africa technical department series adjusting educational...
TRANSCRIPT
I, btO0 52
World Bank Discussion PapersAfrica Technical Department Series
Adjusting EducationalPollc'ies
Conserving Resourceswhile Raising SchoolQuality
Bruce Fuller* ,4
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Recent World Bank Discussion Papers
No. 74 Costs and Beniefits of Rent Control in Kumasi, Ghana. Stephen Malpezzi, A. Graham Tipple, and Kenneth G.Willis
No. 75 Ecuador's Amazon Region: Development Issues and Options. James F. Hicks, Herman E. Daly, Shelton H. Davis,and Maria de Lourdes de Freitas [Also available in Spanish (75S)l
No. 76 Debt Equity Conversion Analysis: A Case Study of the Philippine Program. John D. Shilling, Anthony Toft, andWoonki Sung
No. 77 Higher Education in Latin America: Issues of Efficiency and Equity. Donald R. Winkler
No. 78 The Greenhouse Effec: Implicationsfor Economic Development. Erik Arrhenius and Thomas W. Waltz
No. 79 Analyzing Taxes on Business Income with the Marginal Effective Tax Rate Model. David Dunn and AnthonyPellechio
No. 80 Environmental Management in Development: The Evolution of Paradigms. Michael E. Colby
No. 81 Latin America's Banking Systems in the 1980s: A Cross Country Comparison. Felipe Morris, Mark Dorfinan,Jose Pedro Ortiz, and others.
No. 82 Why Educational Policies Can Fail: An Overview of Selected Afiican Experiences. George Psacharopoulos
No. 83 Comparative Afitcan Experiences in Implementing Educational Policies. John Craig
No. 84 Implementing Educational Policies in Ethiopia. Fassil R. Kiros
No. 85 Implementing Educational Policies in Kenya. G. S. Eshiwani
No. 86 Implementing Educational Policies in Tanzania. C. J. Galabawa
No. 87 Implementing Educational Policies in Lesotho. T. SohI Thelejani
No. 88 Implementing Educational Policies in Swaziland. Cisco Magalula
No. 89 Implementing Educational Policies in Uganda. Cooper F. Odaet
No. 90 Implementing Educational Policies in Zambia. Paul P. W. Achola
No. 91 Implementing Educational Policies in Zimbabwe. 0. E. Maravanyika
No. 92 Institutional Reforms in Sector Adjustment Operations: The World Bank's Experience. Samuel Paul
No. 93 Assessment of the Private Sector: A Case Study and Its Methodological Implications. Samuel Paul
No. 94 Reaching the Poor through Rural Public Employment: A Survey of Theory and Evidence. Martin Ravallion
No. 95 Education and Development: Evidencefor New Priorities. Wadi D. Haddad and others
No. 96 Household Food Security and the Role of Women. J. Price Gittinger and others
No. 97 Problems of Developing Countries in the 1990s. Volume I: General Topics. F. Desmond McCarthy, editor
No. 98 Problems of Developing Countries in the 1990s. Volume 11: Country Studies. F. Desmond McCarthy, editor
No. 99 Public Sector Management Issues in Structural Adjustment Lending. Barbara Nunberg
No. 100 The European Communities' Single Market: The Challenge of 1992for Sub-Saharan Africa. Alfred Tovias
No. 101 International Migration and Development in Sub-Saharan Africa. Volume 1: Overview. Sharon Stanton Russell,Karen Jacobsen, and William Deane Stanley
No. 102 International Migration and Development in Sub-Saharan Africa. Volume II: Country Analyses. Sharon Stanton Russell,Karen Jacobsen, and William Deane Stanley
(Continued on the inside back cover.)
AFRICA TECHNICAL DEPARTMENT SERIES
Technical Paper Series
No. 122 Dessing, Support for Microenterprises: Lessons for Sub-Saharan Africa
No. 130 Kiss, editor, Living with Wildlife: Wildlife Resource Management with Local Participation in Africa
No. 132 Murphy, Casley, and Curry, Farmers' Estimations as a Source of Production Data: MethodologicalGuidelines for Cereals in Africa
No. 135 Walshe, Grindle, Nell, and Bachmann, Dairy Development in Sub-Saharan Africa: A Study of Issuesand Options
No. 141 Riverson, Gaviria, and Thriscutt, Rural Roads in Sub-Saharan Africa: Lessons from World Bank Experience
No. 142 Kiss and Meerman, Integrated Pest Management and African Agriculture
No. 143 Grut, Gray, and Egli, Forest Pricing and Concession Policies: Managing the High Forests of Westand Central Africa
No. 157 Critchley, Reij, and Seznec, Water Harvestingfor Plant Production, vol. U: Case Studiesand Conclusionsfor Sub-Saharan Africa
No. 161 Riverson and Carapetis, Intermediate Means of Transport in Sub-Saharan Africa: Its Potentialfor Improving Rural Travel and Transport
N o. 165 Kellaghan and Greaney, Using Examinations to Improve Education: A Study in Fourteen AfricanCountries
Discussion Paper Series
No. 82 Psacharopoulos, Why Educational Policies Can Fail: An Overview of Selected African Experiences
No. 83 Craig, Comparative African Experiences in Implementing Educational Policies
No. 84 Kiros, Implementing Educational Policies in Ethiopia
No. 85 Eshiwani, Implementing Educational Policies in Kenya
No. 86 Galabawa, Implementing Educational Policies in Tanzania
No. 87 Thelejani, Implementing Educational Policies in Lesotho
No. 88 Magalula, Implementing Educational Policies in Swaziland
No. 89 Odaet, Implementing Educational Policies in Uganda
No. 90 Achola, Implementing Educational Policies in Zambia
No. 91 Maravanyika, Implementing Educational Policies in Zimbabwe
No. 101 Russell, Jacobsen, and Stanley, International Migration and Development in Sub-Saharan Africa,vol. I: Overview
No. 102 Russell, Jacobsen, and Stanley, International Migration and Development in Sub-Saharan Africa,vol. 1: Country Analyses
132 w 1a3 World Bank Discussion PapersAfrica Technical Department Series
Adjusting EducationalPolicies
Conserving Resourceswh'ile Raising SchoolQuality
Bruce FullerAklilu Habte, editors
The World BankWashington, D.C.
Copyright C 1992The International Bani: for Reconstructionand Development/THE WORLD BANK
1818 H Street, N.W.Washington, D.C. 20433, U.S.A.
All rights reservedManufactured in the United States of AmericaFirst printingJanuary 1992
Discussion Papers present results of country analysis or research that is circulated to encourage discussionand comment within the development community. To present these results with the least possible delay, thetypescript of this paper has not been prepared in accordance with the procedures appropriate to formalprinted texts, and the World Bank accepts no responsibility for errors.
The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) andshould not be attributed in any manner to the World Bank, to its affiliated organizations, or to members ofits Board of Executive Directors or the countries they represent. The World Bank does not guarantee theaccuracy of the data included in this publication and accepts no responsibility whatsoever for anyconsequence of their use. Any maps that accompany the text have been prepared solely for the convenienceof readers; the designations and presentation of material in them do not imply the expression of any opinionwhatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning thelegal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitationof its boundaries or its national affiliation.
The material in this publication is copyrighted. Requests for permission to reproduce portions of it shouldbe sent to the Office of'the Publisher at the address shown in the copyright notice above. The World Bankencourages dissemination of its work and will normally give permission prompdy and, when thereproduction is for noncommercial purposes, without asking a fee. Permission to photocopy portions forclassroom use is not required, though notification of such use having been made will be appreciated.
The complete backlist of publications from the World Bank is shown in the annual Index of Publications,which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, andcountries and regions. The latest edition is available free of charge from the Distribution Unit, Office of thePublisher, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., orfrom Publications, The World Bank, 66, avenue d'1ena, 75116 Paris, France.
ISSN: 0259-210X
Bruce Fuller is associate professor of education at Harvard University, and Aklilu Habte is educationadviser with the Education Office of UNICEF in New York.
Library of Congress Cataloging-in-Publication Data
Adjusting educational policies : conserving resources while raisingschool quality / Bruce Fuller, Aklilu Habte, editors.
p. cm.-(World Bank discussion papers, ISSN 0259-21OX;132. Africa Technical Department series)
Includes bibliographical references (p. ).
ISBN 0-8213-1932-91. Education and state-Africa, Sub-Saharan-Congresses.
2. Educational change-Africa, Sub-Saharan--Congresses.3. Education-Africa, Sub-Saharan-Finance-Congresses. I. Fuller,Bruce. II. Habte, Aklilu, 1929- . III. Series: World Bankdiscussion papers; 132. IV. Series: World Bank discussion papers.Africa Technical Department series.LC95.S726A34 1991379.67-dc2O 91-40976
Abstract
African governments and donor agencies have realized in recent years that development programscan. only succeed if built upon sound policy and institutional foundations. Since the mid-1980's, severaleducational policy adjustment programs have been initiated in Sub-Saharan Africa.
This volume explores these early policy efforts - drawing on reports from government leaders anddonor representatives in Ghana, Malawi, and Senegal. Three specific questions are addressed: Whattypes of policy and budget changes have been attempted? What lessons have been learned regardinglocal school and community effects, stemming from central policy adjustments? How can policyprograms better complement long term efforts to strengthen institutions?
Papers in this volume were presented at a conference attended by government officials and donoragencies, co-sponsored by the World Bank and USAID. Highlights of the conference debate also arereported.
v
The Editors and Contributors
Bruce Fuller is associate professor of education at Harvard University. He has worked on policyprograms and school-level research for the World Bank and USAID.
Aklilu Habte is formerly the senior human resources advisor to the Vice President, Africa RegionalOffice, the World Bank. Mr. Habte presently works with the new education office of UNICEF in NewYork.
Birger Fredriksen is principal education planner in the education and employment division of theWorld Bank. He was team leader on the Senegal policy program.
Frances Kemmerer is an associate professor of education, State University of New York, Albany.She worked in Africa under USAID's Improving the Efficiency of Educational Systems project.
Edward Ngaye is chief of the planning division, Malawi Ministry of Education and Culture. Headvises the principal secretary on implementation of the ministry's policy action program.
Vida Yeboah is depuly principal secretary of education, Government of Ghana. She is deeplyinvolved in crafting and implementing policy changes.
vi
Foreword
African governments - since independence - have struggled to mobilizesufficient public resources to broaden educational opportunities. In turn expan-sion of basic schooling, complementing efforts in other sectors, has spurredeconomic growth in many African societies.
The past decade, of course, has been a sobering time for many African states.Efforts to simultaneously expand schooling, maintain educational quality, andmore equitably distribute the benefits of basic education have strained thepolitical will and fiscal fragility of many governments. And the decade-longprocess of adjusting macroeconomic policies has further constrained the educa-tion sector's institutional capacities.
Since the mid-1980s, several African governments have intensified efforts toconserve their educational resources - through significant adjustments inpolicies and institutional structures. Determined conservation of resources throughspecific cost-saving measures often produces more efficient allocation of theseresources to strategies to consistently boost student achievement, equity, andsubsequent economic and social benefits.
This collection of papers reports on initial progress made by African govern-ments in improving their educational policies. The recent conference for whichthese papers were prepared -co-sponsored by the World Bank and USAID -convened government leaders and donors who have worked together in thisnovel area of institution building.
The president of the World Bank recently announced a substantial increase inlending to support policy change and investment in the education sector. It istimely that we reflect upon how policy adjustment strategies in Africa have beenimplemented and how they have affected schools and children. What types ofpolicy adjustments are more likely to succeed? How can policy change strengtheneducational institutions over the long run? This book takes a modest step inanswering these questions.
Ismail SerageldinDirector
Technical DepartmentAfrica Region
Contents
Preface ............................................ xi
Chapter 1Education Policy Adjustment ..................................... I1Defining Central Strategy, Assessing Local Effects - by Bruce Fuller and Aklilu Habte
Chapter 2Policy Reform to Raise School Quality ................................... 14nie Case of Malawi - by Edward Ngaye
Chapter 3Ghana's Policy Adjustment Initiative ..................................... 18; 18Opportunity for Renewal - by Vida Yeboah
Chapter 4Social and Political Constraints on Education Reform .................................... 23The Case of Senegal - by Birger Fredriksen
Chapter 5Inducing and Monitoring Policy Change ............. ........................ 3S- by Frances Kemmerer
Chapter 6Lessons Leamed? ..................................... 43A Colorful Rainbow of Viewpoints - by Bruce Fuller
References ..................................... 48
ix
Preface
Developing countries - over the past five decades - have made breathtaking progress in rapidlyexpanding mass education. More recently, many governments have turned their attention towardimproving school quality, even rethinking the purpose and content of basic education within ruraland urban settings. And most political leaders realize that quick growth in schooling does not nec-essarily lessen inequality in terms of which families gain access to quality education.
In Africa, however, governments face severe constraints in their struggle to expand educationalaccess, raise school quality, and advance equity. Entering the 1990s, per capita income in many Africannations was at or below levels observed in 1960. Resources available for basic education fell in realiterms throughout the continent, due to external economic shocks, political strife, and competingdevelopment priorities.
Following on the heels of macro-economic adjustment programs - undertaken jointly by gov-ernments and development banks - a handful of education policy reform programs were beingdeveloped by the mid-1980s. This modest book reports on these earliest efforts, initiatives partiallyinanced by the World Bank or the U.S. Agency for International Development. The thinness of thisvolume reflects how little we know, at this early point, about the actual effects of policy adjustmentin the education sector.
This book does surround and describe the range of educational policies and budget items onwhich sector-adjustment efforts have focused. We concentrate on experience within three specificcountries: Ghana, Malawi, and Senegal. Our contributors speak to three fundamental issues:
- What types ofpolicy and budget changes have been attempted, and which are morelikely to be successfully implemented? What kinds of policy reforms face stifferconstraints in terms of technical complexity or political opposition?
* What lessons have been learned about the local effects of centralpolicy change? Policyadjustments usually entail conserving resources that are being used inefficient-ly (or inequitably), then reallocating these funds to efforts that help raise theeffectiveness of local schools. But what evidence do we have that more (techni-cally) rational utilization of resources can be attained, and that concrete local ef-fects can be observed?
* How can policy adjustment programs better complement long-term efforts to strengtheninstitutions? Most policy programs are blended with more traditional project as-sistance, technical support, purchase of school inputs, and capacity-building. Akregovernments becoming more skilled in generating policy alternatives and assess-ing implementation constraints? And are donors learning something about howto mix policy and project forms of assistance?
You will quickly realize that we still have a long empirical road to travel before eviidence issufficient to answer these questions.
xi
Chapters in this book were originally presented at a recent conference held in Washington D.C.,aimed at sharing early experience and lessons-learned. The seminar involved African governmentofficials and donor agencies, and was sponsored by the World Bank and USAID. The meeting wasco-chaired by Aklilu flabte and Gary Theisen. The International Science and Technology Institutehelped organize the meeting. Special thanks also are due to Joan Claffey, Martha Engel, Rick Hunt-ington, Irene Landwehr, and Jim Socknat. The papers benefitted from review by Cameron Bonnerand Emil Baran, under the guidance of Peter Moock and Mr. Socknat. Steve Slanerassisted with the editing of chapters, and Arc & Line Design produced the final volume.
Bruce FullerHarvard University
Aklilu HabteUNICEF and the World Bank
xii
Chapter 1
EDUCATION POLICY ADJUSTMENTDefining Central Strategy, Assessing Local Effects
Bruce Fuller and Aklilu Habte
decade ago the World Bank, with its building that is crumbling.A_l client governments, embarked on ini- The chapters which follow define and illus-tiatives that aimed to modify the rules trate the burgeoning field of education policy
and institutions which organize national econ- adjustment. We review variation in specificomies. More recently, donors working in the ed- country strategies. Our contributors assess com-ucation sector have been attracted to the mon policy and budget reforns aimed atpromise of making policy adjustments in boosting the efficiency of different national ed-school institutions. Strategies mounted in par- ucational systems. We report on lessonsticular countries vary significantly. Yet the learned from these initial attempts at sector ad-overall shift is remarkable - moving away justment. Do policy programs yield intendedfrom conventional projects and toward policy policy and budgetary effects within central gov-and budget reform. ernments? Do adjustment initiatives contribute
This sharp move toward education sector ad- to the long-term strengthening of central insti-juslment is motivated by several factors: Gov- tutions and enhance the quality of localernments and donors recognize that explosive schools? These are the basic questions asked bypost-war expansion of mass schooling, while our contributors.yielding benefits, has made educational institu- A debate is unfolding over what types oftioIns more brittle - unable to sustain this pace macroeconomic adjustment strategies are mostof qrowth, unable to backstop eroding levels of effective. The very definition of "effective" iseducational quality, and unable to use scarce evolving, now taking into account effects feltresources more effectively. Continuing conven- by the poor and by fragile governments. Intional project assistance, while necessary, is not contrast, little debate is occurring on altema-sufficient for long-term strengthening of central tive forms of education sector adjustment. Weeducation ministries and local schools. In seem to be drawn to this novel medicine with-many cases, traditional "stove-pipe projects" out critically assessing its multifaceted effects.simrply add categorical programs that may in- More determined efforts are required to un-trigue Western sponsors but which are difficult derstand the effects of adjustment strategiesto sustain over time. Unless a nation's structure within the education sector, then to reflect onof school finance and management provides how policy interventions can be better crafted.inputs and human resources that effectively Careful thinking is needed at the outset to spec-boost literacy, an array of narrow projects can ify what results - within central governmentsonly attack symptoms, rather than strengthen- and within schools - are intended. Short-terming underlying institutional capacities over impacts on government spending are impor-time. It is like putting a fresh coat of paint on a tant, through the containment of inefficiencies
1
or encouragement of greater cost-recovery. Yet Some project officers within donor agenciespolicy change also intends to touch teachers argue that each adjustment package must beand pupils - encouraging central allocations crafted to fit country-specific conditions. Weand school-level practices that more effectively agree, in part. However, given the extent (andboost children's basic literacy and achieve- cost) of sector adjustment operations, wement. We are just beginning to understand the should push to extract generalizable lessonsconditions under which central policy improv- learned and common mistakes made. Onlyments influence local action. then can we define the types of policy initia-
tives and country conditions under which ex-Surrounding and Defining the Field pected (short- and long-term) outcomes will
likely occur.Children throughout southern Africa survive Growth in sector adjustment programs. One
on eating nsima. Like other forms of corn meal, fact is unambiguous: the share of multilateralnsima's nutritional value is sufficient, not im- assistance now linked to policy adjustment haspressive, with long-term effects on growth that risen dramatically over the past decade. One-are significant, not optimal. Is this an apt met- fourth of total World Bank lending now goesaphor for sector adjustment programs? Policy for macroeconomic or sector adjustment activi-reform initiatives are vcaguely tasty and certain- ties. Conditionalities linked to balance-of-ly filling. With little variation in the recipe, and payments support or exchange rate reforms of-buttressed by faith in its effects (as opposed to ten influence the education sector. Ceilings onhard evidence), donors continue to move ag- civil service posts, for instance, limit govern-gressively into the policy adjustment area. ments' capacity to hire new teachers. Since the
The field itself is still in the process of defini- education sector represents 15% to 35% of alltion. Government officials and donors cite the government spending in most African coun-same types of policy changes as exemplars of tries, sector reform not surprisingly becomes anwhat sector adjustment is all about. Others important ingredient of macroeconomic adjust-argue that "adjustment" is simply a euphe- ment strategies (World Bank 1988b, 1989).mism for cutting government spending. While Half of all World Bank loans in educationmacro-economic adjustment does have this fo- now contain policy measures, and this propor-cus, the earliest education adjustment pro- tion is increasing. The most common form ofgrams (described in this volume) focus on education loan is now a hybrid combination ofshifting education resources to more effective policy adjustment and conventional project as-inputs or using these resources to achieve more sistance. With some reservation we introduce aequitable pattems of pupil subsidy. Programs new acronym - SAI - to characterize "sectorin Malawi and Mali actually push govern- adjustment and investment" operations. Mikements to protect or boost expenditures for edu- Wilson's (1988) review of Bank experience re-cation, particularly spending for primary veals a variety of instruments now used ineducation. working with governments, including conven-
Below, we sort the growing list of possible pol- tional projects, sector investments providingicy changes into clearer categories- illustrat- ministries discretion over allocations, and ad-ing the variety of objectives which can be justment loans that tightly link policy and bud-served by sector programns (also see World Bank get reforms to disbursement of loan tranches1988a). The general field, however, remains (see also Chapter 4 by Birger Fredriksen).underdeveloped in terms of linking various pol- Similarly, basic education initiatives under-icy actions to expected outcomes. Our capacity taken by the U.S. Agency for International De-to monitor the effects o:f policy change even velopment (USAID) increasingly combinewithin central governments, let alone local program and project assistance in a variety ofschools, remains quite crude. Nor is our knowl- ways, the former encompassing policy dia-edge systematic as to what combinations of logue, information systems, and research - allpolicy intervention are most feasible or effec- aimed at provoking policy and fiscal reform.tive. Much remains to be learned regarding With limited resources, USAID usually provideswhat implementation steps are most effica- in-country technical assistance to the educa-cious. tion sector and avoids large capital invest-
2
men,ts. Work with African governments in- sector adjustment efforts tend to be drivencreasingly involves policy dialogue and reform more by the proximity of policy and budgetaimed at improving long-term management levers than by any clear understanding ofcapacity, boosting efficiency, and addressing long-term effects.ineqtuities in terms of who benefits from formalschooling. Policy Intervention within
One IJSAID project recently begun in Mali Uncertain Contextsprovides technical assistance to help the gov-ernrnerLt implement policy reforms developed Policy change. Sector-specific policy adjust-in collaboration with the World Bank. These re- ments enter a political and economic environ-forms provide for the shifting of excess resourc- ment that determines whether the specifices from secondary schools (with a low initiative will take hold in the short run. Thispupil:teacher ratio of 14:1) to primary educa- environment has three key features, illustratedtion. USAID assistance will help the govern- in Figure 1.1 (see page 4). The "indigenous" po-merit in developing a modest information litical-economy is constantly driving policy de-system necessary for tracking the shift of teach- bate and occasional change. For instance, aers and expenditures. USAID also is financing government may be pushing to eliminate allan inservice teacher training program to help pupil fees within primary schools at the sameupgrade primary school quality, another policy time that a donor is advocating stronger cost-objective worked out with the Bank. USAID recovery measures. Or the secular governmentand the French aid agency are contributing to may be faced with opposition from religiousa $12 million quick disbursing fund linked to groups (many of which run schools) at the pre-successful implementation of policy and bud- cise time that a donor is arguing to deregulateget adjulstments. private schools.
