Education Reform Support Today - epdc.org Reform... · Many of the examples we could not have compiled without the help of ... existing data to support debate or gather ... Education

Education Reform Support Today

Joseph DeStefano and Luis Crouch

SEPTEMBER 2006

The original Education Reform Support series is available for download:

Volume One: Overview and Bibliography

Luis Crouch and F. Henry Healey

http://pdf.usaid.gov/pdf_docs/PNACA717.pdf

Volume Two: Foundations of Education Reform Suppor t

Luis Crouch, F. Henry Healey, and Joseph DeStefano


Volume Three: A Framework for Making It Happen

F. Henry Healey and Joseph DeStefano


Volume Four: Tools and Techniques

Luis Crouch, Beverly Schwar tz, F. Henry Healey, and Joseph DeStefano


Volume Five: Strategy Development and Project Design

Luis Crouch and Joseph DeStefano


Volume Six: Evaluating Education Reform Suppor t



ACKNOWLEDGEMENTS

This paper could not have been drafted without the collaboration and input of several colleagues. Jonathan Berr y of the Center for Collaboration and the Future of Schooling

provided research suppor t and completed an extensive literature review that allowed us toupdate many of the references in the Education Reform Suppor t series. Boon-Ling Yeo ofthe Research Triangle Institute also helped with research, par ticularly having to do with

the case examples. Many of the examples we could not have compiled without the help ofcolleagues whose experiences in the field shed invaluable light on the practical applications

of ERS. We would like to thank Frank H. Healey, Donna Kay LeCzel, Jef f Puryear, AlastairRodd, Audrey-marie Schuh Moore, and Carmen Siri. In addition, John Gillies responded toour initial draft and helped shape our discussion of the evolution of funding modalities.

Finally, David Plank, Patrick Collins, Ash Har twell, and Audrey-marie Schuh Moore providedvaluable feedback and input that helped improve early drafts of the paper.

Education ReformSupport Today


SEPTEMBER 2006

TABLE OF CONTENTS

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

The Case for Reform Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Education Reform Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

How Education Reform Support Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

The Tools of Education Reform Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

What ERS Projects Can Do. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2. Dealing with Scale and Sustainability: Moving Beyond the Technical Challenges of Education Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Scale and Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Literature Confirms Requirements of Reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Changes in the Donor Environment Make Reform Support All the More Important. . . . . . . . . . . . . . . . . . 9

The Tools and Techniques Discussed in ERS Are Still Valid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

New Focus on the Institutional Environment for Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Implications for Donor Projects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3. Clearing Space, Filling Space, Building Reform Support Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . 15

Clearing Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Filling Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Ensuring schools have the resources and authority they need . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

School-community collaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Public-private partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Accountability for results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Timeliness and effectiveness of school and teacher support. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Filling Space and the Conditions for Success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Reform Support Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4. Education Reform Support Activities and Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Dialogue and Communication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Building Reform Support Infrastructure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Projectizing Data, Analysis, Dialogue, and Support to a Reform Support Infrastructure . . . . . . . . . . . . . 36

5. Limitations (and Advantages) of the ERS Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

6. Justifying an ERS Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

7. Jump-Starting Education Reform Support (What Can Be Done Now?) . . . . . . . . . . . . . . . . . . . . . . . . . . 45

8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Annex A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Education

Reform

Support

Today

1Executive SummaryTHE CASE FOR REFORM SUPPORT

Traditional projects in education introduce inno-vations at the school level, sometimes improv-ing learning in a defined number of schools.The hope is that somehow piloted successescan be replicated or taken to scale. But toooften they are not. Dissatisfied with this,donors may choose policy-level interventionsthat promote resource reallocations, specificpolicy reforms, and investments in administra-tive and management capacity to effect sys-tem-wide change. But the record of policyreforms having impact on learning at the schoollevel is disappointing. If we fund school-levelprojects, the challenge lies in how to createpolicy and institutional reforms that supportreplicable school-level success. If we supportpolicy-level interventions, the challenge lies inhow to ensure that national reforms lead tochanges in the day-to-day practice of schools.Both approaches require effective programs ofwhat we call reform support.

