92

htlaehrehtOsecivres 04 33 23 33 23 63

latoT 001 001 001 001 001 001

Health Expenditures by Percentage of Distrbution—A

that the cost per secondary school student was eighttimes higher than spending at the primary level andthe cost per university student was 254 times greater.

Finally, Hurst observed that, within the educa-tion budget, a growing share was required simplyto cover the cost of salaries and boarding, and a de-clining share was going to learning material. All ofthis underscored his original point that governmenthad to set priorities. Should Kariba opt for CATscans in the university hospital, or malaria treat-ment in the countryside? University education forthe elite, or primary education for all? These deci-sions had been made at the center, yet the minis-tries repeatedly failed to allocate their resources ac-cording to government priorities. In his view, thissituation had to change and it had to change quickly.

09–0891egareva 1991 2991 3991 4991 5991

yramirPevitneverp 73 43 03 53 42 42

yraitreT 63 04 43 53 63 83

noitartsinimdA 72 62 63 03 04 83

latoT 001 001 001 001 001 001

Health Expenditures by Percentage of Distribution—B

0991 1991 2991 3991 4991 5991

yramirP 25 05 35 14 14 04

yradnoceS 62 52 13 43 43 43

ytisrevinU 22 52 61 52 52 52

latoT 001 001 001 001 001 001

Education Expenditures by Percentage of Distribution—A

The Permanent Secretary for Education, ColinMcNamara, could not let criticism of his ministry passwithout a response. Strongly supported by the Per-manent Secretary for Health, he argued that all of theministries agree that any blame for the failure to allo-cate resources in line with Cabinet priorities restedsquarely with the government’s decision two years agoto move to a cash budget. This approach, in which thegovernment spent only the monies that it had on handor could borrow at prevailing market rates, had cre-ated tremendous disruptions in the ability of operat-ing agencies to plan strategically and to implementgovernment policy effectively.

McNamara and Barre reminded their colleaguesthat, to be assured that there was sufficient cash inthe bank to meet outstanding obligations, the Min-istry of Finance only disbursed funds on what wasessentially a day-to-day basis. Thus, while Cabinetmay have allocated resources and Parliament sub-sequently appropriated them, ministries were un-sure of the actual amount of money they wouldactually receive. In practice, disbursements variedconsiderably from the figures listed in the formalbudget, and some ministries received as little as 40percent of what they sought. Furthermore, even thesums the Ministry of Finance eventually disbursedwere released at inconvenient times in an apparentlyarbitrary and non-transparent fashion.

0991 5991

seiralaS 04 65

stsocgnidraoB 03 63

slairetamgninraeL 02 9

noitartsinimdalareneG 01 5

latoT 001 001

Education Expenditures by Percentage Distribution—B

The Cash Budget

Confronted with this uncertainty, ministries couldnot plan from one week to the next—and strategicmultiyear planning is a waste of time. Any funds theMinistry of Finance released were first used to takecare of the backlog of accumulated bills. Remainingfunds were quickly claimed by other commitments,regardless of actual operational requirements. All ofthe ministries knew, McNamara stated, that their un-used funds may be scooped up by the Ministry of Fi-nance at any moment.

Finally, McNamara revealed that, to maximize po-tential allocations from the Ministry of Finance, min-istries tended to submit overstated cost estimates. Anyrational official would do the same, he contended,when confronted with the dysfunctional incentive sys-tem maintained by the central agencies.

This notwithstanding, Auditor General Louis Gerrardreminded his colleagues of the rationale for moving to acash budget: In the early 1990s, the International Mon-etary Fund had become increasingly concerned aboutthe government’s fiscal and monetary policies and in-sisted that this move be a precondition for further aidfrom the international community. The cash budget wasintended to pressure the government into hold expendi-tures within the limits of current revenue. And it hadlargely been successful. Inflation is down and the cur-rency stabilized. The government—by setting prioritiesand making hard choices among competing programs—was meeting its deficit reduction targets.