Education sector adjustment: a slippery con- In addition, external economic shocks fre-cepit. Let us try out a specific definition of the quently constrain African governments' capac-process: Governments and donors negotiate ity to finance sector reforms or to undertakeand implement discrete policy or budgetary any controversial "adjustment" requiring con-changes that attempt to contain government siderable political capital. Finally, policy re-spendiing, allocate scarce resources more effi- forms specific to the education sector mayciently, and/or strengthen incentives for im- conflict with broader economic adjustments be-provement in school-level quality and ing pushed by the same or a different donor. Ef-perforrnance. forts to increase recurrent spending on
WVithin our definition, "efficiency" can relate textbooks, for instance, are stymied at times byto the school's internal impact on children's lit- overall spending ceilings negotiated with IMFeracy and achievement. For instance, govern- or World Bank economists. Given the fluidity ofment and private providers of education the environment, the seemingly tight link be-ideally support those inputs and pedagogical tween policy change and efficiency-effects canpractices that most effectively boost pupil loosen considerably over the long run.achievement. Efficiency also relates to out- Even the term "adjustment" distracts us fromcomnes external to the school. In many African effecting long term institutional change. A dec-settings, the school often imparts skills, knowl- ade ago, adjustment of exchange rates or pro-edge, cnd attitudes useful in urban areas but of ducer prices appeared to be a sufficient short-little relevance to conditions in rural communi- term antidote to economic decline. With great-ties. Thlis represents an external inefficiency. er sobriety, at least in Africa, donors are
Our definition has two distinct characteris- (again) realizing that more careful use of re-tics. First, the emphasis is on outcomes or func- sources requires long-term improvement in in-tiolns of policy adjustment as opposed to, say, stitutional capacity and managementpolicy levers that can be manipulated. Second, The long-term effectiveness of policy changethe definition assumes that these policy levers is constrained by three factors (seen on theare directly connected to the intended out- right-hand side of Figure 1.1). In designing pol-cornes - effects observed both within the cen- icy reform efforts, we must be clear about thetral government and at the school level. Initial nature and location of anticipated effects. Of-
3
Figure 1.1Forces Moderating Actual Effects of Policy Adjustment Strategies
Tirne 1 Time 2
SECTOR-ADJUSTMENT PROCESS ACTUAL FELT EFFECTS within centralBEGINS (within central government) government and local schools
Short run constraints: uLong rn intervening forces:In-country policy and budget debates Unclear links between policy change andExternal economic shocks intended effectsLimited political capital and stability Technical and political complexity of
change required (across organizationalConflicts between sector adjustment and levels)macroeconomic strategy Resistance from current beneficiaries or in-
terest groups, and reticence due to techni-cal uncertainties
4
ten the results of manipulating one policy lever more equitable distributional policies). Butmove across organizational levels in complex seemingly tidy policy initiatives often meet anways. Many SAI operations, for instance, now unpredictable, hostile environment whichrequire governments to shift sector resources erodes even short-run viability. And even poli-from secondary or tertiary institutions to pri- cy adjustments that stick in the short run canmary education, recognizing that gains in ba- be derailed by a hazy conception of where ef-sic literacy most directly benefit agricultural fects are to occur, related technical complexi-productivity and quality of life. While this shift ties, and relentless institutional resistance orcan be monitored within central ministries, inertia.real change requires moving teachers (sincesalaries comprise the bulk of recurrent spend- A Menu of Policy Choicesing) and increases in class sizes at post-primarylevels. Each of these shifts requires a great deal Governments and donors are eacger to act -of diplomatic and technical work - at different to expand schooling, to improve educationalorganizcational levels - before any impact can quality, to push for more efficient use of sectorbe observed. resources. The first generation of SAI programs
now being implemented reflect an earnest fo-Complexity of Institutions cus on the novel intervention of policy adjust-
ment. As mentioned above, we are moreSimilarly, policy initiatives often underesti- intrigued with mounting the strategy and ar-
mate the technical complexity of implement- ranging new sets of inputs than with under-ing mneaningful institutional change, standing the resulting effects. Not knowing justparticul(arly at the school level. Donors are how alternative adjustment strategies yield dif-pushing governments, for instance, to reduce fering outcomes, we continue to define the fieldthe rate at which pupils repeat a grade level or only by the menu of actions taken at the frontstandar(d. The objective is admirable: to con- end of the process. This is rather like definingcentrate scarce primary school resources on agricultural production only by examining iso-those children who progress at reasonable lated inputs, leaving out human and materialrates. But the range of local actions required is facets of the technical process and ignoring thevery complex: assessing pupil achievement, varied quality of outputs. Notwithstanding thiskeeping adequate records (when some urban definitional weakness, a brief review of sectorprimaries have over 3,000 students), and work- policy interventions is a useful exercise.ing with parents to explain the policy change. The chapters which follow illustrate how dif-
Finally, actual effects stemming from policy ferent elements from this menu have come to-change are blocked or watered down by institu- gether within country-specific adjustmenttional resistance - inertia or strong political programs. Figure 1.2 arranges possible policyopposition that can derail the best-designed re- and budget reformns into five categories. First,formn. Policies that bring students into the civil macro adjustment programs may pull the edu-service when they enter a teacher training col- cation sector into compliance with broader polit-lege, for example, are very costly. But they em- ical-economic objectives. Efforts to containploy large numbers of well-educated youth, government spending, for instance, may affectmany of whom come from politically vocal resources available to the education sector. Onfamilies. Efforts to reduce these types of subsi- the other hand, where education receives andies run into institutional traditions (often be- insufficient share of GDP or government spend-gun under colonial admninistrations) and ing, donors have advocated higher levels ofconcrete political resistance. At times, persua- spending. Economic or social policies aimed atsive leaders can push beyond such resistance, changing the structure of opportunities - e.g.,as detailed in Vida Yeboah's discussion of Gha- providing women with access to agriculturalna's reforn experience (Chapter 3). credit or other inputs - may place new exter-
In surn, rational planners of adjustment of- nal demands on the education sector. Turningten focuLs on a discrete policy lever which, when to the education sector itself, adjustment strate-pushed and pulled, should yield a concrete out- gies usually attempt to contain costs faced bycome (lower subsidies, budgetary shifts, or central government. These measures include
5
reducing subsidies, grealer utilization of school penalties but offer little assistance in actuallyfacilities, concentrating resources on children improving pedagogy or school management.who are achieving, or ernploying new technol- In addition, headmasters and teachers are ac-ogies to lower productions costs (associated countable to central government regulationswith producing instructional materials and and dictates, but incentives are decouplednew teachers). from actual pupil performance or to demands
Policy levers also have been identified which expressed by the local community. School fi-aim at more efficient allocation of sector resourc- nance schemes and salary incentives could ad-es. This includes shifting additional resources just incentives to encourage better teachingtoward primary schools, when gains are most and community-level accountability.urgently required in basi.c literacy. Similarly, The columns in Figure 1.2 provide three di-allocative efficiency is low when most resourc- mensions for contrasting the different policyes are going to secondary and university levels, initiatives. The location of impact for eachparticularly when wage employment is limited item is fairly obvious. Yet policies aimed at lo-or declining. Resources within subsectors also cal levels, in contrast to central governmentcan be used more cost-elfectively. For instance, adjustments, call for very different implemen-there is now clear evidence on the positive tation methods and bring forth different tech-achievement effects stemring from infusions nical and political complexities. Our judgmentof textbooks and basic instructional materials on comparative levels of institutional resis-(Fuller 1987). Yet throughout Africa very few tance and technical complexity are rough atresources remain for these essentials, after allo- best Precise levels of resistance and complexi-cating monies for teacher salaries and admin- ty, of course, vary across countries and poli-istration. cies. Our judgments are based on written
Inventive policy reforrns have been suggest- reports and our own experience in-country.ed to help diversify sources of school finance. Few A diversity of policy options. This particularAfrican governments can afford to keep pace framing of alternative policies - looking atwith skyrocketing social demand for schooling. them from three different angles - illustratesMany governments already encourage local the wide variation in the nature and locationcommunities to contribute cash, labor, and/or of intended effects, and emphasizes that imple-materials to help cover capital and recurrent mentation processes also differ depending oncosts. But additional encouragement of private the desired impact. In shaping sector adjust-schools will be needed, particularly if govern- ment strategies, SAI designers must spendments are to conserve resources for maintain- considerable time in negotiating viable agree-ing even minimal levels of educational ments with governments. Long-term technicalquality. and institutional constraints on actual imple-
Finally, donors and governments increasing- mentation receive much less attention. Onlyly are discussing how to create stronger incen- rarely do we think systematically about howtives for teachers and heacdmasters. Little these hindrances to effective sector adjustmentinformation is available on actual achieve- vary across different types of policies.ment levels of pupils. Progression through pri- Within particular countries, these alternativemary school often is used as a proxy for policy ingredients are mixed together in vari-"performance." But this tells us little about ous ways. Early sector adjustment operationschildren's achieved levels of literacy. National focused on containing government spending,examinations are used cLs selection devices; pinpointing where budget cuts could be ac-they are not seen as tools to assess attainment complished without narrowing educational ac-of minimal literacy levels. In much of Africa, cess or quality. In both Ghana and Senegal,headmasters and teachers are expected to me- for instance, subsidies for university studentchanically perform their jobs, often working in housing and auxiliary services were reduced.isolated settings. In the cabsence of incentives Other sector adjustment activities are nowfor professional developrnent, the motivation creatively balancing budgetary reductionsof headmasters and teachers often is quite low. with incentives for local improvement. TheBorrowing from colonial patterns, school in- World Bank program in Malawi, for example,spectors are seen as regulators who can exact inventively provides support to primary educa-
6
FIgure 1.2Educational Policy and Budget Items
Impact Institutional TechnicalLocation Resistance Complexity
1. Fitting the Sector to Economic andFolitical Proritles
Slow sector spending to free capital and contain Central * Uinflation; or raise sector spending to boostproductivity and social objectives
Limit growth in the teaching force to contain the Central U Vcivil service wage bill
Reduce cost of imported school inputs to Central * Uconserve foreign currency
Shift tax burden for vocational training to the Central and local * Uprivate sector
Implement more equitable pricing Central * Upolicies, lowering subsidies forhigher-income parents or for teachers(e.g. housing subsidies)
Open opportunities for females in Local * Uthe economy and education sector,to raise their productivity inagriculture and the informal sector
2. Containing Education Sector Costs
Fully utilize school facilities thru Central and local * Udouble-shifts, multigrade dasses,admission of day pupils (secondaryschools and universities)
Reduce boarding subsidies that benefit Central * Umore advantaged pupils, includingsecondary or university pupils whowill gain higher earnings
KEY: * = High * = Moderate or Low
Notes: 'Impact location' (column 1) refers to whether the policy adjustment intends to influence change within the central gov-ernment, within local schools, or within private firms. 'Institutional resistance' may occur within any of these three institutions.'Technical complexity' relates to implementing the policy change, independent of institutional or political points of resistance.Sources: Chapters by Ngaye and Yeboah (this volume), Mingat & Tan (1988), Wilson (1988), Windham (1988), Woild Bank(11 988).
7
Figure 1.2 (continued)Educational Policy and Budget Items
Impact Institutional TechnicalLocation Resistance Complexity
Limit the number of years of schooling Central and local * Udefined as basic education
Encourage early retirement: of teachers Central * Uand administrators in sub-sectors thatare over-staffed
Raise class size, espedally alt Central and local * Upost-primary levels
Bring the capital budget in line with Central * Usustainable levels of recurrent spending
Introduce new leaming technologies, Central and local * Uincluding distance education programsfor pupils and teacher trainees
Develop teacher training programs that Central and local * Ucombine residential and more fieldteaching practice
Adjust the structure of teaching Central * Uservice pay scales and rates ofpromotion to contain salary costs
Adjust teaching service pay-scales Central and local * Uto equalize incentives and statusbetween primary and secondary schoolteachers
Use teachers more intensively by Central and local * Uincreasing teaching hours, lengtheningthe school year, implementingdouble-shifts, and discouragingteacher absenteeism
Cut school construction costs Central * U
KEY: * = High * = Moderate or Low
8
Figure 1.2 (continued)Educational Policy and Budget Items
Impact Institutional TechnicalLocation Resistance Complexity
Employ new technologies to reduce Central and local * Uproduction costs of textbooksand instructional materials
Reduce rates of pupil drop-out Central and local * Uaind grade repetition
Raise pupil admission standards in Central and local * Usecondary schools and universitieswhere supply of graduates exceedslaboir demand
'rrack and reduce the number of 'ghost' Central * Eteachers who have no school assignment
Increase the ratio of teaching to Central and local * Unon-teaching staff
IReduce output of teacher training Central * Ucolleges when student demand is leveling
3. Allocating Sector Resources MoreEfficientlyRaise the proportion of recurrent Central and local * Asector spending allocated to essentialtextbooks and instructional materials
Consolidate curriculum and focus Central and local * Uinstiructional material expenditureson basic subjects (that are relevantto rural or urban needs)
Allocate teachers to equalize pupil: Central and local * Xteacher ratios across schools andgrade levels
Improve management information and Central * Uaccounting systems to track allocationof teachers, instructional materials,anci capital spending
KEY: * = High * = Moderate or Low
9
Figure 1.2 (continued)Educational Policy and Budget Items
Impact Institutional TechnicalLocation Resistance Complexity
Conduct research to identify those Central and local * Uschool inputs and classroom practicesthat effectively boost pupil achievement
Adjust the school calendar to avoid peak Local * Dperiods of labor demand (during theday or year), lowering the opportunitycost of staying in school
4. Diversifying Sources of Finance(lessening public sector burden)
Raise student fees at secondary and Central and local * Uuniversity levels, reserving a portionof revenue for pupils who cannotafford to pay
Encourage community contributions to Central and local * Uschool/classroom construction andrecurrent costs
Create book user-fees and revolving Central and local * efunds to support textbook production(especially at secondary and universitylevels)
Deregulate govemment controls over Central and local * Uprivate schools
Provide govemmenit incentives to pupils Central and local * Uand teachers in private schools, whencost-effective in expianding schoolplaces or improving quality
Encourage local, private production Central and local * Uof school furniture
Create student loan schemes for Central and local * Uuniversity students
KEY: * = High * = Moderate or Low
10
Figure 1.2 (continued)Educational Policy and Budget Items
Impact Institutional TechnicalLocation Resistance Complexity
S. Creating Incentives for Teachersand Headmasters
Improve national examination systems Central and local * m
to provide feedback to schools oneducational quality, controlling fordifferences in pupil background
Link a portion of teacher salaries Central and local * Uto performance in the classroom
Link a portion of teacher salaries Central and local S U(or stipends) to participation ininservice training
Create 'master teacher' roles within Local S Uschools to encourage professionalachievement and development
Support secondary and university Local * Kschooling through student scholarshipsor vouchers, boosting accountabilityof headmasters and teachers
Decentralize certain management Central and local * Ufunctions to encourage localresponsibility and accountability
Change the role of school inspectors Local * Uand headmasters to encourageimprovement in pedagogy andprofessional development activities
KEY: * = High * = Moderate or Low
11
tion, encouraging lower subsidies and stronger Policy choices and implementation.cost-recovery within the secondary and univer- 4. How can sector policy and budget changesity subsectors. The Bank has matched addi- more directly affect educational quality andtional recurrent expenditures for primary pupil achievement at the school level?school textbooks, staying within the govern- 5. How do altemative policy initiatives differment's aggregate spending ceiling (negotiated in terms of intended effects and likely con-with the IMF and the Bank). Policies aimed at straints on successful implementation?building the capacity of district education offic- 6. What are the critical conditions and keyes and even headmasters, within the Malawi factors that contribute to effective policy ad-operation, also provide positive incentives for justment?improvements in primary school effectiveness. 7. Are first-generation SAI operations send-Here the focus is on boosting pupil literacy, not ing cohesive, consistent signals? Or are theyonly accomplishing discrete shifts in central overly complex and occasionally contradictorybudgets. Also within the joint World Bank/ (for instance, urging school expansion andUSAID program in Mali, budget reform at the quality improvement simultaneously)?center is balanced with technical assistance invillage schools aimed at boosting the skills of Future Improvementsheadmasters and teachers.
8. Second-generation SAIs are now being de-Critical Q.uestions signed by govemments and donors. Can these
efforts encourage more ownership and broaderOur seminar brought together African gov- support within countries? Are we learning any-
emment officials and donors who have been thing about how policy levers can yield strong-involved with this first generation of education er positive effects at the school level?adjustment operations. Participants brought 9. How monitorable are policy reforms? Howtheir recent experience and a tentative list of do we disentangle discrete policy and budget in-lessons learned - from Ethiopia, Ghana, Leso- terventions from on going economic and policytho, Malawi, and Senegal. The chapters which drifts which affect the education sector?follow elaborate these specific country cases - 10. Are SAI strategies enabling governmentsand how a subset of policy initiatives are being to delineate their own policy choices and to de-implemented (or resisted) in particular set- velop information and evidence which informstings. To the government representatives and altemative courses of action?donors who attended the conference, we put Each country case which follows will touchforward a set of questions that were debated upon these issues. Our concluding section willthroughout the two-day session. As you read put forward some responses to these questions,through the country case studies, keep these is- based on the discussion that occurred at thesues in mind. Washington conference.