EDUCATION REFORM SUPPORT

Why is reform support needed? Ten years agoUSAID published the Education Reform Support(ERS) series to answer just this question. ERSrecognizes that the existing arrangements inthe education sector—urban-rural inequities,management environments skewed by bureau-cratic concerns, teaching improvements con-

strained by union pre-rogatives—are not acci-dental. Powerful politicalforces benefit from,shape, and defend thecurrent situation.Changes within the sys-tem cannot realisticallybe implemented without

first dealing with the preexisting institutionalenvironment. Altering that environment meansrecognizing who stands to win or lose from pro-posed reforms, and what incentives signalthem to either work for change or defend the

status quo. The literature supporting such anunderstanding of education reform is rich. ERSdraws on that literature and goes one step fur-ther to outline the tools and techniques for sup-porting and strategically managing the reformprocess.

HOW EDUCATION REFORM SUPPORT WORKS

ERS is organized around three notions:

• Clearing space: making room in the existingpolitical economic landscape

• Filling space: introducing the substance ofsound educational improvement

• Reform support infrastructure: the networkof institutions and actors that clear and fill space

Clearing Space. A major problem in carryingout a reform agenda is the existence of precon-ceived ideas, mental models, interest groups,and institutional traditions. “Clearing space”refers to the need to reduce the intellectualand political space taken up by existing viewsand interests. It involves bringing hard evidenceand countervailing political power to bear, andit implies isolating narrow interests or traditionalbeliefs both intellectually and politically anddemonstrating the negative effect of theseinterests on the general good.

Filling Space. If space is cleared in the educa-tion landscape, there is no lack of “solutions”waiting to fill it back up—often those solutionsproffered by the interests that needed to becleared out in the first place. Space-filling activ-ities should be judged primarily by whether theyare demonstrably effective at improving educa-tion outcomes. And filling space requires notjust introducing the innovations, but also struc-turing them so that their impact can be evaluat-ed, the conditions required for successfulimplementation can be documented, and obsta-cles that may need to be “cleared” for thoseinnovations to be sustained can be identified.

“ERS…goes one step

further to outline the

tools and techniques

for supporting and

strategically managing

the reform process.”

Edu

cation

Ref

orm

Sup

port

Toda

y

2Reform Support Infrastructure. The originalERS series coined the term, “reform supportinfrastructure,” to describe the network oforganizations and individuals that help bringabout successful reforms. This infrastructure is needed because donors, by themselves, cannot do reformist work effectively. The workof a reform support infrastructure includes generating demand for specific reforms, creatingopportunities for and facilitating dialogueamong stakeholders, organizing and mobilizingconstituencies, engaging in strategic communi-cation, and presenting data and analysis incompelling formats. No one organization possesses all of these capabilities. Therefore,a viable reform support infrastructure dependson collaboration and coordination among many actors.

THE TOOLS OF EDUCATION REFORM SUPPORT

ERS is implementable as a set of discreteactivities provided they are chosen and imple-mented strategically. The tools of ERS focus ongeneration and use of data, analysis, dialogue,and communication. ERS also relies on buildingreform support infrastructure.

Data. Empirical data have to underpin analysisand promotion of reform, so donors shouldsupport data efforts. But promoting reformdoes not require that donors spend much energy on creating and maintaining traditional,supply-driven education management informationsystems (EMIS) or general-purpose surveys.The data activities recommended for ERS useexisting data to support debate or gather original data for a specific policy discussionpurpose.

Analysis. Donors can support analysis, but theanalysis ideally must be driven by debate aboutissues that require reform. Analysis of schoolfinance, for example, should be driven by sharppolicy concern over funding equity. Also, policydeliberations are often hampered by the lack ofrigorous analysis of existing educational innova-tions. ERS supports systematic experimenta-tion and evaluation that can leverage broaderimplementation and impact.

Dialogue and Communication. Perhaps the most important element of the ERS approachis fostering communication, debate, and dia-logue. Communication and dialogue should beintegral to data and analysis activities. In fact,without communication and dialogue, demandfor analysis and data is weak. The ERS positionis that sustainable, implementable solutionsemerge from a dialogue-driven process of competition and searching.