Gerrard admitted, however, that the move to a cashbudget had resulted in a number of unintended nega-tive effects. In the first instance, funds had tended togravitate away from less politically sensitive sectors,such as the social and economic sectors, in favor ofdefense and foreign affairs. The cash budget had alsoresulted in a number of hidden “outstanding liabili-ties,” as ministries waited to pay creditors for goodsand services received until the Ministry of Finance ac-tually released the funds. Both expenditures and theannual deficit tended to be understated.

The auditor general noted that, from the perspec-tive of both his office and the Ministry of Finance, thetype of actions Permanent Secretaries McNamara andBarre described provided the rationale for closelymonitoring and controlling the ministries. Staff in thecentral agencies were convinced that, if left on theirown, the ministries would spend the funds allocatedto them without due regard for good cash manage-ment. Ministries frequently incurred commitments upto their authorized limit (and occasionally beyond),even though they knew that actual disbursementstended to be significantly lower. The Ministry of Fi-nance was invariably left picking up the tab.

At this point Permanent Secretary McNamaranoted that it was difficult for authorizing officersto monitor spending effectively given the poor sta-tus of Kariba’s accounting and budget controls. Fi-nancial information on expenditures was typicallyseveral months late, seldom comprehensive, andoften inaccurate.

Gerrard conceded that the government’s accountingsystems needed to be improved. However, in his view,the problem was not just one of receiving accurate andtimely information. Ministries, agencies, and depart-ments occasionally spent funds on initiatives sanctionedneither by Cabinet nor Parliament. Gerrard noted thatfunding for general administrative expenses, which in-cluded numerous trips abroad for permanent secretar-ies and deputy permanent secretaries, was invariably wellover budget in many of the ministries. However, minis-tries often spent only a small fraction of their allottedbudget on recurrent departmental charges, such as sup-plies. As a result, government employees often lacked sta-tionery and equipment, infirmaries were frequently with-out medications, and children had to share or do with-out textbooks.

Gerrard concluded that, in his view, the fundamen-tal problem was that most ministries viewed them-selves as spending agencies and had not accepted re-sponsibility for managing their budget. Expenditure

controls were not practiced rigorously. Commitmentswere not well documented, nor were they closely re-lated to the budget. Accounting control was viewed asthe exclusive responsibility of his office, and many inthe Civil Service did not consider budget managementto be a part of their work program.

Why should they, Barre shot back, when all of theiraccounting personnel were seconded from theAuditor’s Office? These were the officers who con-trolled the checkbook.

At this point, Edward Phillips summarized their con-versation so far, taking care to emphasize the points ofcommonality. He noted that there appeared to be con-sensus among the members of the task force that, as aresult of a move to a cash budget, a wide gap had devel-oped between the Ministry of Finance and other centralagencies and the operating ministries. The Ministry ofFinance was inclined to pursue a control-and-manageapproach toward ministries. For their part, ministriestended to supply the Ministry of Finance with informa-tion that would ensure that they got as much of the avail-able resources as possible regardless of actual costs orneeds. This situation did not augur well for any sort ofcorporate approach to management, strategic planning,or good cost estimation.

Just before assuming the post of budget director,Phillips received a detailed report on Kariba’s currentsystem of expenditure management, drafted byMalcolm “Chips” O’Toole and Steve Michaels, twonotable international consultants who played a majorrole in implementing innovative budget reforms in anobscure British dependency in the South Pacific. Thereport was highly critical of a number of aspects ofKariba’s current system of expenditure management.Phillips took this opportunity to share its contents withmember of the task force.

The report contended that expenditure planninginvolves several interconnected components. These

The O’Toole–Michaels

Report

components included the players—policymakers,planners, and operators—and the programs requiredto deliver the government’s agenda, the proper cost-ing of these programs, and even decisions about themost appropriate planning period. Each componentplayed an integral role in the delivery of thegovernment’s agenda. Moreover, the expenditure planwas only as good as its weakest link. O’Toole andMichaels examined each element of Kariba’s currentsystem and benchmarked it against best practice.