Donors' eagemess to pursue policy change,The context of long-term national develop- though more tempered and realistic than was
ment. apparent within the earliest experiments, will1. Africa's static economic growth and its at- not wane. Africa has been on the periphery of
tempt at economic adjustment: How does edu- the world economy for centuries. Contempo-cational policy change contribute to macro rary manifestations of this status - reliance oneconomic remedies? What windows of oppor- subsistence production, limited value-addedtunity are created for sector adjustment? manufacturing, and erratic commodity prices
2. How does policy change contribute to the and terms of trade with the world economy -long-run objective of strengthening institutions are not new to sub-Saharan Africa. The soaringand their leadership wilhin? level of foreign debt is a more recent facet of
3. How can sector adjustment and invest- political-economy, and one that will continuement strategies serve long-run national aspira- to constrain capital investment and the poten-tions? How will social and economic aspi- tial for growth.rations shape educational priorities, indepen- Thus the topic of macro-economic adjust-dent of rationalized adjustments? ment, and spending reforms specific to the edu-
12
cation sector, will move higher on the agendas the very least, inquire more carefully and in aof both governments and donors. Whether suc- hard-headed manner as to how costly policycessive generations of "adjusting" will contrib- experiments are touching the short-term ca-ute to long-run improvement in human pacity of governments - and the long-term ef-capacities remains to be seen. We should, at fectiveness of local schools.
13
Chapter 2
POLICY REFORM TO RAISE SCHOOL QUALITYThe Case of Malawi
Edward Ngaye
S ince winning its independence - just ventive financing strategies are pursued.25 years ago - Malawi has expanded Within this context Government began inschooling at a rapid pace. Primary 1985 to discuss with the World Bank a "policy
school enrollments have tripled; the number of action program" which includes a series of sec-youth entering secondary school has risen five- tor adjustments. This chapter will describe thefold. Enrollments continue to grow at 3.2 per- elements of this scheme. Throughout my dis-cent annually. Government remains committed cussion I emphasize the importance of moreto expanding access to basic education and in- careful planning within the education sector,creasing literacy.' including the formulation and broad discus-
Gradually, however, an awareness has sion of policy alternatives and choices. I willgrown that our success in expanding schooling not dwell on the merits or specifics of each poli-presents new problems, particularly in main- cy element. Instead, the discussion focuses ontaining minimum standards of quality and in the process by which ideas have been put for-using the sector's resources more efficiently. ward (from inside Government and from do-Downturns in the 1980s throughout the world nors). In addition, I raise the issue of how well-economy reminded educators in Malawi that intentioned proposals actually play out withinhigh aspirations for education must be bal- the broader social and economic environment.anced against resource constraints facing Gov- The paper is meant to stimulate thinking onemnment. Public spending on primary how educational reform can occur within aschooling, for instance, hcas risen just 0.6 per- process of planned change - taking into ac-cent per year (in real kwacha terms) since 1980. count the aspirations of a society or govern-Given that enrollments are growing at over 3 ment.percent, spending per pupil is declining at over2.5 percent each year! Education Development Objectives
Govemment's National Development Planrecognizes that gains in basic literacy are essen- Govemment's policy objective for the educa-tial for raising economic and social standards tion sector is to develop an efficient schoolingof living. Boosting literacy requires a balanced system of a type and size appropriate to availa-strategy aimed at expanding basic schooling ble resources and to the political, social, and ec-and enhancing educational quality. This strat- onomic goals of the nation. Education is seenegy will be effective only if (a) existing sector re- as an important vehicle for promoting nation-sources are focused on those inputs and al consciousness and cohesion, and for further-practices that most directly boost pupil achieve- ing economic independence. Schooling isment, and (b) new sources of revenue and in- viewed as a means for reinforcing the high eth-
14
ical standards of Malawian society, essential if rent government spending should in-social justice is to be maintained in the face of crease from 11 percent to 15 percenteconomic progress. Education also plays a role by 1995. This goal is included in Gov-in providing workers who hold skills that are in emient's current ten-year develop-short supply, thereby aiding national develop- ment plan, and has become anment. important ingredient of ouIr policy
The Second Education Development Plan for action program.the nation (1985-1995) spells out these priority * Variation in access to basic educa-functions of education. In pursuing these tion. Disparities exist in the distribu-broad objectives, the Ministry of Education and tion of physical and humanCllture attempts to focus on three particular resources in the educational system.concerns: improving equitable access to educa- This is a situation that the Ministrytional opportunity, achieving a proper balance intends to redress in our developmentin resources allocated to different levels of in- programs.struction, and raising the efficiency of the edu- * Low internal efficiency. The educa-cational system. tion sector is not as efficient as it
could be, especially at the primaryIdentifying Key Constraints level. Repetition rates are quite high
and Problems in most class levels, ranging from 5percent to 42 percent. Only one-
Overall policy and budget actions must be in fourth of all pupils who enter willline with Government's development objec- complete the full eight years of pri-tives. And policy improvements should address mary school.those pressing constraints that limit the educa- * Inadequate instructional materials.tion sector's capacity to achieve these objec- Basic textbooks, teacher guides, andtives. Delineation of priority issues stems from writing paper are in very short sup-lengthy dialogue with donors, and from Gov- ply. Teachers report that just one Chi-emment's own recent review of the education chewa textbook is available for everyservice (conducted by Price Waterhouse). Princi- four children. The ratio of students topal problems facing the education sector in- pupil desks equals 13:1! Together,clude the following: Government and Malawian parents
(through fees) can afford to spend- Population growth. Child popula- less than three kwacha per pupil on
tion appears to be growing at 3.9 per- instructional materials annually.cent a year, indicating that demand This covers the cost of one-half of onefor formal and nonformal education textbook (pupils are supposed towill exceed supply for many years to have three standard texts). The ratiocome. Formulation of programs that of pupils per teacher has grown dra-ensure adequate access to education matically since 1970, rising fromand training opportunities is essen- 41:1 to 67:1.tial. This will enable youth to play a * Insufficient places in seconrdarycatalytic role in the economic and so- school. The intake capacity of all sec-cial development of Malawi. ondary schools is only 7,000 pupils.
* Limited resources. Expansion of ba- But the number of pupils who passsic education will occur within tight the primary school leaving exam,resource constraints. We believe that and thus become eligible for a secon-the education sector has been under- dary place, now exceeds 60,000 eachfunded relative to other sectors and year. This means that only about 12that funding does not reflect school- percent of qualified pupils are able toing's potent contribution to economic enter secondary school. This bottle-growth. Both the Price Waterhouse neck is the direct result of inadequateteam and the World Bank have sug- infrastructure.gested that the sector's share of recur-
15
A Poicy Action Program 4. The Ministry is attempting to strengthencommercial and vocational subjects at both
Ministries of education are constantly discuss- primary and secondary schools. These effortsing, and pushing for, possible policy change have run into problems: equipment used forwith other levels of government Donors must training is not always appropriate, facilities arerecognize that their particular ideas for policy costly and not fully utilized. But a clear policyor budget reform are injected into institutions focus should lead to improvements in the qual-where many options already are being debat- ity of vocational instruction.ed. In addition, forces outside ministries of edu-cation often are responsible for policy change. Quality and efficiency reforms. Several poli-These "reforms" - resulling from extemal eco- cy actions attempt to enhance educationalnomic or political considerations - are simply quality and raise the efficiency of the educationbeyond the control of the education sector. sector. These reforms also involve two or more
Keeping in mind these two contextual points, levels of the system.let me describe major policy changes that have 5. Efforts are being made to concentrateoccurred, or continue to unfold, within Mala- scarce resources on pupils who are successfullywi's education sector. Several of these reforms persisting through primary school. Almost halfhave been negotiated with the World Bank in of all standard 8 pupils are repeating, trying torecent years. Others sterm from the Govern- get into secondary school. Repetition at allment's own actions as we address the sector's standards increases class size and diverts re-major problems. sources away from pupils who are achieving at
satisfactory levels. The Ministry is proposingOrganizational reforms. The Ministry is imple- that entry to secondary school be heavilymenting several institutional reforms that cut weighted in favor of pupils who have not re-across all levels of the sector. peated. This adjustment to the incentive struc-
1. Since 1977, all teachers have been part of ture may improve the efficiency of the primarythe civil service, helping to raise their status subsector.and pay. In addition, Government has intro- 6. Increased utilization of school facilitiesduced new promotional grades for primary and stronger cost-recovery measures have beenschool teachers and headmasters (a T-1 rank- introduced. Government subsidies for boardinging and three grades of ciploma teachers). facilities, benefiting secondary school and uni-These adjustments to career ladders encourage versity pupils, have been reduced. The Ministrylonger tenure in the teaching service and pro- is considering how more day students (non-vide incentives for higher performance. boarders) can be enrolled in secondary schools.
2. The recent Price Waterhouse review recom- Double shifts of classes are being introduced onmended decentralization of certain administra- an experimental basis. Govemment hastive functions now controlled by the central agreed to raise the pupil:teacher ratio at theMinistry. Many time-consuming problems, university (which now stands at 7:1). The uni-such as teacher leaves and pupil discipline cas- versity has relaxed its policy requiring that alles, now must be decided by high-ranking Min- pupils be in residence and has already admit-istry officials. Decentralization of these ted 36 day students. These measures seek to ex-functions would free time for "policymakers" to pand educational opportunity at littleactually deal with major policy and fiscal is- additional cost.sues. We are presently developing an imple- 7. Government has agreed to raise the pro-mentation plan. portion of education spending allocated to pri-
3. Previously two separate agencies man- mary schools. This subsector exercises theaged national examinations. The Parliament largest impact on boosting Malawi's literacyrecently approved legislation to consolidate rate. In turn, rising literacy should improvethese offices into one Nalional Examinations productivity and the social quality of life in ru-Board, leading to reduced bureaucratic costs. ral areas. (Over 80 percent of Malawi's popula-The new agency is exploring ways of improv- tion reside and work in rural villages.)ing exams and administering them more effi- 8. The finance ministry has agreed to doubleciently. Governmenfts support of essential textbooks
16
and instructional materials. The Ministry will strong political will, technical innovation, andnoNv cover about one-half the cost of these ma- financial resources. These essential ingredientsterials, with parent fees picking up the remain- will come together only through strong, lastinging cost. collaboration between governments and do-
9. The Ministry, working with the university, nors. Participatory planning by all partieshas enmbarked on a research program that should be emphasized, through extensive con-aims to identify those school inputs and teach- sultation and open workshops.ing practices that effectively boost pupil Third, policy change must be combined with pos-achievement. These findings will feed into the itive incentives. These adjustments often createon-going dialogue on what policy and budget anxiety, as change occurs in fee structures andlevers can directly improve school-level actions budgetary allocations, or locally in terms ofwhich increase basic literacy. school-level practices (regarding policies related
to repetition or national examinations). Actorsirnmary school reforms. Several elements of in the educational system should not be ex-
our policy action program speak to particular pected to change in the absence of incentives.subsectors. Let me highlight reformns specific to Our policy action program therefore includesprimary schooling. several positive incentives: efficiency measures
10. The primary school curriculum is being open-up new places in secondary schools andrevised to make it more consistent with our so- the university; preservice and inservice teachercial-cultural environment. The number of sub- training is improved and stipencls encouragejects also is being reduced, allowing teachers to participation; clearer rewards are provided toconcentrate on a more manacgeable curricu- primary school pupils who progress at normallum. levels.
1.1. To increase educational access, the Minis- Fourth, the policy adjustment pnocess must be re-try is proposing elimination of all fees for stan- spectful of the nation's social, political, and eco-dards 1-4. Methods for implementing this nomic aspirations. For instance, donors areproposal, particularly how Government will re- emphasizing the importance of educationalplace the loss in fee revenue, are now being dis- quality and using scarce sector resources morecussel. efficiently. But gains in this area must be bal-
12. The Ministry is experimenting with differ- anced against expanding access to basic edu-ent ways of upgrading the skills of underquali- cation. Just 40 percent of all children in Malawified primary teachers and cost-effective are enrolled in primary school. Improvementsstrategies for producing more teachers. A crash in quality will benefit only the relatively fewprogram to train 4,500 new teachers is being children who actually enter school. The socie-implemented. This effort involves a combina- ty's aspiration for universal access to schoolingtion of training in residence, practice teaching also must be addressed.in the field, and correspondence courses. Fifth, local communities and par-ents must be in-
volved in the policy reform process. At a mini-Lessons Learned mum, local villages should understand why
changes are occurring, for instance, modifica-Discussion of these policy reforms began five tions in the fee structure or in the curriculum.
years ago. The more complex process of imple- To take another example, we cannot simplymentation has been underway for just three decree that pupil repetition rates will be re-years. Several tentative lessons have emerged duced. Central ministries must work with par-as the Ministry of Education and Government ents, teachers, and headmasters if school-levelpush these basic policy and budget changes. practices are to be improved.
First, only a few reforms should be introducedat one time. The idea of policy adjustment is stillnoveL And the human capacity of ministries is NOTE:fixed., Only so many fundamental changes can 1. This paper was originally presented at the USAID/be taken on over a short period of time. World Bank seminar on Educational Policy Adjustment
Second, policy reforms should not be imposed held in Washington. This paper does not necessarily re-from the outside. True policy change requires flect the official policies of any institution.
17
Chapter 3
GHANA'S POLICY ADJUSTMENT INITIATIVEOpportunity for Renewal
Vida Yeboah
"'The process of policymaking is not simple... it involves, for example,having thousands of students protesting outsideyour office, asking whyyou are working with the World
Bank on budget reforms (including cuts in govemment subsidy of pupil boarding costs)."- Dr. Ebi Sutherland Addey
Deputy Prinpal Secretary for Education, Ghana
I n 1987, after a decade of institutional Ghana's relentless economic decline, deeplyerosion, the Government of Ghana em- felt since the mid-1970s, seriously undercutbarked on a broad program aimed at earlier progress in the education sector. Gov-
revitalizing our educational system. The ini- emient's eventual decision to tackle sober-tial thrust was to entitle every child to nine ing adjustments to economic and publicyears of basic education, through junior sec- finance policies, however, has provided aondary school. But the reform effort has be- window of opportunity to move on our broadcome more ambitious, addressing issues of restructuring of the educational system. Poli-access, educational quality, and resource effi- cy and budget adjustments have been veryciency. difficult and painful. But they have sparked
Within Government's broad efforts to renew Government's long-term initiative to revital-the educational system lay several policy ad- ize schooling.justment measures. These policy and budgetactions realign government subsidies of Backgroundschooling in more equitable ways, as by en-couraging higher quality basic education for The desirability of serious educational reformthe country's poorest families. Working with - improving the basic institutional structure,international donors, Government also is at- content, and patterns of finance and subsidy,tempting to manage more carefully budgets and content - had been recognized in Ghanaand expenditures of various educational insti- for 15 years. Indeed, in 1974, the organizationaltutions. We are trying, against considerable and budgetary changes now being implement-opposition, to reduce unnecessary or ineffi- ed were first detailed in the Dzobo Committeecient expenditures. This includes cutting sala- Report and approved by Government ('Theries for non-teaching staff, containing growth New Structure and Content of Education forof the teaching force, and scaling back subsi- Ghana"). A consolidated education service wasdies of secondary school and university stu- created in that year in response to the Dzobo Re-dents when they can afford to cover a portion port.of instructional and housing costs. We are re- Due to previous governments' lack of resourc-allocating the resources thereby conserved to es and political will, however, other major rec-increase the availability of essential instruc- ommendations were not taken up. They simplytional materials, including textbooks, exercise languished for the following decade and a half.books, and writing paper. By generating cost By 1983, Ghana's educational system had sunksavings on one side, we also free up resources to very low levels. The symptoms of declineto expand in-service teacher training. were severe:
18
* Given the mass exodus of well- adjustment. We did not want to be dependenttrained teachers from the country, upon donor resources once these significanthalf of all primary and secondary shifts in structure, policy, and budget were un-teaching posts were filled by un- dertaken.trained or unqualified teachers.
* Enrollment increases had declined to Ghana's Reform Programwell below growth in child popula-tion. Primary school enrollments The education reforrn package implementswere rising at just 1.5 percent per significant changes at all three levels of school-year, and even less at the junior sec- ing - primary, secondary, and the university.orndary level. In addition, adjustments were made in teacherSuccessive governments severely cut training institutions and within the organiza-back on resources allocated to basic tion of nonformal education.education. The proportion of GDP de- Basic educatiorL The Education Reform Pro-voted to education fell from over 6 gram, announced in 1987, implemented a ma-percent to just 1 percent between jor change in the system's basic structure. We1976 and 1983. In 1985, real spend- moved from 17 years of primary and secondaryirLg on education equaled only one- schooling (64-7) to the more typical 12 yearsthird the level of resources allocated (6-3-3). At the same time, Government dedaredten years earlier. that all children are universally entitled to nine
a Miost school children were without years of free basic education, through juniortextbooks, paper, and pencils. secondary school. Our objective in reducing theSchool buildings and furniture had length of secondary school was to lower the op-tloroughly deteriorated, given scarce portunity cost of keeping children in school - aresources for maintenance and re- cost that had been quite high for many parents.placement. The following principles formed the founda-T The sector was devoid of any data tion of our reforms in basic education:and statistics necessary for describing * Government's affirmation of the ma-thle educational system, let alone jor policy decision that every child,planning for the future. age 6-15, has the basic right to be
able to read, write, and function use-I]n 1983, the present government decided to fully in society.
take strong action to arrest this surprisingly * The participation of every Ghanaianrapid decline of the educational system. The is extremely important if the countrynew policy was implemented by the Provi- is to develop. But, sadly, most Gha-sional National Defense Council (PNDC) un- naians are cut off from participationder the broad Economic Recovery Program. In because they are either illiterate or1986, Government pursued more focused dis- have been miseducated and are un-cussions of policy and budget reforms specific able to realize their potential.to the education sector. Top political leaders * Every Ghanaian needs a sense of cul-acknowledged that development of Ghana's tural identity and dignity. Our coun-human resources was a vital prerequisite to try has a heritage of individualsocial and economic growth. ethnic cultures and of promoting a
As a result of these high-level deliberations, unified culture of Ghana to ensure athe new Education Reform Program was sense of national identity. As the na-drawn up to ensure that national educational tion grows stronger, people will begoals could be realistically met. Emphasis was more proud of themselves and theirplaced on increasing access to, and boosting society. A proper cultural identity willthe quality of, basic education. We also in- free our minds from dependence ontended to make basic schooling more relevant the cultures of other people.to the people's socioeconomic conditions. The * Since education has not been rele-program was designed, from its inception, to vant to their situation, many youthbe sustainable after an initial period of policy leave school and cannot find work
19
* Today's world is a scientific and tech- courses; andnological one; even a minimal edu- * reviewing the cost of finishing un-cation must tune eveiry child's mind completed buildings and improve-to this fact if he or she is to live com- ments in university facilities.petently.