Building Reform Support Infrastructure. All of the activities discussed above can be implemented to support a given package ofreforms. Merely funding such activity by itselfdoes not institutionalize ongoing capacity todebate and design reform. The goal of donorassistance should be to help create the infrastructure whereby data-based analysis, policy dialogue, and strategic communicationcan be ongoing and adaptive. This requiresbuilding up the capacity of a variety of institu-tions and individuals and, in particular, invest-ing in helping them network and collaboratewith each other.

The matrix on the following page summarizeshow ERS aligns data, analysis, and communication with the challenges of clearing space, filling space, and buildingreformsupport infrastructure.

WHAT ERS PROJECTS CAN DO

Ideally, we advocate that donors implement allof the above by designing coherent sets ofactivities (projects) whose size and duration aresufficient to have a profound impact. However,as a practical matter, smaller interventions canbe carried out quickly, within the confines of anexisting project. For example:

• A project can inject fresh data on schoolingquality into policy dialogue sessions that fos-ter awareness of the poor quality of school-ing. Dramatic comparisons and illustrationscan be useful, for example, in regard to chil-dren’s reading ability. Such comparisons canbe used to focus attention on resourceinequality or poor and unequal results.

Education

Reform

Support

Today

3Clearing Space Filling Space Reform Support Infrastructure

Data• Using data to redefine

issues or to refute conventional wisdom

• Demonstrating the validity of particularinterventions

• Investing in buildingcapacity of local institu-tions (inside and outsidegovernment) to do data,analysis, dialogue, andcommunication

• Investing in networkingacross institutions withdifferent capacities andpromoting the emer-gence of common agen-das and an articulatedtheory of change

Analysis

• Demonstrating the limitations orinequities of existingarrangements

• Cost-benefit comparisons• Comparisons of

trade-offs (political and in terms of management difficulty and potential impact)

Dialogue &Communication

• Giving voice to divergent points of view

• Creating opportunities for genuine deliberationand debate

• Involving stakeholders indialogue around how innovations can be implemented

• Drawing out broader institutional implicationsfrom pilot successes

Networking &Reform SupportInfrastructure

• Promoting and supporting strategic alliances• Informing and mobilizing specific constituencies• Building strategic management capacity

• At certain points in the policy process, thedebate will peak. Being ready with a profi-cient analysis and position paper in the mid-dle of a peak moment of policy decisionmaking is a good way to have greater impactthan the cost of the activity would indicate,but it requires paying constant attention tothe debate and being ready with useful inter-ventions at a moment’s notice.

• An existing school-level project can also pur-sue data, analysis, and communication toaddress some of the policy issues garneredfrom its on-the-ground experience. Data canbe gathered that show how the conditionsfor success need to be changed at theschool level to maximize the impact of, say,teacher training. Communication and dia-logue can help clear space, mobilize sup-porters, and advocate for establishing thoseconditions on a broader scale.

ERS activities provide high-leverage possibilitiesfor affecting the direction of national reformefforts, and they potentially can increase thescale at which school-level projects can haveimpact. Strategically supporting use of dataand analysis, targeted communication, and dia-logue and networking among reform-mindedinstitutions can advance larger-scale, more sus-tainable reforms in education system policiesand practices.

Education

Reform

Support

Today

51.IntroductionWith 115 million 6- to 12-year-olds still out ofschool, the challenge of improving access tobasic education remains critical. Furthermore, in sub-Saharan Africa, as many as one in threestudents fail to complete even four years of primary school (UNESCO Institute for Statistics2006). With such high dropout levels, and littleif any evidence of learning, improving the effectiveness of schools supersedes the need toexpand access if what we are after is truly edu-cation for all. What is the point of expending everincreasing resources to get more children intoschools that fail to provide adequate education?

Since the early 1960s, the world has witnessedrepeated commitments to ensuring education forall children and recurring calls to mobilize thefunds needed to make it happen. Each time thetarget for achieving universal primary educationis missed, the date is pushed back anotherdecade or so, and a lack of resources andpolitical will is blamed for the inadequateprogress (Clemens, 2004).