As the government’s key policymakers, Cabinet min-isters play a pivotal role in the expenditure process.One of the most important functions of the Cabinetis to establish priorities for the government. This oc-curs first at the macro level, in the form of overall fis-cal targets, and second at the micro level, in the allo-cation of government resources among individual sec-tors. These priorities form a blueprint for the otherplayers to follow. Therefore, the priorities must beclearly established, be transparent, and have the com-mitment of the whole Cabinet. It is critical that, to thegreatest extent possible, policymakers avoid givingmixed signals or frequently amending priorities.

In the case of Kariba, O’Toole and Michaels ob-served that the current situation was far from ideal.While the Cabinet set out its priorities, those priori-ties were often poorly recorded and disseminated, sothat even permanent secretaries were frequently un-aware of them. New developments often forced re-peated adjustments. The forecast of revenues had been,on balance, overly optimistic, and expenditure fore-casts had been routinely understated. The resultingfiscal forecasts lacked credibility and required continu-ous amendments in the expenditure framework.

Furthermore, individual Cabinet ministerstended to pursue their own agenda once they leftthe Cabinet room. In one celebrated incident, theMinister of Defense—one of the most powerful and

The Policymakers

ruthless men in the country—called the previousdirector of budget three months into the fiscal yearand demanded that his entire annual budget be re-leased immediately. The director complied. Thisresulted in a severe reduction in the funds availablefor other ministries and precipitated a minor crisisin the Cabinet that eventually contributed to thebudget director’s abrupt departure.

O’Toole and Michaels went on to note that, once thegovernment’s priorities had been established, the nextstep was to develop an action plan to ensure that thegovernment’s objectives could be realized. Normally, therole of planning was delegated to the central agencies—usually the Cabinet Office and Ministry of Finance. Lineministries could also play an important role in the plan-ning process. Their involvement at this stage helped tofoster corporate vision, which reduced the risk of the planbeing side-tracked by other agendas. Line ministries werealso often best placed to recommend the most efficientand effective programs for delivering the government’sagenda.

Here again, the situation in Kariba was far fromideal. While the central agencies had made someattempts to develop an action plan, the exercise hadbeen found lacking on a number of fronts. As pre-viously noted, the government’s agenda was predi-cated on a set of economic and fiscal assumptionsthat were frequently suspect, which made it diffi-cult for the central agencies to have faith in any plansbased on them.

The involvement of operating ministries in theplanning process to date has been virtually nonexist-ent. In instances where operating ministries providedcost estimates, the central agencies either heavily dis-counted or ignored the estimates. Thus, the ministrieshave few incentives to produce good cost estimates orto suggest cost-containing alternatives to deliver thegovernment’s agenda.

The Planners

The permanent secretaries of the various ministriesshould play a pivotal role in implementing thegovernment’s agenda. Their role is to translate intothe actual programs the priorities established by theCabinet and laid out in the action plan developed bythe Cabinet Office and the Ministry of Finance. How-ever, in Kariba, ministries tend to follow their ownagenda. Since permanent secretaries report to indi-vidual ministers, their first allegiance is to their im-mediate superior, and the corporate priorities of theCabinet are relegated to lower status. The CabinetOffice neither reviews the permanent secretaries norsanctions failure to pursue the government’s agenda.

Within the ministries, few permanent secretariesspent much time addressing budgetary issues. Theyperceived this task to be tedious and somewhat artifi-cial and delegated it to junior clerks within their bud-get offices.