* Every Ghanaian must be taught The committee has made a full series of rec-about his or her environment. Prob- ommendations which Government is presentlylems such as deforestation, low agri- reviewing. In the meantime, Government hascultural productivity, and widespread moved to reduce subsidies for instruction anddisease will be minimized if Ghana- housing for those students who can afford toians are taught how to prevent these share the cost burden. These policy and budgetafflictions as a part of their basic edu- reforms have been painful and politically haz-cation. ardous. But they demonstrate Government's re-
solve to make educational finance moreThe nine-year basic education program is ter- equitable and to conserve public resources for
minal for many children. Some school leavers basic education.may enter apprenticeship or other training. Nonformal education. Adult literacy pro-Many youth will continue on to junior secon- grams have grown dramatically in Ghanadary school after completing the primary stan- since the late 1970s. This occurred, in part, duedards. to rising drop-out rates in the formal school sys-
Senior secondary education. The traditional tem. The reform program has attempted to ra-secondary school was sharply criticized for tionalize management and contain costs in theplacing too much emphasis on academic nonformal subsector.work and for being far removed from local sit- The first step was the creation of a formal Di-uations and national manpower require- vision of Nonformal Education within the Min-ments. The senior secondary program will istry. This allowed us to consolidate a wideseek to correct these defects and provide stud- array of programs that had arisen over the pasties that promote individual and national de- decade to meet rising popular demand. As wevelopment. The new senior secondary see the social costs of adjustment ever more viv-curriculum builds upon skills provided in ba- idly in Ghana, the ministry is redoubling its ef-sic education, then strengthens intellectual forts to boost adult literacy.competencies and knowledge required for ei- Teacher education. The reform programther higher level employment or university also seeks to improve teacher training institu-education. tions. Indeed, teachers constitute the key fac-
Universities and tertiaiy education. In 1986, tor to the entire school improvement process.the Ministry of Education and Culture created Much depends upon teachers' commitmenta committee of experts to study the current and competence. Consequently, we are intro-institutional structure and its problems. The ducing into teacher education programsstudy's objective was to provide basic descrip- training in the new basic education course oftive information and recommendations on study.medium-term development for university In the past, Ghana had both three- andcampuses and other tertiary training schools. four-year preservice training programs, de-Specifically, the committee focused on: pending upon the teacher candidate's prior
level of schooling. Under the reform program,* the cost and pricing structure of uni- Government is phasing out the four-year pro-
versities and an assessment of where grams. Beginning three years from now, theexpenditures could be increased or improved three-year preservice programs willdecreased so as to imaprove alloca- admit only applicants who have completedtional efficiency; senior secondary school. This means that new
* assessing the demand for, and sup- teacher trainees will have higher-level aca-ply of, existing course offerings and demic skills and more relevant knowledgerecommending how costs could be when they enter the training college. At thecut through a more rational array of same time, unit costs of graduating a new ba-
20
sic education teacher will decline. We expect and secondary schools.that the teacher training curriculum will be A public information campaign was under-more professional, focusing on pedagogical taken during the first three months of 1987. Wemethods rather than spending time reviewing described to numerous groups the state of basicthe primary or secondary school curricula education and explained why the Education Re-(which trainees should already know). form Program was necessary. As mentioned
above, several policy and budget adjustmentsDonor Collaboration and Funding were not easy, undercutting the interests of well-
established groups in Ghanaian society. WeGovernment approached several intemation- therefore wanted to carefully and persuasively
al agencies in 1986, seeking their support and articulate our goals and the specific changes be-financial backing for the Educational Reform ing undertaken within the Ministry.Program. By late 1987 we had successfully ne- In late 1987, the first tranche of funds fromgoticited with the World Bank what became the World Bank and co-financiers was released.known as the education sector adjustrnent credit This financing recognized our success in adjust-(or EdSAC). A major component of EdSAC in- ing subsidy and pricing structures in the sector.volved the series of policy and budget adjust- Of equal importance, these funds allowed us toments which helped to reinforce Govemment's intensify work in the following areas:broader program.
As surnmarized at the beginning of this chap- * Finalizing the curriculum develop-ter, these policy measures realign Govemment ment strategy for junior secondarysubsidies of education in more equitable ways, schools;encouraging higher quality basic education for * Producing exercise books and sta-the country's poorest families. Unnecessary or tionery;inefficient expenditures on student housing, sal- * Publishing textbooks within Ghana;aries for non-teaching staff, and the old lengthy * Procuring teaching materials for ba-secondary system were cut significantly. Gov- sic science, agriculture, and technicalernrnent is attempting to manage more careful- training;ly budgets of various educational institutions. * Initiating inservice training for juniorAs we conserve budget resources, we can raise secondary teachers;spending on instructional materials, including * Setting up distribution systerrms fortextbooks, exercise books, and in-service train- the new instructional materials;ing. * Preparing new inspectors and re-
This agreement with the World Bank on poli- vamping the supervision system; andcy adjustments helped draw resources not only * Creating a new unit for project man-from IBRD (US$34 million) but from bilateral agement.donors as well, induding Norway, Great Brit-ain, Switzerland, and the OPEC fund. The Min- In addition, a National Planning Committeeistry is now beginning to implement a second was formed to help implement the junior secon-budget support and technical assistance pack- dary school program. Members were drawnage financed by the Bank and USAID. from the education ministry, Ghana Education
Service, the national teachers association, uni-Implementation Process versities, and the National Service Scheme. Re-
and Constraints gional and district implementation committeeswere formed to facilitate improvements in the
Prior to announcing the reform program, six junior secondary schools. A national task forcemonths were spent planning steps toward im- also was formed to work exclusively on logisti-plermentation. The Ministry of Education and cal issues, linked to dissemination of instruc-Culture led this process of laying out the basic tional materials and the renovation of selectedstrcategy, then articulating concrete implemen- educational facilities.tation tasks. Our basic management data were Following a review of our progress, the secondsketchy - pertaining to the teaching force, en- tranche of donor funds was released in mid-rollments, and even the location of primary 1988. We are presently working to complete the
21
transition to the three-year junior secondary motivated by political or personal interests, con-school. This involves redrafting syllabi and tinue to loudly oppose the reform program.some textbooks, and providing new equipment. In a way, student protests against the reformsIn addition, we are trying to upgrade facilities can be said to be motivated by a vested interest.at senior secondary schools. The issue of cost-recovery involves removing
Implementation constraints. The shortage of boarding and feeding subsidies. This has gener-time represents a major constraint on our mo- ated numerous protests from university studentsmentum. Since none of the policy measures who have refused over the years to consider al-seemed particularly neiw within ministry circles, ternative means of finance. Within the pastwe may have underestimated the time required year, these protests have led to closure of theto explain the reform program. Opposition to university. The debate continues over how thepolicy and organizational changes keeps recur- cost of housing can be shared, lessening thering, requiring new rounds of explanation and burden placed on the government budget andargument. Major planning and implementa- increasing the equity with which education sub-tion activities are coordinated by a relatively sidies are distributed.small group within Government. Implementa- Other policy conditionalities have been diffi-tion suffers when key members are pulled away cult to maintain over time. The general freezeto respond to political resistance and continual on the recruitment of untrained teachers stillmarketing of the progr(am. stands. This has been extraordinarily difficult,
The structural adjustrnent program was ini- however, since rural education authorities havetiated before we had time to improve basic traditionally hired untrained teachers outside ofmanagement structures within the Ministry of the regular hiring process. We also have re-Education and Culture. This process has proven duced the number of excess non-teaching staffquite complicated. In addition, the comprehen- within junior secondary schools. Generatingsive reform program involved a sharp restruc- cost savings, then ensuring that these resourcesturing of junior and senior secondary schools. are allocated for increasing instructional mate-Even as Ghanaian parents were figuring out the rials, is difficult politically and technically com-nature of these new inslitutions, they were told plex to implement. Central determination ofthat fees and prices would change significantly these policy changes is the critical first step.as well. Too many changes - both policy and Long-term implementation, and the persever-organizational - may have been attempted si- ence required, have proven just as problematic.multaneously. We are now trying to consolidatethe office that should hold authority for specific In Summaryareas of reformn. Originally many branches ofGovernment were involved both in the educa- The period of planning and implementation,tion ministry and in thE Ministry of Finance and as described above, has been long and fraughtEconomic Planning. with unanticipated difficulties. We have
In the past, the Ghana Education Service and learned much about trying to simultaneouslythe universities have operated as semi- adjust policies, budgets, and school structures.autonomous units under the education minis- Shifting the organization of central ministrytry. The budget of each institution had grown in management, while at the same time reachinga rather uncoordinated fashion. Even where in- out to various public constituencies, also hasefficiencies were apparent, the education minis- proven very difficult.try's authority to remedy the situation was Despite these complexities and constraints,ambiguous. We are now working with both or- the PNDC government is committed to fully im-ganizations to set budgeting and manpower tar- plementing the Education Reform Program.gets that help keep expenditures under control. This will ensure that our educational system
Another implementation constraint involves ceases to promote only the interests of the few.on-going opposition from vested interests. It has Instead, we are making basic schooling morebeen possible to win gradually the support of accessible to more people at increasingly higherthose who had genuine fears about the chang- levels, enabling Ghanaians to develop theires, concems resulting mainly from a lack of own capacities and thus accelerate develop-knowledge about the program. But other actors, ment of the nation.
22
Chapter 41
SOCIAL AND POLITICAL CONSTRAINTS ON EDUCATION REFORMThe Case Of Senegal
Birger Fredriksen
f'LThis chapter reviews Senegal's efforts However, at independence, enrollment ratiosduring the 1980s to accelerate the de- were modest at these levels of education as wellvelopment of primary education, -2 percent in secondary education and 0.5
mainly through policy reforms aimed at allo- percent in higher education, as compared tocating scarce resources more efficiently. I intro- SSA-averages of 2 percent and 0.2 percent, re-duce this discussion by describing the general spectively.development of education policy in Senegal During the first two decades of independence,during the 1980s. Yet I focus on issues related Senegal made major strides in developing pri-to the preparation and implementation of the mary education, despite weak economiceducation policy package promoted through growth (GCP per capita declined at an averagethe Primary Education Development Project annual rate of around 0.5 percent between(Education IV), supported by the World Bank 1965 and 1980). Strong factors fueling theand the African Development Bank.1 The dis- growth in the supply of education includedcussion highlights social and political con- sharp increases in the overall Government bud-straints faced by governments in countries such get (from 8.4 percent of GDP in 1960 to 20.3as Senegal in their attempts to introduce educa- percent in 1980) and in the share devoted totion policy reforms. The chapter concludes by education (from around 15 percent in the earlysuggesting lessons that may be drawn from the 1960s to 23.5 percent in 1980). Together, theseSenegal case. two factors resulted in an increase in the share
of GDP devoted to education from around 1The Setfing percent to around 5 percent during this twenty-
year period.Developments up to 1980. Despite Senegal's Despite the fact that these efforts permitted a
traditional role as a center for education and tripling in the number of primary school pupilstraining in French-speaking West Africa, at in- during the first two decades of independence,dependence development of primary education the gap in coverage of primary education be-(27 percent gross primary enrollment ratio in tween Senegal and the rest of SSA widened. As1960) was well below the sub-Saharan Africa a result, Senegal entered the 1980s with a pri-(SSA) average of 38 percent2 Senegal's reputa- mary enrollment ratio of 46 percent, comparedtion as an education center derived mainly to the regional average of 73 percent.from its institutions of secondary and higher Recommendations of the Education Com-education which, during the colonial period, mission. The slow development of primary ed-served not only Senegal but most other French ucation, combined with factors such ascolonies in West Africa, particularly the Sahel. dissatisfaction on part of parents with educa-
23
tion content, desire to clive national languages universal primary education by the year 2000.a more important role as a means of instruc- One main cause was accelerated growth of thetion, and budgetary constraints led the Govern- child population (of primary school age), fromment in January 1981 to convene a major 2.6 percent per year during the 1970s to a pro-national conference (les Etats Generaux) on jected average annual rate of 3.1 percent forthe future of education in Senegal.3 This confer- the period 1980-2000.ence made recommendations regarding: Second, the education sector's increasingly se-
vere squeeze between rapidly rising costs and* the creation of a "new school" that rigid budget ceilings made it clear that it might
would gradually lead to universal ed- not be possible even to continue the past trend.ucation for the age-group 3-16 years; The country's economic condition had wors-
ened, unit costs in primary and secondary re-* the resources required to implement mained higher than in other low-income
this new policy; and African countries, and the share of the Govern-ment budget devoted to education was as high
* the training and conditions of em- as 26.5 percent in 1984. While there is no obvi-ployment of education personnel, ous ceiling for the share of public budgets thatparticularly teachers. can be devoted to education, in practice each
country has a threshold beyond which the fi-These recommendations were adopted by the nancial burden of education may seriously
Government in February 1981, and a National cripple the supply of other public services. AndCommission on the Reform of Education and in light of the increasing burden of debt service,Training (CNRED) was established to work out it could be argued that this threshold was beingthe details of the reform as regards objectives, approached in Senegal.resources and structure of the new school. The Third, while the Government was able to in-Commission made its recommendations in crease the share of the education budget devot-July 1984; these were approved with minor ed to primary education from about 40 percentmodifications by the Govemment in January in 1980 to 46 percent in 1985, it became appar-1985.4 ent that further reallocations in favor of pri-
The recommended reforms aimed to make mary education would be necessary in order tothe educational system. more equitable, effi- maintain sufficiently high enrollment growth.cient and responsive to national development This would imply continued containment ofneeds. One key objective was to attain, by the expenditures in higher education, particularlyyear 2000, 30 percent enrollment for the age- of student subsidies which in 1985 accountedgroup 3-6 years, and 100 percent enrollment for 36.5 percent of the higher education budgetfor the age-group 7-16 years. Within the latter, (25 percent for scholarships and 11.5 percentthe group 7-12 years would be taught in regu- for campus services providing subsidizedlar schools to be provided at the village level, meals, lodging, transport, and medical servic-while education for the group 13-16 years es). Sixty percent of all university students re-would be provided in local centers and would ceived financial assistance. Over half of therely heavily on cooperation with local produc- scholarship budget was for study abroad, thetion units in the traditional and modem sec- per-pupil cost of which was five times highertors. than at the University of Dakar.
Key sector issues in the mid-1980s. While Fourth, education quality was low and de-these policy reforms were being detailed, the se- clining at all levels. As regards primary educa-riousness of constraints on future education- tion, the success rate at the primary schoolsector growth became even more evident. First, leaving examination declined from 59 percentwhile enrollment in pri.mary education grew at in 1966 to 42 percent in 1983. The factors ad-an average annual rate of 6.8 percent over the versely affecting quality included: poor leamingperiod 1980-85, resulting in an increase in the efficiency due to severe shortage of textbooksenrollment ratio from 46.0 percent to 55.5 per- and other training materials, overcrowdedcent, continuation of this rather healthy trend classrooms in urban areas (in 1983-84, 27 per-would not lead to attainment of the target of cent of urban classrooms had between 70 and
24
120 pupils), a short effective school year (due to and construction and rehabilitation of class-many holidays and long school vacations, rooms.even the official school year was less than 120 The policy component has two main elements.days), and inadequate physical facilities (about The first includes measures designed to increasehalf of the classrooms had an insufficient num- primary school enrollment through reductionber oif pupil benches, and one quarter of the in unit costs by means of more efficient use ofclassroorns were in a state of serious disrepair); teachers and facilities. The second element in-overly academic content of the curriculum and cludes measures aimed at containment of ex-pupils' often limited comprehension of the lan- penditures in post-primary education andguage of instruction during the initial years of reallocation of the savings to primary educa-schooling; and inadequate management and con- tion while maintaining the share of the Gov-trol at the regional level due to a lack of means emment's budget devoted to education at itsfor inspectorate staff to visit schools and en- 1985 level (23.2 percent).5 More specifically, thecourage good teaching performance. measures included in each of these two ele-
In response to these constraints on the educa- ments are as follows (the description below istion sector, the Govemment developed a set of based on the Staff Appraisal Report).sectoral adjustment policies aimed at promot- Reduction of unit costs. This element in-ing the clevelopment of primary education cludes five measures, four of which are directlythrough more efficient allocation and use of related to the training and use of primaryexisting resources. The Government's strategy school teachers:also included a number of actions to improve 1. Recruitment of candidates into primary schoolthe quality and relevance of education, and teacher training programs at an instituteur-adjointlfixed the year 2000 as the target for attainment instituteur ratio of 80:20 as of the stcart of theof universal primary education, with an inter- 1986-87 school year.6 In 1984-85, 47 percent ofmediate target of 60 percent at the end of the the primary school teachers in Senegal were in-Seventh Plan in 1989. This strategy became an stituteur-adjoints and 53 percent instituteurs.integiral part of the Education IV, to which I The objective of this measure is to r educe unitnow tumn. costs since the salary difference between the
two category of teachers is about 30 percent.Adjustment Policies Promoted 2. Redeployment into teaching positions of 400
through Education IV primary school teachers holding adminnistrativeposts in inspectorates and secondary schools -
Education IV became effective in September 200 persons would be redeployed cluring 1986-1987. Total project support equals US$19.6 mil- 87 and 1987-88, and 200 during 1988-89 andlion, indcluding $5.3 million in co-financing by 1989-90.the African Development Bank. Although a 3. Introduction of double-shift teaching in urbantraditional investment operation, the project il- areas and multi-grade teaching in rural areas. Thelustrates the new type of policy-oriented pro- former system was to be tested in 75 classroomsgrams which emerged in the World Bank's during two school years starting in 1987-88.West Africa office during the mid-1980s. The The system chosen would include reducing theadjustment measures promoted through the number of weekly curriculum hours for stu-project are in most respects as comprehensive dents and increasing the teaching load foras those promoted through today's sectoral ad- teachers, so that one teacher, usinig one class-justrnent projects or through hybrids (programs room, would teach two groups of pupils, one inthat provide quick-disbursing funds released in the morning and one in the aftemroon. The sys-tranches against progress in implementation of tem would gradually be applied in Grades 1-4agreed-upon education reforms). and was estimated to allow about 20,000 addi-
Education IV has two main components. The tional pupils to gain access to prinmary educa-investment component focuses almost exclusive- tion by the end of the project (199 4). Multi-ly on1 piimary education - development and grade teaching would be experimented with inprinting of textbooks; strengthening of inspec- 25 classrooms. Teachers involved in these twotorates, education planning and research capa- types of teaching would receive a salary in-bilities, and project implementation capacities; crease of 25 percent.
25
4. Increase the pupil-teacher ratio in primary about 80 percent of the primary school budgetschool teacher training colleges from 7:1 to 15:1 in 1985-86.by merging two existing colleges by 1987-88, Apart from monitoring project implementa-and two more prior to 1988-89. tion during regular supervision missions, the
5. Increase student-teacher ratios and reduction Credit Agreement called for the organization ofin operating costs of specialized civil service train- two special implementation reviews (in Juneing schools. Complete feasibility study and im- 1988 and June 1990) to assess progress inplement plan by February 1988. Consolidate achieving the targets of the adjustment pro-schools as per plan of 19188-89. gram. Disbursement on 350 of the 400 class-
Containment of expenditures in post- rooms to be financed under the project wouldprimary education. be conditioned on satisfactory implementation
6. Reduction of subsidies to higher education by: of the policy program; funds for 175 classroomsreducing the scholarship budget from the 1985- would be released following each of these re-86 level by 8 percent in :L987-88, and by 3 per- views, implementation progress permitting.cent annually from 1988-89 through 1994-95; Before reviewing progress in implementingestablishing a cost-recovery system for higher the above program, three aspects related to itseducation based on a feasibility study (to be formulation should be emphasized. First, thecompleted by June 1988) on the introduction of criteria for the June 1988 and June 1990 imple-a student loan system, which would be tested mentation reviews include a number of activi-by making a few loans to third- and fourth- ties other than the seven sets of measuresyear students during 1988-89 through 1990-91, described above. These other activities relateand by evaluating the percentage of loan re- mainly to the strengthening of institutional ca-covery between 1991 and 1993; and reducing pacities in areas such as: (a) facilities planningindirect subsidies to higher education, meaning and maintenance, construction managementthat the budget for the University of Dakar and procurement, and design of constructionCampus Services would be reduced in real techniques for durable, low-cost primary schoolterms from its 1985-86 level by 8 percent from classrooms; (b) education planning and cost1987-88 through 1994/95, and also that a sur- control; (c) education research and develop-vey would be conducted by March 1988 on cost ment of textbooks; and (d) decentralized educa-reduction measures to be introduced for the tion management. These measures are closelyprovision of essential services, including privat- related to implementation of the project's in-ization of certain services and cost-recovery vestment program.methods. Second, all reforms included in the program
7. Establishment of ceilings on budgetary growth were designed to be implemented gradually:for recurrent expenditures on central administration for example, redeployment of administrativeand on secondary and higher education. Limits staff into primary schools would take placeequal to one percent annually for central ad- over a period of four school years; curriculumministration and higher education, in real reforms, textbooks, use of double-shift andterms, and up to 1.5 percent annually for sec- multi-grade teaching techniques, and introduc-ondary education (both rates in real terms). A tion of low-cost dassroom construction tech-study on cost-saving measures in secondary niques were all to be properly tested prior toand higher education would be completed by implementation. To facilitate development ofJanuary 1988. these measures, several studies would be con-
ducted and financed under the project.General Elements of Third, while practically all the project invest-the Policy Package ments benefitted primary education, many of
the most difficult reforms - particularly thoseIt was estimated that iimplementation of the calling for budgetary containment - would af-
above policy package would allow for an addi- fect post-primary education. As we shall see,tional enrollment increase of 180,000 pupils, or this has become a serious problem during pro-31 percent, between 1987-88 and 1994-95. In ject implementation, particularly after a separ-monetary termns, the cumulative cost savings ate Ministry of Higher Education wasover the project period would correspond to established in early 1988.