Despite this seemingly endless cycle of calls toaction, missed targets, and recalculated fundinggaps, over the years, many millions have beenmobilized to support each push to achieve education for all. By some estimates, low- andmiddle-income countries spend about US$70billion a year on education (Abadzi, Crouch,Echegaray, Pasco, & Sampe, 2005). Mostrecently, the Education for All (EFA) Fast TrackInitiative (FTI) was created to mobilize additionalmultilateral and bilateral resources to supportcountries’ plans for achieving universal accessby 2015. For the 20 countries that have qualifiedthus far for Fast Track support, US$600 millionhas been pledged on top of regular, ongoing aidtransfers for education. But, even this heroic effortto support EFA falls short by US $510 millionthis year alone, based on the FTI’s own estimates.Another 40 countries are slotted to access theFTI by 2008, pushing the estimated additionalfunding gap to US $3 billion a year for the nextdecade or so (FTI Progress Report, 2006).

Even under the best of circumstances, it is hardto imagine where this additional funding willcome from. As long as we define the problemas mobilizing sufficient amounts of additionalresources, we will continue to be disappointedwhen large increases in foreign assistance failto add up to what is needed. To talk realisticallyabout improvements that can dramaticallyincrease the effectiveness of schools in mostdeveloping countries (and in many developedcountries), we must talk about transforming theway education systems are organized and managed, not just whatadditional resources must be “…we must talk

transforming the

education syste

are organized a

managed, not ju

additional resou

must be spent.”

about

way

ms

nd

st what

rces

spent. Tangible interventionsthat can increase educationaleffectiveness are availableand needed (e.g., a focus onmethodologies for teachingreading in early primarygrades), and resources canand should be directed topromoting them. However, such innovations, ifthey are to operate at scale and be sustained,necessitate a host of accompanying changes ininstitutional incentives (Pritchett, 2004).Therefore, reform, whereby the policies, prac-tices, and organization of education are altered,needs to remain on the front burner.

Like the recurring calls for achieving educationfor all, the drumbeat of reform has been soundedrepeatedly. Countless programs and projects,citing reform as an objective, put resources onthe table to support policy development, analysis,and institutional capacity building. Projects havesupported reorganization of ministries, trainingof education officials, and massive purchasesof computers and information technology. Despitethese investments, however, most donor-instigatedpolicy reforms are not formulated well enoughto be implementable. Those that are decentlyformulated too often remain at the level ofintention. In the rare cases where policy reformsare implemented reasonably well, it is not clear

Edu

cation

Ref

orm

Sup

port

Toda

y

6what school-level impact they have, at least onquality. If after decades of investing in institu-tional capacity, a lack of such capacity still con-strains improving education, we have to askwhether our methods for investing in capacitydevelopment are effective.

Often, the alternative to putting resources intogrand plans for sector reform is to investinstead in projects that directly introduce inno-vation and improvement at the school level.Introduction of new teaching techniques or cur-riculum improvements, teacher and administra-tor training, and improved school-communityrelations and interactions are often the stuff ofprojects seeking to improve quality and effec-tiveness. Sometimes real successes areachieved—access is increased, quality isimproved, and learning can be demonstrated—and the hope is that somehow these success-es can be replicated or taken to scale. But toooften they are not. A good case in point isInteractive Radio Instruction (IRI). In spite ofsuccessful experiments, and studies thatdemonstrate the effectiveness of this interven-tion, scale-up and sustainability of donor-basedIRI interventions is quite disappointing. Therecord is better where such technologies havebeen developed endogenously.

If we fund school-level projects, then the chal-lenge is how to create the policy and institu-tional reforms that support, extend, and sus-tain school-level and child-level success. And, ifwe support policy-level interventions that targetnational policy and institutional reforms, thechallenge we face is how to ensure that thosereforms are implemented in a way that leads tochanges in schools’ day-to-day practices. Inboth cases, the challenge is designing effectiveprograms of what we call reform support.

In 1997, USAID published the Education ReformSupport (ERS) series. The ERS approach contends that successful reform programs needto direct assistance to overcoming the politicaland institutional obstacles that too often preventnational-level policy changes from being identi-fied, pursued, and implemented. In revisitingand updating the ERS approach 10 years later,

it is clear that those same obstacles must stillbe overcome, whether we are designing policy-level programs that help shape national reformagendas or school-level efforts that work directlyto introduce innovations that can be taken to scale.