Permanent secretaries were also fairly disengagedfrom the whole process of program monitoring andevaluation. O’Toole and Michaels noted that the nor-mal review of programs by operating ministries hadbeen the exception rather than the rule, and programswere typically reviewed only by the central agencies.Any change to existing programs was almost alwaysdriven by the central agencies, which—in light of theirgeneral unfamiliarity with the issues at stake—occa-sionally made costly mistakes. This approach alsomade it difficult to identify the responsibility centerfor a program that was not generating the intendedresults. The central agencies alleged that operatingministries would ascribe any shortcomings to prob-lems in the design of the program by central agencies.

O’Toole and Michaels noted that Kariba had experi-enced a long and painful series of unsuccessful effortsat budgetary reform. On the recommendation of sev-eral donor agencies in the late 1960s, the Ministry ofFinance attempted to implement a program planning

The Operators

Budget Systemsand Time Lines

and budgeting system. The objective of this effort wasto move away from the incrementalism and heavy biastoward historic expenditure patterns implicit withina traditional line-item budget and to use techniquessuch as cost-benefit analysis and cost effectiveness toput Kariba’s expenditures on a more rational and ob-jective basis. This attempt failed for a number of rea-sons, including the difficulty in defining and accuratelycosting government programs, the rigidities involvedin changing or adjusting programs once established,and weak analytic capacity in the line ministries andprovinces. In the 1970s, the country briefly flirted witha zero-based budgeting system, but it proved too cum-bersome to implement and failed to correct the strongbias toward historical expenditure norms.

Kariba currently uses a traditional, incremental line-item budgeting system for recurrent expenditures and adevelopment budget and public investment program forcapital expenditures. The former is managed by the Bud-get Office in the Ministry of Finance, and the latter bythe Ministry for National Economic Recovery and De-velopment. Under the current budget process, the Min-istry of Finance receives macroeconomic forecasts for thenext fiscal year (which is identical to the calendar year)from the National Economic Planning Board in March.It then issues a call circular to the ministries outliningthe general revenue and expenditure parameters underwhich they are to prepare their budgets in April. Minis-tries prepare their budgets during July and August, typi-cally by multiplying their historic expenditure levels bythe figures they receive in the call circular. They meetwith the Ministry of Finance in September and Octoberto discuss and negotiate these sums. The budget is final-ized in November and debated in Cabinet in December.The president then presents it to Parliament for approvalin mid-January, which typically takes from four to sixweeks. Often the budget for the existing fiscal year is notoften approved until late February or early March.

In theory, the capital budget is derived from thefirst year of the public investment program, which is

rolled over annually. However, O’Toole and Michaelsnoted that the links between the public investmentprogram and the annual recurrent budget were poorlydeveloped. It was not uncommon for projects to beapproved with significant downstream expenditureimplications without the knowledge or approval of theBudget Office. The public investment program cur-rently contains more than 150 projects, several ofwhich have been on the list for over a decade. The to-tal cost of these projects amounts to 8 percent ofKariba’s current gross domestic product (GDP),whereas annual funding for capital expenditures hasbeen running at around 2 percent of GDP. In light ofthe large gap between the demand for capital invest-ment and the actual funds available for it, the govern-ment has been forced to spread its limited resourcesthinly over a large number of projects, delaying theestimated completion dates for many of them butavoiding the political difficulties involved in cancel-ing sensitive projects.

The government has marketed the public invest-ment program aggressively to bilateral and multilat-eral donors. It has achieved considerable success ingaining funds for historic preservation, cultural en-richment, and the environment, with the promise thatit will put up significant matching funds in return.

The report went on to note that choosing the appro-priate period for setting the expenditure framework isvery important. On the one hand, too short a planningtime may not allow delivery on the government’s agenda.On the other hand, too long a time may prove unrealis-tic given the number of factors that could give rise to theneed for change. The challenge, O’Toole and Michaelsargued, is to choose the time horizon that avoids the pit-falls of either extreme.