26
erage pupil:teacher ratio remained a low 8:1.Implementation Progress The review team concluded that, in fact, the
and Difficulties target of 15:1 could not be reached under exist-ing teacher recruitrnent policies ancl that it
There vvas a time span of almost two years would be wrong to further reduce teacher train-between appraisal (October 1985) and imple- ing capacity, since this would be required in or-mentation (September 1987), suggesting that der to cater to the planned enrollment growththe project became effective nine months later in primary education once the benefits fromthan anticipated. While this delay resulted in a double-shift teaching and staff redeploymentcorresponding implementation delay for some had been reaped. The annual outpult of pri-policy measures, the introduction of others mary school teachers was down from a high ofstarted well before project implementation. The 1,379 in 1980-81 to 159 in 1988-89, whidc walstwo tranche release reviews related to the dass- less than the minimum of about 200 newroom construction program were held in June teachers needed annually just to compensate1988 (as scheduled) and in December 1989. For for attrition.reaso:ns explained below, the latter was six The limitation on teacher recruitment hasmonths ahead of schedule. turned out to be one of the key problems in the
The main condusions of the Bank team adjustment program supported through Educa-which conducted the June 1988 review were tion IV. The limitation derives from the overallthat: (1) although performance differed mark- constraint on civil service recruitment estab-edly among individual measures, on balance lished under Senegal's macro-economic adjust-implementation of the five measures designed ment program.to redLuce unit costs were ahead of target levels, Measure 2. Redeployment to teaching dutiesand (2) implementation of cost-containment of teachers engaged in administrative tasksmeasures for post-primary education, while had gone beyond the initial target A total oflagging, was generally in line with overall tar- about 1,270 persons had been redeployed asgets for the 1987-88 school year. Thus, the mis- against the target of 400.sion recommended that subject to reservations Measure 3. Introduction of multi-grade and(explained below) the second tranche for the double-shift teaching also was well ahead ofschool construction program be released. schedule. Double-shift teaching had been intro-
However, at the same time, the mission not- duced in 712 classes (82,927 pupils), and multi-ed several disturbing aspects in current trends grade teaching in 147 classes (7,984 pupils).and warned that, in the absence of corrective Net enrollment gains in 1987-88 due to thesemeasures, achievement of the main objective two measures has been estimated at someof the project, especially accelerated develop- 40,000 pupils. This corresponded to 7.2 percentment of primary education, would be threat- of the enrollment in public schools, and result-ened. Before discussing these warning signals, I ed in savings of about 580 teachers (5 percentbriefly review implementation performance for of the teaching force) and 786 classrooms.ecach policy adjustmnent measure. Measure 5. The study aimed at identifying
Measures 1 and 4. Progress had been limited ways of increasing pupil:teacher ratios in spe-in reducing the share of instituteurs in the cialized civil service training schools had notteaching force and in increasing the pupil: been launched, and only limited progress hadteachter ratio in primary teacher training col- been made in developing adjustanent measuresleges. This was largely due to the Government's in this sector (two institutions - the Centre definan,cial inability to recruit teachers which, in Formation et de PerfectionnementAdministratifturn, led to sharp reduction in admission into and the Ecole Nationale d'Administration et deteacher tlaining programs. New entrants are Magistrature - were merged in October 1986).pre-recriited as teachers upon entry into the Measures 6 and 7. One main problem in re-prograras; consequently, new admission de- viewing progress in attaining the budgetarypends on the number of vacant/new budgetary targets for post-primary education was that thepositions available. Although the consolidation monitoring system required was not yet fully inof training colleges had progressed as planned place at the time of the June 1988 review. The(two out of seven colleges were closed), the av- need to establish a reliable monitoring system
27
stems from the fact that actual expenditures for the University (these are nominal increas-may differ from budgeted expenditures. es). Taken together, the delay in launching the
Another problem specific to the Senegal case studies and the proposed budgetary increasesis that the adjustment targets are established in suggested future problems in meeting the ad-real terms. But no price index is available de- justment program's targets for containment ofsigned to reflect price increases in the education expenditures in higher education.sector, the consumer price index or the GDP de- Second, the main objective of the project is toflator is generally used. accelerate the development of primary educa-
Keeping in mind the foregoing, the review tion. As already explained, the number of pri-team concluded that performance against tar- mary school teachers to be trained andgets set for the reductions in student subsidies recruited would have to increase sharply inover the two-year period had been quite good. years to come in order for the enrollment ratioBased on a GDP deflator of 8.4 percent for to continue to increase towards the target. To1985-86 and 5.6 percent for 1986-87, the indi- achieve this within the overall ceiling of civilcators for higher education appear in Table 4.1 service recruitment included in Senegal's macro(at appraisal, US$1.00 = CFAF 360): adjustment program, a substantial reallocation
The smaller-than-projected decrease (in real of budgetary posts in favor of primary educa-terms) in the scholarship budget is a direct re- tion would be required, mainly from outsidesult of the widespread student unrest during the education sector.
Table 4.1Higher Education Spending Trends
1985-86 1986-87 1987-88 Percent real growth, 1985 thru 1987(in billion nominal CFAF) Target Actual
Scholarships 2,159 2,159 2,419 -8.0% -2.5%Campus services 988 988 988 -14.5% -14.5%Dakar University 3,671 NA 4,252 2.0% 1.3%
the school year. The 1987-88 budget foresaw a In view of these two issues, the review mis-freeze in nominal terms (a 14.5 percent de- sion recommended that the second phase ofcrease in real terms over the two-year period), the construction program be released followingbut CFAF 260 million were added during the a confirmation on the part of the Governmentyear to provide for scholarships for new stu- that it would: (1) recruit a sufficient number ofdents admitted to the university.7 students into teacher training colleges as of Oc-
Issues arising from the June 1988 review. tober 1988 to staff the schools to be construct-The main conclusions of the review mission ed; (2) within the limit of the ceiling on recruit-may be summarized as follows. First, despite ment of civil servants stipulated in its adjust-the modest performance on containing the ment program, transfer to the Ministry of Edu-scholarship budget in higher education, the cation a number of budgetary posts sufficientimpressive performance on the cost-saving to recruit these teachers; and (3) given the in-measures for primary education would satisfy creases in student subsidies proposed in theagreed-upon policy conditions for release of the draft 1988-89 budget, confirm its commitmentsecond tranche for the classroom construction to the adjustment program targets for contain-program. Little progress had been made in ment of expenditures in higher education up tolaunching the eight studies induded in the ad- 1994.justment program. The second tranche of the dassroom con-
Furthermore, continued student unrest had struction program was released in Octoberled the Government to propose substantial in- 1988, following the Government's confirma-creases for higher education in the 1988-89 tion of its agreement to the above three points.budget, a draft of which was fumished to the In addition, the Bank agreed to the Govern-review mission: 16.5 percent for scholarships, ment's request to advance the June 1990 re-35 percent for campus services, and 8 percent view by six months so that classrooms to be
28
constructed under that phase would be ready classes than in regular single-shift classes; (2)by October 1990. The Government hoped to teachers and parents generally have a negativeprovide sufficient classrooms at the start of opinion of double-shift classes, as t]hey believeschool year 1990-91 to enable the pupils that the reduction in curriculum hours will de-who staited in double-shift cLasses in 1986-87 crease achievement - a negative opinionto revert to single-shift cLasses upon reaching largely caused by lack of information on theGrade 5. probable impact of this approach on learning,
and particularly prevalent among parentsPolicy and Investment whose children were not enrolled in double-
Implementation through 1989 shift classes; (3) pupils in classes taught byinstituteurs-adjoints had, on average, higher
A Bank review team returned to Senegal in achievement levels than pupils in dassesDecember 1989 to assess progress toward fulfill- taught by instituteurs, whether in single-shift oring the policy conditions set for release of the double-shift classes - a difference not ex-thirdL tranche. Three main conclusions may be plainable from the survey data in terms ofdracum from its review. The first finding was achievement levels; and (4) the positive impactposilive, regarding ongoing progress in intro- of access to a textbook (reading or math) wasducing double-shift and multi-grade teaching. greater in double-shift than in single-shiftThe two other conclusions were cause for con- classes.8
cern, as they confirmed the difficulty of con- In view of this positive evaluation of the im-tainiing expenditures on student subsidies in pact of double-shift teaching on pupil achieve-higher education and the danger of stagnating ment, and given the shortage of classroomsenrollment growth in primary education. and teachers, the Government decided at the
Double-shift and multi-grade teaching. The end of 1989 to extend the use of this methodpositive development with respect to increased through Grade 6 for pupils already enrolled inuse of these two teaching techniques has con- double-shift classes.tinued. At the start of 1989-90, double-shift Spending on post-primary education. Theteaching was used in about 1,000 classrooms, figures in Table 4.2 summarize the develop-enrolling a total of 110,020 Grade 1-4 pupils. ment of actual expenditures during the periodMulti-grade teaching was used in 221 class- 1986-87 through 1988-89 for key indicators ofrooms, enrolling 11,647 pupils. This represents the adjustment program. As will be noted,a 33 percent increase in enrollment in double- some of the figures for higher education forshift classes over the two-year period, while en- 1986-87 differ markedly from the budget figures.rollment in multi-grade classes increased by 46 Note that growth rates refer to a two-year peri-percent. od and are expressed in real terms.
An evaluation of the experience with double- These figures clearly show that, instead of de-shift teaching was completed in June 1989. The creasing, student subsidies increased at a verystudy (based on a sample of 16 double-shift rapid rate. This tendency continued in the bud-and 16 single-shift classes) concluded that: (1) get voted for 1989-90, which includes CFAFpupil achievement was no lower in double-shift 3,456 million for scholarships and CFAF 2,343
Table 4.2Secondary and Higher Education Spending Trends
1986-87 1988-89 Percent real growth, 1986 thru 1988(billion nominal CFAF) Target Actual
Central adm. 1,764 1,817 2.0 -5.1General secondary 10,398 12,299 3.0 10.1Scholarships 2,145 3,036 -6.0 33.4Campus services
- budgeted 1,725 2,028 -16.0 95.1-actualexp. 2,162 3,737 -16.0 266.3
Dakar University 3,375 3,819 2.0 5.0
29
million for campus services (current prices). derway to prepare an extensive joint review be-Furthermore, preliminary results of an ongoing tween the Government and concerned donorsstudy on the functioning of these services show of policy and investment options. Studies fore-that actual spending in 1989-90 would be seen under Education IV's adjustment programabout CFAF 4,124 million. The study also will constitute one important input to this re-shows that, in early 1990, these services had ac- view. As of April 1990, the Government hadcumulated arrears arnounting to more than completed most of these studies. A second in-four billion CFAF. These were paid by the Min- put will be provided by an education sector re-istry of Finance in February 1990. From now on view prepared by France as the outgrowth of athe campus services will no longer be allowed donors' conference on education in Senegal asto accumulate arrears. Indications are that the planned in 1990. A third input will be providedbudget of Dakar University also has increased by several studies, especially on professionalfollowing the strike of the teaching staff in and technical education, conducted as a part1989. of the preparation of a proposed Fifth Educa-
The demand for higher education remains tion Project. An agreement on the future devel-strong in Senegal as evidenced by a sharp in- opment of the higher education sector iscrease in enrollment between 1988-89 and expected to play a crucial role in the Bank's de-1989-90 (from 14,833 to 18,800). There was cision as to whether or not to participate in thealso a marked increase in the number of stu- financing of such a project.dents who received scholarships (from 5,439 to Given the Government's overall ceiling on6,490). In response to this demand, the Govern- civil service recruitment, the main problemment has indicated its desire to open a second for primary education has been to ensure a real-university at a new, unused campus construct- location of budgetary posts in favor of primaryed for this purpose several years ago in the education. In its discussions with the Govern-North. ment, the Bank has called for the recruitment
Stagnation in primciry education. The grad- of a minimum of 600 new teachers each year,ual increase in the enrollment ratio for primary induding 200 to compensate for attrition.education (from 46.0 percent in 1980-81 to This minimum number has not yet been at-55.5 percent in 1985-86, and 56.6 percent in tained. However, within the framework of its1987-88) came to a halt in 1988-89 (56.4 per- Fourth Structural Adjustment Program (SALcent). The stagnation was caused by a sharp IV), supported by the Bank through a US$53decline in the number of new entrants to Grade million credit approved at the end of 1989,1: from 121,390 in 1987-88 to 104,709 in 1988- the Government has committed itself to89. This negative development took place hire a total of about 1,900 primary teachersmainly in rural areas where the enrollment ra- during the three school years 1989-90 throughtio declined slightly to 34 percent. 1991-92. Given that the total number of civil
The shortage of primary school teachers is a servants would be reduced over this period,principal cause of this stagnation in enroll- the importance accorded in SAL IV to thement The number of primary school teachers protection of primary education is noteworthydeclined by some 1.6 percent between 1987 since primary school teachers would constituteand 1989 and, while reliable data are not about half of all new hires during these threeavailable, many classrooms in rural areas years.(200-300 according to some sources) are empty Because of the problems discussed above,due to lack of teachers. The decrease in the the Bank decided not to agree to an early re-teaching force has, in turn, led to a decline in lease of the third tranche for the school con-the share of the education budget devoted to struction component. In addition to the factprimary education, from 47 percent in 1986-87 that policy conditions for such a release wereto 45 percent in 1988-89. not fulfilled, it would be of little use to construct
World Bank reactiors. In its policy dialogue classrooms that might not be staffed. And thewith the Government over the past two years, Government's decision to extend the use ofWorld Bank staff have consistently underscored double-shift teaching to all six grades of thethe disturbing trends summarized above. As re- primary cycle for those already enrolled in thisgards higher education, several activities are un- system reduced the need for new classrooms.
30
Lessons Leamed in Senegal rather than through enrollment expansion.However, the implementation of this strategy
Drawing on our experience in Senegal and is difficult for the reasons discussed earlier. Andother Bank-supported education policy opera- to be successful in a highly sensitive sector suchtions, this chapter concludes by reviewing fac- as higher education, any adjustment measuretors that shape the success of such programs. I would need to be based on a high degree of na-start by stressing a few key issues arising from tional consensus. This was not the case in Sene-the Senegal case. gal where the adjustment measures proposed
The most important lesson learned from this for higher education were hardly known to keycase involves the difficulties of budget adjust- players in the sector, let alone discussed andments in higher education. This is particularly agreed upon.acute in countries such as francophone Saheli- The second major lesson concerns the difficul-en countries where education systems generally ties in giving increased budgetary priority toare small, fairly costly, and essentially elitist, primary education in systems characterized bywith primary education catering to mainly ur- strong demand for higher education, unrestrict-ban areas and post-primary education benefit- ed entry for all high school graduates, and gen-ing the urban elite. The explicit objective of erous student subsidies. Education IVrecent Bank-supported projects in the Sahel, in- attempted to achieve this by establishing ceil-cluding Education IV in Senegal, has been to ings on growth in the budgets for post-primarydevelop more equitable and relevant systems education. The difficulties in implementing- where resource allocation is more efficient, such a strategy in the absence of a nationalwhere priority is given to quality improvements consensus, and a government that has suffi-at all levels and to expansion of primary edu- cient political will and strength to defend it,cation, and where the development of higher have been clearly demonstrated. 9education responds better to labor market de- A third set of lessons concerns certain designmands. features of the policy package. In general, the
Progress toward this objective has required package included measures that were both nec-cost-saving and resource reallocation measures essary and sound from a professional view-that have a clear negative impact on the edu- point. However, from the perspective ofcation benefits enjoyed by privileged groups. implementation, it would have been better notSince these groups are more vocal and influen- to fix the budgetary target in real terms and de-tial than those who would benefit from the re- vote more attention to the establishment of aallocations, governments often find it difficult monitoring mechanism for the adjustmentto implement these types of policy measures. measures. The problem of how to treat higher
While it is easy to agree that adjustment in education in a project principally dealinghigher education is a problem, it is more diffi- with primary education has already beencult to agree on what exactly must be done. In mentioned, as has the need for national con-cases such as Senegal, many Bank staff argue sensus-building in the development of policythat, given (1) low rates of primary enrollment packages.and adult literacy, (2) open unemployment of A fourth lesson is the need for flexibility onmost types of university graduates, (3) low the Bank's side when assessing whether or notqucility of higher education (caused partly by tranche release conditions have been met. Indistribution of limited resources over too many many cases, the conclusion is not obvious asstudents), and (4) severe budgetary constraints, progress on some conditions may have over-it is necessary to: (a) contain enrollment increases shot the targets in the policy matrix (e.g., intro-in higher education and reallocate new en- duction of double-shift teaching), whiletrants toward fields that offer better employ- progress towards other conditions may be mod-ment prospects, and (b) use any additional est at best. This requires a clear understandingresources available to the sector for quality im- of the relative importance of the various policyprovements. This strategy recognizes the need to conditions.strengthen higher education in these countries Finally, it should be stressed that most of thebut argues that, in the medium tern, this problems experienced in the Senegal case wereshould be done through quality improvements directly related to the political difficulty of in-
31
troducing adjustment in higher education dur- reforms, it is crucial to their success that they being a period of general fiscal austerity. The based on a high degree of consensus amongproblems were not caused by weak implemen- the population groups concerned, and thattation capacity in the traditional sense. As a their content is "correct" in terms of the desiredmatter of fact, the pro ject has benefitted from long-term development effect. While the latterthe support of a strong project coordination appears rather obvious, ensuring that reformsunit, led by a capable director. are professionally sound is not always easy
considering the time pressures, resource con-Improvinlg the Policy straints, and high degree of uncertainty oftenAdjustment Process surrounding the preparation of projects. Fur-
thermore, our knowledge as regards the mostThe starting point for this discussion is that cost-effective way of achieving key education
major reforms aimed at improving the efficien- objectives is less than perfect. To minimize thecy and effectiveness of resource use in the edu- risk of introducing reforms that later turn outcation sector generally have important both to be misguided, several points must be ad-short- and long-term effects on the lives of large dressed as serious policy dialogue begins andsegments of the population, as well as on the moves toward specific proposals:nation's prospects for social and economic de-velopment. Consequently, such reforms entail * Reforms must be based on thoroughGovernment decisions that generally are politi- sector knowledge. In many cases thiscally very sensitive. For example, many reforms means conducting a number ofimpact on the working conditions of teachers, country-specific studies prior to thewho frequently constitute the largest single design of the reform package to begroup within the civil service (sometimes 30-40 supported through a particular lend-percent) and form the strongest and most vocal ing operation. The amount and na-trade union. They also affect the lives of par- ture of sector work required dependsents and pupils. In the Senegal case, important on the type of policy measures con-segments of both teachers and parents resisted sidered. For example, while measuresthe introduction of double-shift teaching. to improve the external efficiency of
Another example derives from the fact that vocational training programs willreforms often implicitly or explicitly change the normally require thorough studiesdistribution of education costs and benefits covering both the supply and de-among different population groups. This is a mand sides of the market for the typevery important aspect since - as modem sec- of labor to be trained, the decisiontor employment is becoming increasingly whether or not to introduce double-scarce and dependent on education qualifica- shift teaching in a country such astions - the benefits derived from public spend- Senegal, where almost half theing on education are becoming an ever more school-age population is out ofimportant determinant of the distribution of in- school, would normally not requirefluence and wealth in the society. Consequent- extensive studies.ly, factors such as selection criteria for * Highly competent staff not only helpadmission and for awarding scholarships, real- to ensure development of sound re-location of resources in favor of primary educa- forms but also serve to establishtion, and location of education institutions credibility and trust between the twobecome major political choices which directly parties in the policy dialogue pro-affect the lives of various social groups. As a cess. It is difficult to overemphasizecorollary, it is important, prior to the introduc- the importance played by an opention of these types of policy change, to assess and sustained policy dialogue in thetheir likely impact on different population development of a well-designed poli-groups and, further, to ensure that this impact cy package. As indicated above,is properly monitored iduring implementation. many required adjustment measures
Because of the high political sensitivity and are painful. Planning these changesoften profound long-term effects of education requires both significant breadth and
32
depth of sector knowledge and the troduced and that, given presentcapacity to identify and explore poli- budgetary constraints and the urgentcy options in a collaborative fashion. need for quality improvements, canIn this process, negotiating skills are no longer be postponed. Such pastas important as technical skills. proposals constitute in many cases aAdjustment measures in the educa- good starting point for the develop-tion sector must be consistent with ment of an adjustment package.'Ooverall macroeconomic policy objec- * The policy package must be dis-tive.s. Sector staff must, therefore, cussed and explained to the mainwork closely with the country econo- parties involved. This is a critical as-mist to ensure such consistency. pect in assuring cacceptance of policyOnce the pre-project sector work is reforms. It is sometimes a delicatecormpleted, competent teams are in matter to discern the extent to whichplace on both sides, cnd initial policy staff from donor agencies should bedialogue established, work can begin involved in this effort. Some govern-on forging an agreement on a set of ments clearly do not want outsiderspolicy measures addressing the sector involved in discussions with tradle un-issues which have been identified. ions or students. What is important isSeveral key factors determine wheth- to ensure that the Govemment con-er a thoughtful and efficacious policy ducts such discussions and explainsdialo(ue can be sustained. They cen- the hows and whys of policy chang-ter on consensus-building. es. For example, in the Senegal case,
policy conditionalities on expendi-As wide a consensus as possible tures and scholarships in higher edu-should be built regarding policy alter- cation were never really explained tonatives. Essential elements in this students and staff at the university.process: * Donors' discussions of the policy
- Most sector work should be conduct- package with the Governmented by nationals, assisted by short- should not be limited to Ministry ofterm outside consultants if necessary. Education officials but, to the extentThere are a number of reasons why they touch upon cost, financing andttis approach should be the rule. For labor market issues, should includeexample, the discussions that take representatives from ministries con-place within the national adminis- cerned with these issues, as well astration during the preparation of pol- representatives of private employers.icy-related studies are often more One of the beneficial effects of cur-irnportant to the development of a rent budgetary exigencies has beennational consensus on policy chang- the realization by ministers of educa-es than the reports themselves. In tion of need for cost-effective use ofparticular, this process helps ensure the public resources under their do-Government ownership of the policy main.package and plays an important in-stituition-building role. Of the factors listed above, consensus-building
* Afost countries have, over the years, is the most important prerequisite for successfulconsidered the introduction of re- implementation of difficult policy reforms. Thisforms that for a variety of reasons is particularly true for policy reforms that re-have not been implemented. Many quire changes in central and local behavior, orof these represent policy changes that that influence the distribution of educationsooner or later would have to be in- costs and benefits among different groups.