This paper takes an updated look at theEducation Reform Support series and attemptsto consolidate many of its theories, frameworks,and operational guidelines. Annex A summarizesthe original body of work. Ten years of experiencehave reconfirmed the ideas on which ERS wasfounded and have added a multitude of examplesto show how strategically targeted and oppor-tunistic support can move a reform agenda forward—instigating demand for change, promoting dialogue around and new insight intoissues, leveraging implementation, or removingpolitical and institutional barriers to success.

This paper starts with the more general andtheoretical considerations, and then moves tothe more operational and concrete issues. Insection 2, we explain why there is a need tohelp countries sustain their own reforms, ratherthan simply ameliorate conditions via school-level projects and donations. We note thatthere is a growing literature in this area. Thesection provides the “theoretical” or “develop-mental” justification for why donor agenciesneed to support education reform. Havingestablished the importance of the issue, insection 3 we lay out some basic concepts thatare helpful in understanding how to supportreforms. Section 4 provides a concrete catego-rization of possible reform support activitiesand offers plenty of examples of such activi-ties. Section 5 explains why not all countriesand situations are equally suitable for reformsupport and suggests how to adapt reform sup-port ideas to difficult circumstances. Section 6provides operational (as opposed to theoreti-cal) justification for reform support in terms ofthe imperatives of most donor agencies andthe professionals that work in them. Finally,section 7 provides some examples of howdonor agency officials can “jump-start” reformsupport activities, even in projects that may nothave been designed with this purpose in mind.

Education

Reform

Support

Today

72.Dealing with Scale and Sustainability:Moving Beyond the TechnicalChallenges of Education ReformSince publication of the ERS series in 1997,interest in understanding education reform hasonly increased. The recent literature, on whichwe draw throughout this document, has beenlargely consistent with the motivations and principles the ERS approach first laid out adecade ago. However, the literature does notoffer many “practical” guides to providingreform support assistance. It seems to implythat the typical reform process is so messyand so political that it is too difficult to offerspecific suggestions about tools and techniquesthat can help in its management. While thismay be satisfactory in an academic study, it is of little help if the aim is actually to design projects to support reform. Providing tools andtechniques was a hallmark of the ERSapproach, and this is taken up again in thisupdate. This section looks at the last 10 years,notes what has changed in the world and whatother donors and experts have concluded during that time about reform processes, and explains that the need for a systematicapproach to supporting reform is as greattoday as it was 10 years ago.

SCALE AND SUSTAINABILITY

The central challenges of scale-up and sustainability have not gone away. We wrote 10 years ago in the ERS series that good educational practice can be found anywhere,but that the central challenge of educationreform is that good educational practice is notfound everywhere. To make good practice thenorm rather than the exception, ERS stipulatedthat reformers recognize that programs or projects often do not go to scale and are not

sustained, understand why not, and developmeans to address the blockages that preventscale-up and sustainability.

For one, the last decade of research and experience reconfirms the Education ReformSupport explanation of why scale and sustainability, while always pursued, are rarelyobtained. Implementing large-scale change isinherently different from introducing minoradjustments or small-scale, pilot initiatives(Elmore, 1996; Samoff & Sebatane, 2001).The existing situation in a given education systemis not an accident. Well-entrenched interestgroups have worked to ensure that whicheverpolicies, resource decisions, and institutionalarrangements are pursued are those that willdirectly benefit them or, at a minimum, will notthreaten the status quo from which they arealready benefiting (Moe, 2003; Hess, 2004b).When a proposed school-level reform is seenas only affecting a small part of the system,then important interest groups may not bethreatened. When large-scale, policy-level changethat entails resource reallocations with tangiblewinners and losers is proposed, then the interestgroups resistant to change see the need torespond and to do so forcefully (Inter-AmericanDevelopment Bank and David RockefellerCenter, 2005). The natural outcome of thisdynamic is often partial implementation of manyprograms, rather than full-scale implementationof any particular reform. People end up settlingfor what they think they can accomplish giventhe local political realities, rather than whatthey should accomplish (McDermott, 2000).