In the case of Kariba the government has tended toconcentrate on a very short period—usually the cur-rent fiscal year. One can appreciate the motivation forchoosing such a period in light of the government’sprecarious fiscal position and its need to carefully

monitor the cash budget. However, concentration onthe short term has effectively precluded any strategicmedium-term approach to planning. To adopt a morestrategic approach to investment programming andto ensure that the downstream implications for cur-rent spending decisions are fully captured in the re-current budget, O’Toole and Michaels strongly rec-ommended that Kariba adopt a medium-term expen-diture framework.

Finally, O’Toole and Michaels observed that Kariba’scurrent budget does not accurately reflect the cost ofproviding public goods and services. They recom-mended that the government give due considerationto both input factors—wages, supplies, capital, etc.—and output factors—the quality and quantity of thegoods or services to be delivered. On the input side,the costing should be driven by current costs as op-posed to historical costs and it should be based on truecosting and not simply “incrementalism.” The inputfactors should be determined not on the basis of thestatus quo but on the most efficient way of deliveringthe programs in the future.

Phillips concluded his summary of the O’Toole–Michaels report. From the comments around the table,he noted that the participants generally agree with itsdiagnosis, but less so with its prescriptions.

Patricia Barre and Colin McNamara, the perma-nent secretaries for health and education, again tookthe lead. They argued that they in fact have much lesscontrol over their budgets than the report implies. Thisis particularly true in the area of personnel emolu-ments, where the central agencies exercise great influ-ence over the bulk of the line ministries’ recurrentbudget. They also contended that the general hiringfreeze imposed by the center is making it very diffi-cult for ministries to deliver on their mandate.

The Costingof Programs

Coping withthe Wage Bill

José Eduardo, the elderly but influential permanentsecretary from the Cabinet Office responsible for pub-lic service management, intervened in the discussion.He noted that, in the case of Kariba, personnel emolu-ments were both the largest single component of re-current expenditures and the fastest growing. Over the15-year period from 1975 to 1990, public sector em-ployment increased from 20 percent of total employ-ment to over 40 percent. The public sector had nowbecome the main employer in Kariba, and both thecentral and local government currently accounted forroughly 150,000 employees.

The wage bill accounted for about 4 percent of theGDP and consumed an ever-growing share of recur-rent expenditures. Corrective action was, Eduardocontended, required to address the situation so as tofree up the funds necessary to finance the supplies andcapital expenditures required to properly deliver gov-ernment programs. He observed that in 1990 the fundsavailable for basic supplies were roughly equal to thewage bill; however, by 1995, this ratio had declined byroughly 50 percent.

Eduardo acknowledged that central agencies hadexercised a relatively high level of control over wagesand salaries, but he argued that this was true in manycountries. In Kariba, the need for strict central con-

noitacudE 009,46

htlaeH 008,81

redrodnawaL 007,81

lareneG

srotartsinimdA 009,53

snaicinhceT 000,9

sreganaM 001,2

latoT 004,941

Estimated Employment in Local and Central Government

trol over both the size of the wage bill and the numberof people permanently employed stemmed from anumber of factors, including the past performance ofthe ministries, the tight fiscal position of the govern-ment, and the fact that the center would probably haveto bear any additional costs in reducing excess labor.

Eduardo noted that wages were among the mostrigid and inflexible public sector expenditures, sincereducing the size or mix of permanent employees of-ten gives rise to expensive incentive packages to in-duce them to leave. Kariba could achieve reductionsin the size of the public sector labor in several ways.The most cost effective form of rationalization wouldbe through normal attrition, often accompanied by ahiring freeze. (In some instances, there may not be amatch between those who remain in the public ser-vice and the skills required to perform particular tasks,so some selective hiring may be in order.) This, henoted, is normally carried out on a case-by-case basiswith the concurrence of a central agency.

To the extent that normal attrition is insufficientto accommodate the planned reduction in the laborforce, other options were often required. One approachwas to reduce the work week, with a correspondingreduction in the wages and salaries paid to each per-son. This approach would not increase the unemploy-ment level—a major concern in the case of Kariba.