33
NOTES:
1. The project was appraised in October 1985 and became the strike. At this time, disturbances spread to the Uni-effective in September 1987. versity, which the Govemment then closed. The stu-
2. If not otherwise indicated, all education statistics quot- dents reacted by joining what was becoming aed in this chapter are derived from World Bank (1986) nationwide youth revolt, which was later joined by theand World Bank (1988) as regards years prior to 1983 teaching staff. In April, there was a ministerial reshuf-and from UNESCO (1989), Ministere de l'Education Na- fling - the Ministry of Education was split - and thetionale (1988) and other national sources for later years. remainder of the academic year was used to negotiateData related to economic growth are from World Bank a solution to the crisis. However, schools remained(1989). closed until October 1988, meaning that all students
3. For a detailed description, see Ministere de l'Education lost one school year.Nationale (1986). 8. Cf. Ministere de I'Education Nationale (1989).
4. Cf. Ministere de l'Education Nationale (1986). 9. For example, in February 1990, Niger experienced seri-S. "Reallocation" is defined in relative terms in the sense ous student unrest caused by Government reforms de-
that, over time, the budgelt for primary education signed to give increased budgetary priority to primarywould grov faster than thset for otherimary lsof educatn education. The students' main complaint was that, inwould grow faster than those for other levels of educa- order to implement a gradual reallocation of educationtion, resulting in an increase in the share of the total sedn nfvro rmr dcto fo 5preducation budget allocated to primary education. cent of the education budget in 1989 to 54 percent in
6. These two grades are common for primary school 1995), the Govemment proposed inter alia to freeze theteachers in francophone African countries. The inshtu- budget for student subsidies at its 1989 nominal level,teur is the highest grade and normally has (as in Sene- and to abolish its policy of pre-recruitrnent to civil ser-gal) four years of training :n a teacher training college vice employment of all students benefitting from schol-following completion of lower secondary education arships. In 1988 student subsidies accounted for 64(Grade 10), or one year of t.eacher training following percent of the higher education budget, and about two-completion of upper secondary education. The institu- thirds of all university students received scholarshipsteur-adjoint has one (as in S;enegal) or two years of train- (the amount is about twice the legal minimum wageing following completion of lower secondary education. for unskilled workers) to help cover living expenses
(there is no tuition). While the outcome of the Govem-7. The 1987-88 strike of students in public secondary and ment's current negotiations with students on these
higher education in Senegcal began on October 15, points is not yet known, it is dear that the influence on1987, at the start of the new academic year when pu- Government policy of the 4,000 students at the Univer-pils at a secondary school in Thies demanded the rein- sity far exceeds that of Niger's 800,000 children of pri-statement of one of their colleagues who had been mary school age who are unable to attend primarysuspended for ransacking tfe home of the headmaster school due to a lack of facilities (only 29 percent of allseveral months earlier. When the authorities refused to children are enrolled in Niger).accede to this demand, the pupils boycotted their dass- 10. An example of a situation where considerable effortses. Other schools joined and before long the original were devoted to developing highly necessary reformsgrievance had been superseded by others related to the which had not been implemented, and which wereeducation conditions in the schools. By February 1988, picked up in a Bank-supported adjustment operation, is80 percent of secondary education students had joined given in the case of Ghana presented in this book.
34
Chapter 5
INDUCING AND MONITORING POLICY C]HANGE
Frances Kemmerer
his past decade has witnessed the de- native use of resources (Wise 1979). lEducation-J velopment and rapid sophistication al efficiency is defined, in turn, as the least-costof education management informa- means of producing gains in student learning
tion systems (EMIS) in developing countries. or the most productive use (in terms of studentPaper and pencil tallies of the numbers of stu- achievement) of a given allocation of resources.dents, teachers, and schools have been re-placed by computerized analyses of of greater Information and Institutionaldepth and breadth. The data sets now being Changeconstructed include information on the num-bers and location of complete and incomplete Information is the key to policy change. Thiscycle schools, the characteristics of teachers, proposition, which has served as the basis forschool-level data on the gender and age distri- heavy donor investmnents in EMIS, is generallybution of students by grade, and school-level interpreted as requiring quantitative data onachievement (where such data are available). system characteristics.
While much has been written about the de- There is little evidence, however, to suggest asign cnd development of EMIS (Chapman direct relationship between information on the1989, Chapman and Boothroyd 1988, Wind- numbers and characteristics of schools, teach-ham 1988, Snyder and Nagel 1988), and re- ers, and students and effective policy change insearch is emerging on the actual use of the either developed or developing nations. A di-data collected, less attention has been paid to rect relationship presupposes that the sole goalthe relationship between information and ef- of the educational system is to produce learn-fective policy change. ing. Nowhere does this seem to be true. As
In the interest of exploring the relationship Windham (1988) has pointed out, statementsbetween information and change, the paper of the goals of education tend to be weightedfirst presents three propositions governing this more heavily toward rhetoric than substance.relationship and then discusses the implica- Even as goal statements are overly inclusive intions of those propositions for donors attempt- terms of child development and child welfareing to induce, monitor, and evaluate change. objectives, they leave unstated the functionalEffective policy change is defined here as move- role of the educational system in the provisionment from a less efficient to a more efficient ed- of public sector jobs and on-the-job training.ucational system. Thus it purposefully excludes Yet in most developing countries, govemmentthe type of policy pronouncements, so popular is the chief employer in the monetary sector,in developed countries, which legislate goals and within govemment the ministry of educa-rather than facilitating or evaluating the alter- tion has by far the largest payroll. In Somalia,
35
for example, almost hcalf of all government doubt that the strategy will increase studentemployees work in education (GSDR 1984) and achievement but it will not do anything formany of the leadership positions in other sec- teachers."' In this case the construction of pre-tors are occupied by former teachers. stigious normal schools providing secondary
Acknowledging the fact that employment equivalency to teachers with only primary edu-and other political goals are rival to the pro- cation, and the employment of secondary leav-duction of student learning goes a long way to- ers (of whom there was a surplus), not onlywards explaining the absence of policies and competed with learning objectives but had apractices designed to deal with well-known sys- higher priority. Consequently, the rewards allo-tem deficits. Poor teacher attendance, non- cated to gatekeepers were based on their re-functioning materials cind supervisory support sponse to political priorities rather than tosystems, and the reluctance to adopt instruc- educational "policy" statements or systemtional technologies which obviate the need for needs.ever more highly "qualified" teachers are the The difficulty involved in improving educa-rule rather than the exception in much of the tional efficiency under the condition of rivaldeveloping world. Similarly, allocation patterns goals is compounded by the fact that politicalfavoring employment rather than productivity priorities affect what information is collected,and higher education as opposed to primary the accuracy and reliability of the information,education appear rational only if the goals of the sharing of information within and amongschooling are defined in terms of both educa- ministries, and the use of information. For in-tional and political objectives of the system. stance, in many developing countries, there are
The second proposition, then, is that the rela- far more teachers on the payroll than there aretionship between system irzformation and system actually teaching. And of those teaching, at-change is mediated by political objectives. This tendance at school is sporadic. Yet the phe-leads to the third propos.ition: namely, that any- nomenon of what the Haitians call "zombie"one desiring to induce change (from inside the sys- teachers is rarely discussed with outsiders, andtem or without) must have detailed knowledge not few hard numbers are available. Informationonly of system deficits but also of the political calcu- on the non-serving but salaried portion of thelus governing the current distribution of costs teacher force is not collected simply becauseand benefits, as well as individual perceptions the political environment is such that it cannotof those costs and benefits. be used.
In short, framing effective policy reform de- Similarly, collection of data on teacher atten-pends not only on infornation on educational dance was not solicited from school principalsinputs, processes, and ouitputs but also on un- because ministry personnel do not believe thederstanding which system objectives individu- data should or would be used. In Somalia, theals within the system ari trying to maximize argument of both planning department andand why. In the absence of this knowledge, school personnel against the collection ofeven success in inducing change is unlikely to teacher attendance data is that teacher salariesresult in success in effecting lasting policy are so far below the cost of living that it wouldchange, since the underlying incentives and be unfair to penalize teachers for high rates ofdisincentives for individual performance absenteeism. Interestingly, however, teachersthroughout the system will remain unchanged themselves do not hold this view. Teachers, pre-(Windham 1978). sumably those who attend regularly, resent the
In Somalia, for example, a crisis in the pri- fact that their salaries and benefits are themary sector precipitated by declining enroll- same as those of their less conscientious col-ments and severe fiscal constraints led to leagues (Said and Jama 1989).development of a donor-inspired strategy In ancient times, the bearer of bad news waswhich would have targeted all available Minis- usually killed. While this is no longer a com-try of Education resources directly on the mon practice, educational planners who con-schools. While the Ministry endorsed the strate- sistently report declining enrollments, highgy, it was in fact never implemented. When a dropout rates, and poor examination resultskey actor in the Ministry was questioned as to are not in an enviable political position in anywhy this was so, the response was, "I have no country. More often than not when these con-
36
ditions prevail, planners are not put to the test Ministry of Education and Culture estimatessince individuals at lower levels of the system were the product not of inaccurate reportingwill perceive an advantage in concealing infor- but of different divisions collecting the samemation which produces negative benefits. In data at different times in the school year andsome cases, there is a high cost attached to pro- estimating data from schools which had not re-viding any information at all. ported by the time each needed to complete its
Liberia provides a dramatic example of this. data setWorld Bank and USAID support in the 1970s To summarize to this point, where competingand early 1980s resulted in the development of goals exist, effective policy change requires in-sophisticated educational censuses and analy- formation not only on the status quo but alsoses supported by a mainframe computer in the on the benefit system supporting the statusMinistry of Education. By 1988, however, the quo. Unless the new policy succeeds in alteringEMI', was survived by a handful of statisticians the balance of incentives and disincentives inwith no data to analyze and a mainframe delivering educational services, the desired re-computer in disrepair. form of practice is unlikely to occur. To the ex-
The remote cause of the collapse of the EMIS tent that political goals dominate educationalwas the persistent ascendancy of the political goals of the system, however, much of the in-goals of the educational system over education- formation necessary to frame policy change isal goals. In lieu of a social welfare system, polit- apt to be unavailable or of unknown reliabili-ical efficiency was based on a patronage ty. In such situations, donors are faced with thesystem with the Ministry of Education as the difficult choice of minimizing project activity orchief employer. Within this context, indicators collecting original data and thus adding sub-of educational efficiency, if not irrelevant, were stantially to the cost of the intervention.of secondary importance.
The proximate cause of the collapse of the Inducing Policy ChangeEMIS, however, was a policy change which re-quired that pupil registration fees collected at Given explicit information about and recog-the school be sent to the Ministry of Education nition of the political context of education,together with enrollment data. The rationale there are a number of strategies for inducingfor the new policy was that the ministry was change. The most obvious approach is to re-able to capture economies of scale in bulk pur- duce the competition between political and ed-chasing (of school supplies and equipment) ucational objectives governing educationalwhich were not available to individual schools. systems by providing sufficient resources suchAs ihe ministry budget was reduced to only sal- that one set of goals need not be sacrificed toary expenditures over time, however, the min- realize the other set of goals. For instance, inistry itself needed the funds to maintain its the example cited earlier of Somalia, the priceactivities, and supplies were not sent to the of a concerted effort to improve what happensschools. Schools, district, and regional offices in schools might well be the building of one oralso needing funds, therefore, had a strong dis- more teacher training colleges.incentive for reporting accurate enrollments or In this case, however, accomodation of politi-reporting at all. As a result, it soon became im- cal and educational goals will promote educa-possible to establish the current status of the tional efficiency only if three conditions aresysteem in terms of numbers of students, student met. The first is that government has the ab-attrition and repetition rates, class size, or the sorptive capacity to assume the recurrent costsdistribution of teacher characteristics within or of these institutions. The second condition isacross counties. that teachers remain in the system long
Skepticism over the reliability of data based enough for government to recover its invest-on awareness of the disincentives faced by data ment in training. And the third condition isproviders is not reserved to donors. Chapman, that normal schools present the most cost-in a recent study of data quality and use in Ne- effective means of improving educational effi-pal (forthcoming), discovered that aggregate ciency. If the first two conditions were met, in-enrollment data were much more accurate formation needed to test the third conditionthan policymakers had expected. Conflicting would require in-depth study of the relation-
37
ship between teacher training (given a particu- ciency at the margin.lar curriculum) and teacher performance. The The USAID IEES Project experience in collabo-costs and benefits of tlle normal school ap- rating with governments on sector reviews (inproach would then have to be compared to the Botswana, Somalia, and Liberia) suggests thatcosts and benefits likely to be generated by oth- policy proposals that had been thwarted whener approaches (such as programmed teaching, routed through normal ministry channels wereinteractive radio, etc.), In the case of Somalia, facilitated when the debate was raised to ahowever, institution-building violates the first higher policy level.2 Frequently, would-be re-condition (absorptive capacity), and therefore formers in the ministry had adequate knowl-the strategy, while politically popular, would edge of system deficits but lacked the power tonot likely result in increased efficiency. change system priorities.
In general, however obvious this approach is In Botswana, for instance, the fact that theto the solution of the problem presented by ri- locally funded community junior secondaryval goals, the reasons for not taking it are com- schools were a source of both system inequitypelling. Investments in institution-building and inefficiency was not new infornation. Dis-leave a legacy of recuirent fund obligations cussion of the sector review findings in a multi-that most developing countries do not have the agency setting, however, led to the building ofcapacity to absorb. Liberia, for instance, has a a strong finance and education reform coali-number of institutions which act as a resource tion which was able to induce change. At thedrain on the entire system. The 1989 Education same time, the USAID commitment to assist inand Human Resource Sector Assessment found, funding the change provided the ministry withfor example, that $500,000 was spent annually an incentive to make change a priority. Inon two rural teacher training institutes with a short, there were no real losers. As a result, stu-combined enrollment of 29 students. Other in- dent entitlement to education through the jun-stitutions, including a college of technology ior secondary level acquired broad politicaland several secondary high schools presented support both in the ministry and throughoutsimilar pictures of cost inefficiency. government.