Edu

cation

Ref

orm

Sup

port

Toda

y

8For this same reason, large-scale, policy-levelreforms are often difficult to sustain. Interestgroups opposed to reform may be overcome ona given issue, but they do not then just go quietlyinto the night. They use their political andstrategic skills and resources to fight back(Moe, 2003). The example of the teachers inSan Antonio (see text box) illustrates this perfectly. The superintendent was able to pushthrough the curricular changes she wanted (andthe district needed), but ultimately the teacherspushed back, not only eliminating the new mathcurriculum, but also persuading the board ofeducation that the superintendent had to go. Infact, it is easier for organized resistance toreform to block change, which sometimes onlyrequires that one refuse to take action. It ismuch harder to successfully carry out a reformthat often requires numerous strategies, multiplelayers of implementation management, andconstant negotiation of a variety of political hurdles (Moe, 2003).

LITERATURE CONFIRMS REQUIREMENTS

OF REFORM

Reform needs to be seen as a dynamic, largelypolitical process that unfolds within particular,perhaps changing, institutional contexts. A series of strategic decisions determineswhether entrenched political and institutionalinterests and biases can be altered or overcometo achieve the desired change (Grindle, 2004).Successful reform therefore requires leadershipthat recognizes the political battles and trade-offsto be fought or negotiated. Grindle’s review ofeducation reforms in Latin America in the1990s highlights how political leaders overcameobstacles to reform by controlling the timing ofreform, setting the terms of political debate, andweakening opposition within key institutions bydrawing on competing institutional resourcesand mobilizing broad public support (p. 21).

In addition to leadership that can navigate thepolitics of reform, successful educational reformrequires attention to the institutional frameworksthat govern operation of the sector. Within eacheducation system, a complex set of institutionalrelationships and arrangements will have grownup over time, imparting behavioral norms and

codes, prevailing models for how work getsdone, and entrenched interests happy with thebenefits that accrue to them under the statusquo. Operational changes in the management,administration, financing, and accountabilitysystems will bump up against these institutionalcontexts, often threatening the existing political

In the late 1990s, the San AntonioIndependent School Districtexperienced unprecedented gains instudent outcomes for a large, urbanschool system in the United States. Forexample, mathematics achievement forAfrican American and Latino studentsmore than tripled in just three years.The district had introduced a newmath curriculum that teachers andadministrators recognized as a sound,high-quality, successful program.However, when the teachers’ associationrequested that staff be allowed tovote on continued use of the program,70 percent of district teachers votedno. That’s right; they voted not to usea program that they admitted worked.Shortly thereafter, the superintendentwho had led the impressive gains inSan Antonio was fired.

Why did this happen? The teachers’association leadership said they likedthe program, but they did not like theway it was being implemented. Theysaw it as too forceful an obligationimposed on them without (in theireyes) sufficient consultation. Thesuperintendent, who gained recognitionaround the country for her academicleadership, failed to build the politicalsupport she needed in her own districtto carry through the reforms shechampioned.

What’s the moral of the story?Technically sound school-level solutionscan be implemented and have goodeffects, but by themselves they cannotaddress all of the challenges ofreforming and improving education.

Education

Reform

Support

Today

9winners with a potential loss (whether real ormerely perceived) of status, authority, oraccess to resources. At a minimum, reformrequires additional capacity beyond what theinstitutions in the sector are accustomed to doing.

Reforms seeking to improve education (makingschools more effective, producing better educational results) are necessarily aboutchanging how resources are used. When theprimary concern is expansion of access, reformsare often about how to use the additionalresources that would be required—more moneyin the budget, more schools to be built, moreteachers to be hired. When the central challengeis ratcheting up the quality of education, thenthe use of existing resources needs to be challenged more directly. This raises the stakesfor the political and institutional interests in the sector (Inter-American Development Bankand David Rockefeller Center, 2005).

Last, since education is often the largest,most visible public service, there is amplescope for civil society to play a variety of rolesin either supporting or opposing educationreforms. In democratizing contexts in particular,the roles of public constituencies are magnified,as elected officials will feel pressure to respondto the issues (Grindle, 2004). Organized interests, of course, will have an advantageover diffused sets of actors, so teachers’unions, for example, may exercise even moreclout given their well-practiced ability to dominatecivic debate. However, there is opportunity forother civic actors to organize and present their points of view and, perhaps, marshal constituencies large enough to exert counter-vailing political weight.