Other options tended to be more costly but are other-wise effective. Departure incentives could be offered toredundant employees, which could take the form of earlyretirement for employees with a relatively long record ofemployment in the public sector and who meet certainage criteria. For other employees, cash incentives may benecessary; these tended to be on the order of two weeks’pay per year of service. Another alternative would be tofreeze the salary of surplus employees or alternatively togradually reduce their current salary over a specifiednumber of years. This tended to provide a strong incen-tive for affected employees to seek alternative employ-ment; it was also preferable to freezing or awarding lower

increases to the whole of the public service, which couldhave an adverse effect on morale, result in an exodus ofthe brighter and more valuable employees, and make re-cruitment more difficult.

At this point, the auditor general observed that thepromotion of employees through the ranks appearedto have been a function of the number of years a per-son had been in the public service, rather than on theneed for the function or the person’s performance.There was no evidence, to his knowledge, of any in-ternal audit function on matters related to perfor-mance, classification, continuing need for a function,or the actual verification of a match between the indi-viduals on the payroll and the individuals in the laborforce. This lack of control, he suggested, resuld in per-sons being paid for jobs that they were not qualifiedto perform; persons being paid for jobs that are re-dundant; and incidences of bogus employees on thepayroll.

Gerrard observed that, to help control the increasein the wage bill, the government had often grantedacross-the-board wage increases at a level below thelevel of inflation. This resulted in a decline in real wagesper employee of some 50 percent over the past twodecades. The end result was that the average wage ofpersons on the payroll was, for the most part, sub-stantially below the poverty level, and many were en-gaged in part-time employment in the private sectorto supplement their income. Levels of absenteeismwere high, morale and productivity were low, and re-quests for “incentive payments” to speed up the deliv-ery of government goods and services were common.

Many participants noted that attempts to under-take any meaningful downsizing have been, for themost part, unsuccessful. Under current contractualarrangements, redundancy measures are extremelycostly and may total as much as four or five times anemployee’s annual salary. These incentives must bepaid in cash, which is particularly problematic for agovernment operating on a cash budget regime.

The permanent secretaries from the operating min-istries blamed the central agencies for the current situ-ation, pointing to the control they command on staff-ing, wage settlements, and contract arrangements withemployees’ unions. The central agencies blamed theoperating ministries and cited incidents such as thelarge number of redundant positions, improper orinflated position classifications, disregard for numer-ous staffing freezes, bogus employees on the payroll,and the high incidence of temporary employees onministry staffs. The director of the Budget Office notedthat the lack of clear delineation of the responsibili-ties related to the management of the public service isa matter of urgent concern.

Edward Phillips thanked the members of the task forcefor what had in his opinion been a lengthy but produc-tive meeting. He observed that there was general con-sensus regarding the critical need for a well-articulatedplan to effect changes in both the way in which the ma-chinery of government was functioning and the mannerin which scarce resources were being allocated. He askedthe participants to consider before their next meetingwhat should be the fundamental elements of an actionplan for rectifying the problems they identified.

As the task force members departed, Phillips pon-dered the appropriate course of action. Clearly, thecurrent system lacked credibility and was not support-ing the efficient and effective allocation of resourcesin accordance with government priorities. Equallyclearly, fundamental reform was necessary. But beyondthese basic points, opinions differed widely andsharply. What were the most pressing issues to be ad-dressed? How could the task force develop a viableaction plan for implementing them? How could theyconvince the participants in the budget process to goalong with their recommendations?

Conclusion

Questions 1. How credible is the current budget system? Whatare its principal weaknesses?

2. What are the principal issues that should guideany reform of Kariba’s budget system?

3. Which Ministries will support efforts to move toa more transparent and predictable budget system?Which Ministries will oppose such efforts? Why?

4. What should Phillips do? What should his nextsteps be?