In practical terms, this suggests that it is vir- A second strategy which holds promise fortually impossible to accomodate rival political relatively small donors involves donor coordi-and educational goals and that reprioritization nation in the identification of system needsof the goals of education is necessary such that and remedies. Division of responsibility amongthe stated objectives (student learning) gain as- donors for supporting different facets or levelscendancy over the political objectives (employ- of an educational system is not likely to in-ment, national prestige as conferred by crease efficiency unless donors subscribe to auniversities, facilities, etc.). mutually agreed-upon strategy for quality en-
For this reason another and more preferable hancement For example, the cost-effectivenessapproach to inducing change is policy adjust- of the decision to use programmed teaching atment lending, as described in the other chap- the primary level rests on the fact that little orters of this book. Essentially, the policy no preservice training is necessary and that theadjustment approach requires that priorities for programmed teaching materials obviate theeducation be realigned, as a condition for lend- need for textbooks. However effective pro-ing. For contingency-based loans and grants to grammed teaching proves to be in improvingbe effective, however, the amount of support skills, its potential efficiency will remain unre-offered must be large enough so that refusal to alized if other donors continue to support tradi-meet the conditions carries a political liability, tional teacher training and textbookthereby making effective policy change politi- development and production. There is mount-cally as well as educationally efficacious. As a ing evidence, in fact, that competitive strategiesresult, it is unlikely that donors who are unwill- leave a legacy of confusion and waste of re-ing or unable to place a priority on invest- sources, while efforts at coordination help to setments in education can employ this strategy. the stage for system-wide reform.The question thus remains as to what ap- Other strategies which have a record ofproaches donors with rmore modest amounts to mixed success are marketing the desired policyinvest can take to influence educational effi- change as an "experiment" and thereby defer-
38
ring the political debate until the effectiveness Unintended effects of the policy change, stem-of the desired policy change has been estab- ming from incomplete knowledge of how deci-lished or using a bottom-up rather than a top- sionmakers at all levels of the system willdown approach. Success with the experimental respond to the change, must be addressed (e.g.approach, however, is likely to be realized only teachers do not understand the new curriculumif the donor is willing to subsidize the institu- so continue to teach the old curriculum). Andtionalization of the innovation after comple- finally, difficulties in the operation of the sup-lion of the experiment and has effectively port systems needed to enable implementationmarketed the innovation to relevant interest of the change must be explicitly recognizedgroups during the experimental phase. If nei- and dealt with (e.g., texts are not delivered, su-ther of these conditions apply, the donor is pervisors do not have the gasoline to visitforced to argue for implementation on the ba- schools, etc.).sis of technical rationality alone in an arena This suggests not only that policy implemen-where political rationality is dominant and tation increases the need for information on thewhere stakeholders' opinions of expected bene- operation of the educational system but alsofits, however correct or incorrect, have already that the policy, no matter how carefullybeen formed. thought out and designed, will have to be mod-
The benefits of the bottom-up strategy, which ified over time to suit local conditions if it is tohas proved so successful in the agricultural and be effective. In sum, all policy should be subject-heallh services sectors (Rogers 1983), may be ed to a formal formative evaluation process.limited due to the predominance in developing In Botswana, for example, the government/countries of highly centralized educational bu- community partnership in the provision of jun-reaucracies, the central provision of teachers ior secondary schooling has proved problemat-and other resources, and the fact that the bene- ic. Local ingenuity in exporting costs hasfits to communities and clients of improved ed- increased the burden to govemment beyonducational services may not be immediately an acceptable level (Swartland and Taylorobvious. The lessons to be learned from the suc- 1988). While government has not attempted tocess of microlevel strategies are, however, that systematically study the problems, micro-levelwhatever the strategy used to induce policy research indicates that some of the difficultychange, the policy package itself must remain stems from widespread misunder-standing offlexible enough to permit local adaptation, the the community's role, the existence of unin-effects of the change must be observable to the tended effects derived from changing the or-adapters, and the institutional structures neces- ganizational structures of already existingsary to support the policy change must be oper- schools, and definition of the school "commu-able (Foster 1974, Kemmerer and Wagner nity" as inclusive of more than one traditional1986, Rogers 1983, Thiagarajan 1985). community (Molutsi 1988). Since none of these
are irremediable problems, it would seem thatM:onitoring Policy Implementation micro-level research tightly focused on imple-
mentation issues would suggest appropriateThe process of monitoring and evaluating remedies.
change is fundamentally different from that of While the requirements of formative evalua-inducing change. The tasks related to inducing tion in terms of clinical and greenhouse testschange are largely political and focused on the are not politically or technically feasible withmacro policy level, while those related to im- regard to some types of policy, they are widelyplernentation and evaluation are primarily applicable to policy components which dictatetechnical and focused on the micro level. Suc- a change in the way resources are allocated atcessful implementation of policy depends, how- intermediate or local levels of the system. Theever, on intense communication among levels advantage of the so called "greenhouses" orof the system. staged implementation is that by reducing
Technical difficulties with the innovation it- scale, information can be obtained more rapid-self must be identified and solved (e.g., new ly from the field and necessary adjustments fedcurriculum is written for 180 days of schooling, forward into policy modification.while the average rural school is open 70 days). Where immediate full-scale implementation
39
is the only practical course, the process of ob- information needed to ensure effective policytaining the necessary information is both more implementation. The listing is clearly not ex-complicated and moire expensive. This suggests haustive, since different policy changes carythat when major policy change is undertaken, with them the need for different types of data.thought and resources must be given simulta- While system and school-level quantitativeneously to improvement of the existing EMIS data should be collected annually, qualitativesystem. Most EMIS systems do not track the ac- data can be collected less frequently after im-tual availability of inputs, with the exception plementation is well underway. A regular cydeof teacher characteristics, at the school level. for the collection of qualitative data is impor-Yet the collection of such data (see Figure 5.1) tant, however, since such data (together withrepresents little additional cost or expertise if inspectors' reports) provide a necessary checkadded to already existing school surveys. on the accuracy of the annual school survey
More costly and difficult to collect, of course, data.are process data - which answer the questionsof how the new resources are being used and if Conclusionsthey are being used inappropriately, why? Inthe initial stages of implementation, the an- Efficiency in the delivery of educational ser-swers to these questions are often provided by vices is based on the integration of instruction-the technical assistants hired to assist in moni- al and instructional-support subsystems.toring and evaluatin( implementation of the Consequently, it is almost impossible to im-new policy. Ideally, however, before implemen- prove one aspect of an educational systemtation is completed, the planning departments while ignoring others. Further, the complexityshould have the capacity to manage qualita- of even small systems indicates, first, that poli-tive as well as quantitative data, with one set of cy change is best considered as a long-termdata informing the accuracy of the other. modification process rather than as a short-
In contrast to the formative evaluation ap- term intervention; and, second, that informa-proach to policy change, the summative evalu- tion demands for both maintenance and modi-ation approach offers neither the information fication increase rather than decrease as thenecessary to modify policy nor confirmation subsector develops or efficiency increases. Thesethat the policy has in fact been implemented. realities underscore the suggestions made earli-Achievement data, even when available on the er that external interventions by donors arelevel of the school, are largely uninterpretable likely to prove effective only where there isunless disaggregated information is also avail- long-term donor involvement, where there areable on both inputs and processes and their ef- coordinated donor approaches to the subsectorfect on individual decisionmaking. While data problems, and where the necessary capacity-on student achievement have little utility by building is accomplished.themselves, they are cf central importance to Long-term Donor Involvement. Donors whothe validation of policy change, since in most remain involved in an educational subsectorcases student achievement is the object of such over the long term have a comparative advan-change. tage in both inducing and effecting policy
But valid data on achievement, as on educa- change since they have the information neces-tional processes, come at great cost. Even in sary to frame and market change. Moreover,those countries where national examinations long-term involvement makes it more credibleare given, there is little evidence that the curric- that the donor will assist the Ministry in ad-ulum is sufficiently developed to serve as a reli- dressing any serious unintended consequencesable basis for a national examination system, that may occur.that the test what is actually taught, or that Donor Coordination. The size of many edu-test taking and grading are accomplished un- cational subsectors suggests that no single do-der professional conditions. In many countries, nor can long remain the only benefactor fortherefore, the developmnent of the curriculum any given level. Sharing information and close-and tests are part of the price of effective inter- ly integrating efforts to improve educational in-vention in a subsector. puts and processes is therefore necessary if
By way of summary, Figure 5.1 outlines the system efficiency is to be maximized.
40
Figure S.1Data Needed to Monitor Policy Implementation
Quantitative System-Level Data
* Per pupil cost of resources* Output of teacher training institutions* Number of teachers at different levels of the salary and
allowance schedules* Birth rate by region
Quantitative School-Level Data
* The number of days school was open last year * Number of exercise books and pencils available* The average length of the school day last year * Number of visits by school inspectors* Enrollments by age, grade and gender * Number of students who dropped out of school* Teacher tumover rate during the course of the year by age, grade, and* Average teaching load per week gender* Average daily student attendance by gender and * Number of students promoted to the next grade
grade by age, grade, and gender* Average teacher attendance by gender and grade * Resources provided by the community in the past
level taught academic yeare Number of textbooks per class * National examination results by grade and sub-* Number of dassrooms with blackboards and ject, age, and gender
chalk
Qualitative School and Classroom-Level Data
* Amount of class time spent on instruction* Type and amount of homework assigned by grade* Average time spent using different instructional tech-
nologies in class1* Amount of teacher time spent in preparation for class* Substance and regularity of dassroom supervision by
headmasters and inspectors* Provisions made for remediation* Provisions made for acceleration* Provisions made for dass coverage during teacher
absences* Number of teachers with second jobs* Standard of living of headmaster and teachers* School/community relationships* Parental involvement* School climate (discipline, headmaster/teacher
relations, teacher/student relations, etc.
1lnstructional Technology is defined as a specific combination of teacher time, student time, and material resources.Note: See also Windham (1988) and Snyder and Nagel (1988).
41
Capacity Building. Strengthening the gov- Planners, no matter how much discretionary pow-emnments own ability to frame, monitor, and er they have, never determine the actual outcomeimplement change is essential if efficiency con- of policy. They can only set in motion forcessiderations are to be institutionalized. This re- which they anhicipate, with or without rationalquires development of capacity, within the ap- justification, will have certain effects. The effectspropriate ministry, for the routinized collection themselves are the result of the millions of micro-of quantitative data on the availability of in- decisions made by individuals responding to theputs and on outputs and the collection and planners'policies in terms of(a) the actual pat-analysis of qualitative data on system processes tem of rewards (positive or negative) which theirand longer-term outcoimes. decision matrix presents, and (b) their perception
This chapter closes where it began, with the of this pattem. The second condition is at least as1978 warning by Doucglas Windham on the important as the first because individuals' in-critical role of informalion on the distribution formation about incentives determines the degreeof individual costs and benefits in inducing, to which they will respond according to planners'implementing, monitoring and effecting expectations. (Windham 1978, p. 4.)change:
NOTES:
1. Conversation with a Ministry of Education official, mance of their educational systems and in strength-Mogadishu. ening their capacity for educational planning, man-
2. The IEES (Improving the Efficiency of Educational Sys- agement, and research. Participating countries have in-tems) Project is a ten-year initiative funded by USAID. cluded Botswana, Haiti, Indonesia, Liberia, Nepal,The objective of the project, now in its sixth year, is to Somalia, the Yemen Arab Republic, and Zimbabwe.assist developing countries in improving the perfor-
42
Chapter 6
LESSONS LEARNED?A Colorful Rainbow of Vi'ewpoints
Bruce Fuller
What Are We Leaning about temal forces is just as important as crafting theAdjusting Policy? internal elements of a policy program. The ad-
visability of key policy adjustments rnust be ar-Only recently has the first generation of poli- ticulated, and this involves understanding
cy adjustnment programs - in Ghana, Malawi, extemal economic and political winds. Third,and Senegal - approached their completion conceptually simple central policy changes aredates. Five years into these pioneering initia- easier to implement (holding political opposi-tives -- perhaps this is Lesson 1 - we are real- tion constant) than are policies requiring com-izing that it is difficult to assess local effects of plex alteration of local behavior. Attempts tocentral policy change when evaluation infor- legislate improvements in pupil repetitionmation is (and remains) scarce. Any lessons rates, for instance, have proven futile. Policyput forward must be considered quite tentative. and budget tools can be efficacious; but theyLong-term results are yet to be observed. [More must be applied realistically to effectively touchrecent policy programs, such as the joint World the behavior of local actors.Bank/USAID operation in Mali, are building in The complexity issue is particularly impor-more careful evaluation components.] tant in trying to improve institutional capaci-
We do know, as evidenced by chapters in this ties. For example, broadening the leadershipbook, that governments at times are able to role of headmasters, boosting local families'make central policy changes - some are even contribution to the village school, or improvingsustained: Malawi still maintains higher recur- textbook distribution can involve policyrent spending for textbooks; Ghana has adjust- change and manipulating signals sent fromed student subsidies toward greater equity; central government. But shifts in social rolesSenegal has rationalized its teacher allocation and gains in institutional capacities - aimedpractices. Difficult policy options have become at raising school quality and efficiency - alsoreal choices which are now largely institution- require complex, even subtle changes on thealized. ground. Policy levers provide a beginning
Implementation. A bundle of related lessons point, but they are insufficient in accomplish-emercge from these chapters related to how gov- ing sustainable organizational change locally.emments can increase the odds of successful Local Crafting of Policy Options. Finally,implementation. First, broad participation is es- what are we learning about how to improvesential, keeping in mind that interest groups governments' willingness and technical capaci-vary in the extent to which they want to see ty to generate policy alternatives? Our chaptersthe central state become more effective and le- from government leaders - Mrs. Yeboah andgitimate. Second, reading and moving with ex- Mr. Ngaye - reveal a certain enthusiasm, a
43
sense of challenge in trying to implement im- strong leaders can look realistically at itsportant policy and budget reforms. But policy problems and develop clear commitment, of re-recipes continue to be advanced largely by in- leasing energies of the people to address devel-ternational agencies. Too often, government opment and productivity. In Lesotho, we tooleaders are in the position of reacting to these are going through a process of identifyingoutside proposals. more realistic goals and programs."
At the same time, education and finance "We must face the political realities and con-ministries often articulate why certain policy straints. The minister of education must devel-and budget shifts are advisable - to conserve op a steel stomach to deal with protest andresources, to support those school inputs and resistance to change."human ingredients thcat effectively boost pupilachievement, and to equalize who pays and Tesfaye Dubalewho benefits from schcoling. A major chal- Ministry of Educationlenge is to raise governments' own ability to Ethiopiacraft their policy options and to assess likely ef-fects on teachers, families, and children. Even "Donors have focused on buying specific in-when policy adjustments yield benefits, if they structional materials or inputs. But these toolswere authored by outsiders, gains in internal should be framed in a broader policy context.political will and technical prowess - neces- The shift of resources toward primary educa-sary in continuing policy momentum - will be tion - assisting the masses - is a political im-slight perative that should be applauded within
initial sector adjustment efforts."A Variety of Voices: Comments and "We must focus on specific priorities - in
Conference Reports terms of how policy and budget improvementmatch targeted investments on school inputs.
The chapters in this book represent just a And we must think through how to encouragefraction of the many viewpoints and ideas ex- greater effort by local peoples, sorting out thepressed during the Washington conference on proper role of the education ministry and fami-Education Policy Adjusltment. Appearing lies."below is a sampling of the colorfully diverseopinions and insights into the process of adjust- Peter Moocking government policies and budgets. In addi- Africa Region, Technical Departmenttion, conference participants worked in small World Bankgroups, addressing a number of specific issues.Reports from each group also appear below. "Ghana's policy reform was first proposed in
1973 but it had drifted for almost 15 years.L. B. B.J. Machobani Then, a very courageous group in the govern-Minister of Educationi ment decided that action must be taken, other-Lesotho wise the educational system would continue to
decline. You must have this strong leadership"We must develop institutions - efforts within the country, otherwise little can be done
which are self-initiated - that are more rele- by outside donors or advisors."vant to the educational needs of our countries.A mood of introspection is moving across Afri- Gary Theisenca - growing from a feeling that false starts USAID Office of Educationwere made in the 1960s. But [today's] pro-grams and institutional, improvements must be "It is critical to sort out those adjustmentssustainable. We must resist prescriptions that that are devised with donors versus those policyare imposed from the outside. If governments and budget changes that unfold within govern-are able to work with donors, like the Bank, in ments, independent of action by internationaldeveloping policy options, this [cooperative] agencies. Also we have heard around the tablestrategy is more likely to work." the critical element of political leadership
"Ghana is an excellent example of how and the importance of constructing policy strat-
44
egies that flow from this commitment" alternative policy proposals. Broad discussion"We must put in place benchmarks and of sector problems early on - working with
ways of monitoring how policy change is various national actors - can help build atouching those who reside in classrooms: teach- consensus, reinforce the authority of the leader-ers and children. We know very little about ship, and minimize political risks. This careful,how adjustments in central policy actually in- deliberative strategy must be balanced againstfluence the motivation of teachers and the "windows of opportunity" that may arise.achievement of students." One example is how the tools of policy anal-
ysis might be used with teacher union leader-Wadi Haddad ship - particularly how alternativeSenior Education Advisor investments and options can lead to differentWorld Bank policy objectives. This should emphasize that
inaction will yield certain effects, such as con-"World Bank experience in lending for policy tinuing certain pattems of subsidy or the ero-
change is not limited to one mechanism but in- sion of educational quality in certaincludes sector adjustment loans and investment situations.loans that resemble more conventional pro- Donors can provide technical assistance tojects. Over the past five years, the Bank has education ministries in presenting their case tomade 91 loans in the education sector, just various stakeholders, from their own financethree have been pure sector adjustment loans. ministry to teacher and student organizations.Sector adjustment loans tend to focus on gener- What can governments and donor agenciesal budgetary and policy reform; more conven- do to help ensure sustainability of policy ad-tioncL investment loans address school quality justments once they are announced? An im-and efficiency in more specific ways. Some sub- portant first step is to establish concretesectors, such as vocational education, have not benchmarks to assess the implementation andyet been pulled into policy-intensive lending effects of policy change at the school and class-operations." room level,
Donors also can play a big role in helping toREPORTS OF WORKGROUPS finance the "transitional cost" associated with
implementing policy and budget changes. IfGROUP I - Political Constraints and subsidies are being shifted from higher educa-Consensus Building tion to primary education, donors' support is
needed to cushion short-term hardship linkedNoel McGinn, Harvard University to such structural changes.Rcapporteur
GROUP 2 - The Content of SectorOur group looked at several related topics: (1) Adjustment Programs
the economic costs of political mistakes, (2) in-country stakeholders who influence, and are Victor Bames, USAIDaffected by, policy change, (3) how to develop Rapporteura common framework and way of defining sec-tor problems for discussion among constituen- We agreed that sector analysis was an impor-cies, (4) how to construct a national consensus tant starting point But governments and do-ancl how donors can help, (5) how govern- nors also should carefully examine politicalments can move forward in the absence of a feasibility, key points of opportunity, and theconsensus, and (6) how policy or budget institutions (central and local) that are beingchange can be reinforced, made more sustain- called upon to implement sustainable, oftenable over time. controversial structural change.
We must recognize that sector adjustment Sector analysis needs to be dynamic, not stat-mcay incur political costs - in terms of teacher ic. As policy change is implemented, results (in-or student resistance which has been substan- tended and unintended) should be tracked.tial in a few cases. Thus we should focus on the Donor agencies, in general, are not set up to dostakeholders, who gains and who loses under long-term analysis. Governments also should
45
be assisted to examrine the social impact of ing or ways of informing people at the grass-sector adjustment efforts. Overall, governments roots about why adjustments or policy im-should gain greater control of the analytic provements are being undertaken, and (3)process. strategies for coordinating donor support, to
Sector analysis could better define a continu- ensure that costs associated with adjustmentum for policy change, for example, a set of are eased by external resources.strategies aimed at improving educationalquality, rather than proposing a set list of poli- GROUP 4 -Improving Information forcy or budget reformns. An array of increasingly Tracking Progress in the Education Sectorstrong measures might be explored within agiven policy objective such as reduced subsi- James Socknat, World Bankdies, improving the mix of instructional inputs, Rapporteuror raising capacity to monitor student achieve-ment First, we should sort out who is demanding
The group emphasized that the issue is not the information and what types of informationsector adjustment versus conventional project are useful to whom. This includes central gov-support. Policy adjustment in education often ernments, donors, local education offices, head-requires more traditional project inputs and masters, and communities. Second, wetechnical assistance. We also have seen that emphasized that information for monitoringpolicy change will not stick without deter- progress (say, with indicators of school quality)mined, long-term efforts at improving institu- differs from information required for evaluat-tional capacities. In addition, traditional ing adjustment programs or more convention-project lending often sets the stage for deeper al education projects.policy or fiscal change. Where USAID has sup- Information also varies in terms of types ofported management information development interventions undertaken by governments oror policy research, for instance, these activities donors. Tracking shifts in govemment or sectorhave prompted discussion of policy options budgets involves fairly straightforward data.within governments. Data on the distribution of textbooks or the
long-term effects of teacher training improve-Group 3 - Implemeriting Policy Adjustments ments become more complex and more costly
to collect.Julie Rea, USAID Information on actual pupil achievement isRapporteur most important. Then we should back up and
assess whether central policy change or the in-The group focused on concrete steps that fusion of alternative school inputs is correlated
might be taken to (1) construct, then reinforce, with actual pupil achievement Where infor-a broad in-countiy consensus for policy or bud- mation on this complete sequence (from cen-get change, and (2) how to generate alterna- tral action to school-level effects) is nottive policy changes, but then limit structural available, monitoring of selected policies or in-reformns actually undertaken - to increase the puts is a useful first step.odds of sustainable institutional change. We As we invest in education management in-generally talked about the conflict between formation systems (EMIS), we should ask, Whysome donors' push for short-term, measurable is this type of inforrnation useful? What does itchanges (such as specific budget changes) ver- tell us about policy action, input levels, andsus a nation's need to carefully build and im- school-level effects? Early MIS systems - nowprove capacities of institutions. in place in many education ministries - are
A government-set task force - with represen- embedded in old questions linked to howtation from many constituencies - might ini- quickly the education system is expanding. Sotiate the policy dialogue, conduct a careful enormous effort is spent on collecting data onsector assessment, and begin to develop policy numbers of schools, students, and teachers. Rel-and budget choices. This task force should look atively little information is available to assistat (1) issues of timing and when certain poli- governments in making decisions about im-cies should be implemented, (2) social market- provements in quality. Counts of basic inputs
46
(textbooks, teacher guides, exercise books) are data. This depends on demonstrating to head-rarely made, or are unreliable. Data on actual masters how the data can be useful to locallevels of pupil literacy are usually not available school staff and the central government. Evalu-and are rarely linked to variation in different ation is critically important to the educationinputs. Improvements in ministry EMIS systems sector in developing a self-critical process andshould address these issues regarding educa- in taking control over the generation of policytional quality. and fiscal alternatives. Building the capacity
Ideally, better monitoring systems and con- for evaluation and monitoring - conducted bycrete indicators (of school quality and efficien- governments and in-country researchers - iscy) can inform governments and donors on the an important long-term project that requiresimpacit of policy adjustments. Progress must be sustained support and patience on the part ofmade on increasing the reliability of school donors.