While all the above is important in consideringwhat issues need to be addressed for reformto succeed, this paper does not try to explainall the complexities and interactions of everyconceivable factor that will influence the outcomeof an education reform. Education ReformSupport by definition is preoccupied with whatoutside agents can do through programs andprojects to increase the probability of reform

success. Since most development projects payattention to the technical aspects of reform, weleave those aside. Conversely, since most projects fail to address the process and politicsof reform, it is on those that we focus. In particular, ERS concentrates on the political andinstitutional dimensions of reform. We touch onleadership, institutional capacity, resources,and civil society issues, but within the contextof how one uses them to define, advocate for,advance, and carry through reforms in the faceof political and institutional obstacles.

CHANGES IN THE DONOR ENVIRONMENT MAKE

REFORM SUPPORT ALL THE MORE IMPORTANT

In the 1990s when the ERS approach wasbeing documented, education reform effortswere shaped in large part by the structuraladjustment approach of the World Bank. Reformprograms adhered to the notion that getting the policy framework right would allow sectordevelopment to take off. Sector policy adjust-ments were needed as education systems wereoften plagued by inadequate or inappropriateresource allocations, poor management, andmisdirected policy. Both the World Bank’s sector adjustment programs and efforts suchas USAID’s nonproject approach tended toassume that policy change is a technical matterof shifting a “parameter” or two (devote morefunding to girls’ education, set up a properEMIS unit, set up a quality assessment unit),or that, if politics was involved, a quid pro quoof financial support in exchange for policychange would be sufficient. At most, formal,one-shot “dialogue” sessions of a few hours, ina large gathering, would be carried out. ERSwas formulated to make the case for greaterattention to the politics of reform and to highlight the inherent differences betweenmacroeconomic “adjustments” and the implementation-intensive reforms required inthe education sector. Changing managementand administrative structures in education sothat more effective teaching techniques can beused in classrooms across an entire countrypresents so many opportunities for reforms tobe sidetracked, resisted, distorted, or ignored,compared to, say, getting the central bank to

Edu

cation

Ref

orm

Sup

port

Toda

y

10drop a pegged exchange rate. ERS made thecase for devoting resources to the strategicmanagement of the politics of reform, becausecentrally planned sectoral programs thatassumed a macro adjustment approach wouldwork, when applied to the education system,were stumbling at both policy formulation andimplementation.

Since the late 1990s, the dominant internationaldevelopment philosophy has maintained a sector reform approach, but added to it a more coordinated donor effort to invest in poor countries that can demonstrate good governance.Sector-wide approaches (SWAPs) emerged inthe late 1990s to address the challenge ofpoor local ownership, uncoordinated and competing donor priorities and strategies, and excessive and uncoordinated reportingrequirements. The SWAP structure is still beingdeveloped, but its core intent is that donorsand governments agree on a single program,and donor funds are “pooled” in support ofthat common program. This, in theory, enablesgreater donor leverage for policy change,improved donor coordination, and reducedreporting and management burdens for ministriesof education. An increased use of general ordirect budget support funding mechanisms, aswell as basket funding among multilaterals andsome bilateral programs, is intended toaddress the shortcomings of project-based aid,such as high transaction costs, unpredictablefunding, donor-driven agendas, and parallel systems that undermine government capacityand accountability.

These recent trends in funding mechanismshave moved hand in hand with the increasedcoordination of donors and recipient countriesat an international level through establishmentof international goals: the Education for All initiatives begun in Jomtien (1990) and Dakar(2000), the Millennium Development Goals(2000), the Monterrey Summit (2002), theRome declaration on harmonization (2003),and the Paris Declaration on aid effectiveness(2005). In 2004, the United States establishedthe Millennium Challenge Corporation based onthe premise that development assistance is

more effective when it is provided to countrieswith policies in place that promote economicgrowth, poverty reduction, and good governance.The emphasis is on countries demonstratingtheir policy commitment in well-defined plansand strategy papers. The emerging paradigmstresses country-led development and, therefore, takes on the fundamental issues of ownership and coherence.