47
Reftrences
Chapman, D.W. (1989) The Role of Management Information Systems in Improving Educational Quality. Chapter 11 inChapman, D.W. and Carrier, C.A. (eds.) Improving Educational Quality: A Global Perspective. Westport, CT: GreenwoodPress.
Chapman, D.W. and Boothroyd, R.A. (1988) Threats to Data Quality in Developing Countries. Comparative EducationReview 32:416-429.
Chapman, D.W. and Dhungana, M. (forthcoming) Education Data Quality in Nepal. Program Evaluation and Planning.Albany: State Universily of New York.
Chapman, D.W. and Windham, D.M. (1987) The Evaluation of Efficiengy in Educational DevelopmentActivities. Tallahassee,FL: IEES Project.
Clark, P. et al. (1988) StrrcturalAdiustment in Lowinca. A Case Exercise in Economic PolicyAnal&sis. Washington D.C.: WorldBank.
Education and Employment Division (1991) Annual Operational Review: Fiscal 1990 Education and Training. WashingtonD.C.: World Bank.
Foster, P.J. (1974) Reflections on the Waigani Seminar. In Education in Melanesia. Australia: The Research School of Pacif-ic Studies, the Australia National University and the University of Papua New Guinea.
Fuller, B. (1987) What School Factors Raise Achievement in the Third World? Review of Educational Research, 57:255-292.
GOL and IEES Project (1988) Liberia Education and Human Resource SectorAssessment. Tallahassee, FL: IEES Project.GSDR and IEES Project (1 984) Somalia Education and Human Resource SectorAssessment. Tallahassee, FL: IEES Project.
Gulhati, R. (1988) The Poitical Economy of Reform in Sub-Saharan Africa. Washington D.C.: World Bank.
IEES Project (1986) The Feasibility of Integrating Programmed Learming with Conventional Instruction in Liberian PrimaryEducation. Tallahassee, FL: IEES Project.
Intemational Monetary Fund (1988) The Implications of Fund-SupportedAdjustmentPrograms on Poverty Washington D.C.Institute for Developmeni Studies (1989) Adiustina Education to Economic Crisis. Special issue of the IDS Bulletin. Sussex: Uni-
versity of Sussex. (This monograph includes case studies on Chile, Malawi, Nigeria, and Zambia.)
Israel, A. (1987) Institutional Development: Incentives for Performance. Baltimore: Johns Hopkins University Press.Kemmerer, F. and Wagner, A. (1986) The Limits of the Use of Educational Technology in Developing Countries. Paper pre-
pared for presentation at the Intemational Conference on Economics of Education: Tackling New Policy Issues, Dijon,France, June 23-25.
Messec, J.L. (1987) Issues and Opportunities for Energizing Educational Systems: Preliminary Review of EMIS-related Litera-ture. Tallahassee, FL: IEES Project.
Mingat, A. & Tan, J. (1988) Analvtic Tools for Sector Work in Education. Washington D.C.: World Bank.
Ministere de l'Education Nationale (1986) Etude de secteur educahon formation. Dakar.
Ministere de l'Education Nationale (1988) L'Ecole Nouvelle. Dakar.Ministere de l'Education Nationale (1989) Rapport d'Evaluation des classes a double flux. Dakar.Molutsi, P.P. (1988) Strengthening Local Educational Capacity through Community Involvement. Tallahassee, FL: IEES
Project.Nicholas, P. (1988) World Bank Lending forAdjustment: An Interim Report. Washington D.C.: World Bank.Policy Clarification Seminar on Priorities (1988) Special issue of Primary Education News. Govemment of Lesotho and
USAID.Rogers, E. M. (1983) Diflsiion of Innovations. 3rd Edition. New York: Free Press.Said, H.M. and Jama, Y.S M. (1989) Final Report of the Teacher Incentives Policy Research Initiative. Mogadishu, Somalia:
Ministry of Education.
Serageldin, I. (1989) Povertv. Adjustment, and Growth in Africa Washington D.C.: World Bank.
Swartland, J.R. and Taylor, D.C. (1988) Community Financing of Schools in Botswana. Chapter 12 in Bray, M. (ed.) Com-munity Financing in Education: Issues and Policy lmplications in Less Developed Countries. Oxford: Pergamon Press.
Thiagarajan, S. (1985) Factors That Contribute to Inefficiencies in Primary Schools in Developing Nations: A Survey of theLiterature. Project Report for USAID Honduras.
Thomas, IA. and Kemmerer, F. (1983) Money, Time, and Leaming. Final research report to the National Institute of Edu-cation Chicago, IL.: University of Chicago.
48
Tomatzky, L.G. et al. (1980) Innovahon and Social Process: A National Experiment in Implementina Social TechnoloaY Elmsford,NY: Pergamon Press, Inc.
UNESCO (1989) Statistical Yearbook 1989. Paris.
Verspoor, Adriaan (1991) Lending for Learnina - Twenhy Years of World Bank Support for Basic Educahon. Washington D.C.:World Bank.
White, L. (1988) Imnplementina Economic PoliLc Reforms: Policies and Opportunities for Donors. Washington D.C.: USAID Officeof Rurcal Development (processed).
Wilson, M. (1988) Current Status of SectorAdjustment and Investment Loans in the Education Sector. Washington, D.C.: WorldBank (processed).
Windharn, D.M. (1978) Incentive Analfsis and Hiaher Education Planning: Altematves in Theo,v. Research. andPolicy. ]Research Report. No. 29. Paris: Intemational Institute for Educational Planning, UNESCO.
Windham, D. (1988) Indicators of Educational Effectiveness and Efficiency. Washington, D.C.: USAID and Florida StateUniversity.,
Windharn, D.M. (1988) Indicators of Educational Effectiveness and EffidencV. Tallahassee, FL: IEES Project.Wise, A.E. (1979) LeWislated Leaming. Berkeley: University of Califomia Press.World Bcink (1986) Republic of SenegaL Cost and Finandng of Education. Westem Africa Project Department, Education
Division.
World Butnk (1988b) Education in Sub-Saharan Africa: Policies for Adiustment. Revitalizahton, and Expansion. Washington, D.C.World Bcnk (1988a) Financing Educahon in Developing Countries: An Exploration of Policv Operahtons. Washington, D.C.World Bank (1989) Sub-Saharan Africa - From Crisis to Sustainable Growth. Washington, D.C.Zuckerman, E. (1989) Adiustment Programs and Social Welfare. Washington D.C.: World Bank.
49
Distributors of World Bank PublicationsARGENTINA FINLAND mEXICO SOUTH AFRICA, OTSrWANA
C.to Htiech, SRL Akatenn.rKi.jkri,p NOTEC Ftorigk tMt
Galaia Gueom PO Box 128 Apsitdo Postal 22-60 Ofod Univeoity Pre
Floida 165,4thP-oor453/465 SFP6l0 14W60ThIpr.MeM D.F. ShuthernAhi
1333 8ruenoe Aires Heintkl 0 P.O. Bo 1141MOROCCO CapeTow-MM
AUSTRALIA. PAPUA NEW GUINEA. FRANCE Sodit d'Etrdes Marketing Macaine
FIlL SOLOMONISLANDS. Word BankPublications 12 rueMox-t Sd. d'Anf Foor4rti
VANUATU, AND WESTERN SAMOA 66 a-vesr d'na C_sabantc international Su.ooiptpin Svio
DA Se.&Jou=ab 75116 Pis P.O. Box 41095
64 Whitehoe Road NETHERLANDS CglrlU
Mitduitn 3132 GERMANY rnOr-Pbrkeat b- vjIohnug 2D24
V,idria UNO-Velag P.O. Box 14Poppel.dorfe All.eS 7240tBA lod- SPAIN
AUSTRIA D-53W Born 1 Mundi-Prerm Ubn SA.
Gerrd -d Co. NEW ZEALAND C_Ldlo37
Gnben 31 GREECE Hils Ubray and Infoation Sevice 25651 Mduld
A-1011 Wien KEME Pri-tSeBg24. Ippodamou Street Pl Pai Ne Maret Librfa Intenroional AEDOS
SAHRAIN Ather11635 Aadland CaeUi deCent 391
labrain Research -nd CoAtny CB0b9rl1celona
Asociates lid. GUATEMALA NIGERIA
P.O. Box 22103 Llbrenia Pied,. Sant. Univ-raity Pre Uinted SRI LANKA AND THE MAIDrVES
Maaes 7own 317 S.. Call. 7-SS ThreCr aBuildingleicho Lake Hous Bookohop
ZCaI I Private Mal Sag 50S9 P.O. Box 244
BANGLADESH Guatteanla City Ibadam 100, Slr Oditamplm A.
Mic rdndustries Devopmet GardineM-rM th
Aextnrr Sotiety (IIDAS) HONG KONGr MACAO NORWAY Clombo 2
Ho,ue S, Road 16 AA. 20 Ltd. Nre,et bIfoomation Center
haonmo.di R/A-e 4-4 Wyrdhm Stret Book Depantmet SWEDEN
Lka 1209 Witring Certe P.O. Box 6125 Ettenrd For kIa tilt.:
2nd Fwr N-06C2 O. 6 FPtzrs FP5bokf-tetst
BSered°. dE Central Hrg Kong Reerg tan 12IBox 16356
Main Rood OMAN 5-103 27 Stxihold n
Maq'deCor.t INDIA MEMRB nfomationServir-
Noakhali -356 Allied Pubhabs Pirvate Ltd P.O. Box 1613, Seeb Aireit For s ordos
751Mount Road Muscat Wwemmgen-WdihansA8
76 KDA. Av-e. Madas-600002 Ba3 04
K h. PAKISTAN S-104 25 Stdhotlr
Btaed qi= Mitrz Book Agency
BELGIUM ISJN. erediMarg 65, Shha-e-Qoaid-e-Aa SWITZERLAND
Jea nDel lnoy Ballad Estate P.O Bo No. 729 FergxStit
Av. du Rot 202 bombby-400 038 Lahore3 Lirarie Payot
1060 B tru..o 6, rue Greeut
13/14 A. Ali Road PERU Cse pobtale 381
CANADA NewDe.hi-11I002 Editoial Deraeolo SA CHI 1211 Ge-a 11
LeDiKf r Apatado 3524
CP. 8,1901rne Ampinfe 17 Chittarj Avenue Lin. For-.Ipto rdar
oud1evi1e. QudFec Cakutta -70D072 Ubrarie Payot
J45 SE6 PHILIPPINES Sevkie d Abonneent.
Jiy.dn a14t Bduilding Internral Book Cente, Casepoaae3312
CIHINA Sth Main Road GandbiragPr P.iB Flor, Thlipina. Ufe Building CH 1002 Losomnne
China cil & Erooric Bmgal ore- 56009 Ayal Ave-., Mkakt
Publishing Howe MetroMamila TANZANIA
d,D. oSiDrrg]te 3-5-1129 Kochigda Cr. Rood Of-d Univreity Press
belpng Hyde-bad- 50027 POLAND PO.Box..99ORPAN DaresSalaam
COLOMBIA Pnrthas RtLa. 2nd Floor Pac K.luroy i NarridInfo_c_ Lud. Near Thakore B Nvrongp-ra 0401- Wae.a THAILAND
Aptado Ar 34270 Amadred -380059 Centml Deprtorent Store
Bootk DE. PORTUGAL 306SDom Road
Palala Houe Uvrai Potugal Bangkok
COTE DIVOIRE 16-AbA hokMorg Ru Do Cnrmon70-74
Centre d'Edition et deDiffion Ludar 226001 126 ULsbon TRINIDAD & TOBAGO. ANTIGUA
Aircatne' (CEDA) BARBUDA,BARBADOS,
04 B.P. 541 INDONESIA SAUDI ARABIA, QATAR DOMINICA, GRENADA, GUYANA.
Abidjsn 04 Pltau Pt Indir Limited air Boak Stre JAMAICA, MONTSERRAT, ST.
1. Sam Rhtroingl 37 PD. Box 3196 KITTS & NEVIS, ST. LUCIA
CYPRUS P.O. Box 181 Riyadh 11471 ST. VINCENT A GRENADINES
MEMRB Eforttion Seaites J3kta Prat SyAteatics Sudie Unit
P. Box. 1E= MEMRB InfrtionSevi 099Watt.Street
Nicosa ITALY Borch q7i8a Curepe
ULi. Commtaiionvi S anoi SPA Al Ala. Sbe Trinidd, Wet Ilndies
DENMARK Via Benedetto Fordni, 12D/10 Al Dahne Center
Sanfumd. ittot-r Casella Poble 552 Fti FlRo. UNITED ARAB EMIRATES
Rrr.a-oenonAAl#II 5D125 Floretre P.O. Box 7188 MEMRB Gulf Co.
DK-1970 Frdeikol,r C Riyadh PO Box 6097JAPAN Sh.*ah
DOMINICANREPUBLIC EasteBookS-ice H jiAbdulla Ali. Buidindg
Editor. Taller, C pr A. 37-3 Hongo 3-Chmre EBunkyo,ku 113 King KhIaled Street UNITED KINGDOM
Res.Lracir a Ioabetla Ca6dlica 39 Tokyo P.O. Box369 Mirao&nfoLtd.
Aprtado Pora 2190 Dmnman P.O. Bo 3
SantoDonitg. KENYA Alto, Hampohire GU34 2PG
Africa Book Savior (EA.) Lid. 33, Moranuned Haan Awrd Stret Eneand
EL SALVADOR P.O. Bo 45245 P.O. Box 5978
Fuades Nairobi jeddah VENEZUELA
Aveidan Msr.u EtiqueAr-upj o3531 hieb e del Eate
EdiSfioSISA,ler. Pio. KOREA. REPUBUC OF SINGAPORE TAIWAN, Aptdo. 60337
Sa SmSlvad Pn Ko BookCorporai MYANMARBRUNEI Carcas 1060-A
P.O. Box 101, K ngwrmun hrfr,tion PublicatioEGYPT, ARAB REPUBLIC OF Sorl Privae. Ltd. YUGOSLAVIA
Al Alrhnm 02-C61 t PL. Pei-F. Industrial ]ugdovaera Knjinga
Al Gala. Stret KUWAIT Bldg. P.O. Box 36
Cairo MEMRB Information SevIces 24 New Indr&iril Rood Trg Republike
P.O. Box 5465 Swngapore19S3 YU-1100OBeigrade
TIreMiddle Ba tObr41, Sherf Stret MALAYSIACairo Univdrsty of Malaya Cooperatve
Bookahop, LiortdP.O Box 1127. Jab.a P-mt i BruKuala umnpw
Recent World Bank Discussion Papers (continued)
No. 103 Agricultural Extensionfor Women Farmers in Africa. Katrine Saito and C.Jean Weidemann
No. 104 Enterprise Reform and Pnvitization in Socialist Economies. Barbara Lee and John Nellis
No. 105 Redefining the Role of Govemment in Agriculturefor the 1990s. Odin Knudsen, John Nash, and others
No. 106 Social Spending in Latin America: The Story of the 1980s. Margaret E. Grosh
No. 107 Kenya at the Demographic Turning Point? Hypotheses and a Proposed Research Agenda. Allen C. Kelleyand Charles E. Nobbe
No. 108 Debt Management Systems. Debt and International Finance Division
No. 109 Indian Women: Their Health and Economic Productivity. Meera Chatterjee
No. 110 Social Security in Latin America: Issues and Optionsfor the World Bank. William McGreevey
No. 111 Household Consequences of High Fertility in Pakistan. Susan Cochrane, Valerie Kozel, and Harold Alderman
No. 112 Strengthening Protection of Intellectual Property in Developing Countries: A Survey of the Literature. Wolfgang Siebeck,editor, with Robert E. Evenson, William Lesser, and Carlos A. Primo Braga
No. 113 World Bank Lendingfor Small and Medium Enterprises. Leila Webster
No. 114 Using Knowledgefrom Social Science in Development Projects. Michael M. Cernea
No. 115 Designing Major Policy Reform: Lessonsfrom the Transport Sector. Ian G. Heggie
No. 116 Women's Work, Education, and Family Welfare in Peru. Barbara K. Herz and Shahidur R. Khandker, editors
No. 117 Developing Financial Institutionsfor the Poor and Reducing Bamers to Accessfor Women. Sharon L. Holtand Helena Ribe
No. 118 Improving the Performance of Soviet Enterprises. John Nellis
No. 119 Public Enterprise Reform: Lessonsfrom the Past and Issuesfor the Future. Ahmed Galal
No. 120 The Information Technology Revolution and Economic Development. Nagy K. Hanna
No. 121 Promoting Rural Cooperatives in Developing Countries: The Case of Sub-Saharan Africa. Avishay Braverman, J. LuisGuasch, Monika Huppi, and Lorenz Pohlmeier
No. 122 Pe!formance Evaluationfor Public Enterprises. Leroy P. Jones
No. 123 Urban Housing Reform in China: An Economic Analysis. George S. Tolley
No. 124 The New Fiscal Federalism in Brazil. Anwar Shah
No. 125 Housing Reform in Socialist Economies. Bertrand Renaud
No. 126 Agricultural Technology in Sub-Saharan Africa: A Workshop on Research Issues. Suzanne Gnaegy andJock R.Anderson, editors
No. 127 Using Indigenous Knowledge in Agricultural Development. D. Michael Warren
No. 128 Research on Irrigation and Drainage Technologies: Fifteen Years of World Bank Experience. Raed Safadi andHerve Plusquellec
No. 129 Rent Control in Developing Countries. Stephen Malpezzi and Gwendolyn Ball
No. 130 Patterns of Direct Foreign Investment in China. Zafar Shah Khan
No. 131 A New View of Economic Growth: Four Lectures. Maurice FG. Scott
The World Bank
Headquarters European Office Tokyo Office1818 H Street, N-W. 66, avenue d'lena Kokusai Building ____._l_
Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chomeChiyoda-ku, Tokyo 100,Japan o
Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00Facsimnile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 tTelex: wui 64145 WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657
RCA 248423 WORLDBK Telex: 26838Cable Address: INTBAFRAD
WASHINGTONDC
00
IZ
C
0
10
0
ISBN 0-8213-1932-9~~~~~~t/