These recent trends are encouraging in thatthey at least promise to improve the ways inwhich funding agency intentions and programsinteract. What they fail to address are the internal dynamics of any country that actuallydetermine whether the well-defined plans andstrategies can effectively translate into imple-mented system-wide reforms. What we know isthat even when policies can be well formulated,centrally planned reform does not automaticallytranslate into sustained policy change, muchless into implemented improvements at theschool level (Moulton, Mundy, Welmond, &Williams, 2001). Current thinking, epitomizedby Easterly’s (2006) critique of developmentassistance, argues strongly for moving awayfrom a central, grand, comprehensive plan and, instead, supporting local capacity to innovate and “seek” solutions, and to unleashentrepreneurial energy in response to develop-ment challenges such as providing effectiveeducation for all children. Emphasis amongeducation reformists has shifted away fromattention to resource allocations and “grand”policy formulation as ends in and of themselves.Policy reforms and resource reallocations arestill called for, but attention must be paid tohow one sustains and improves on policychange, constantly and at the margin, andmust be approached from the perspective ofhow those changes will lead to more effectiveschools.

The ERS approach has always emphasized theneed to work on political and institutionalobstacles to supporting and sustaining school-and community-level changes. Funders are stillinvolved in supporting policy-level reforms,whether in the context of the FTI, a SWAP, ordirect bilateral programs. In these policy-level

Education

Reform

Support

Today

11programs, the key question remains, “How donational reform efforts translate into realchanges in the institutional environments thatcreate conditions for success at the schoollevel?”

As funders also invest resources and effort inprojects that can demonstrate impact in termsof improved student outcomes, the tendencywill be to do more direct piloting of school-levelinterventions. These kinds of targeted interven-

tions have to developERS specifically

addresses the steps

needed to ‘walk up the

ladder’ from projects

to policy reforms.”

“ explicit, strategic theories of change thatstate how the school-and community-levelsuccesses will be usedto leverage institutional,policy, and resource

changes in the system. ERS specificallyaddresses the steps needed to “walk up theladder” from projects to policy reforms. Thosesteps, by necessity, involve identifying the insti-tutional requirements of the “reformed” sys-tem—that is, changes in how teachers interactwith students, how schools are organized, howministerial personnel hold schools accountable,and how the public stays informed aboutresource use. Considering how to put in placethose institutional requirements forces pro-grams and projects to take on the political andinstitutional interests wedded to the existingrules of the game.

Whether school-level programs are trying tohave an impact and be sustained beyond theimmediate scope of a project, or assistance topolicy-level reforms is going to lead to impactat the school and community levels, we areforced to deal with the politics of reform andthe institutional contexts that govern the education sector. Because ERS is primarily aboutaddressing the politics and institutional contextof reform, we elaborate on these two below.

THE TOOLS AND TECHNIQUES DISCUSSED IN ERS

ARE STILL VALID

Ten more years of experience and research only reiterate the primacy of politics in educationreform. ERS introduces some tools and

techniques for identifying, formulating, andimplementing strategies to deal with those politics, and stresses data and information ascritical to shaping the dialogue and debatearound what is needed. In addition to analyticaldata, ERS promotes communication and marketing approaches to information use—forexample, surveying stakeholders to understandtheir points of view, market-testing differentways of defining or highlighting issues to seehow key interest groups react, and promotingpublic dialogue and deliberation over policyoptions. ERS also emphasizes the need to network with a variety of actors and stakeholdersto form coalitions that could support reformsand work to overcome other interest groupsthat may oppose them.

What recent research and experience haveadded to ERS’s original set of political tools andtechniques are a greater emphasis on collabo-ration and purposeful constituency building insupport of reform. It is not enough merely toidentify and seek to include key stakeholders indialogue. For example, Navarro points to theinherent disadvantage government has whenconfronting union leadership that has a longerplanning horizon. Unless other forces are mobilized, this disadvantage leads to unionsgetting what they want more often than government getting what it needs (Inter-AmericanDevelopment Bank and David RockefellerCenter, 2005). What adds to the challenge isthat the entrenched interests are well organizedand have sharp, narrow objectives, while thebeneficiaries of reform are loosely organizedand often unaware of the their interests andpotential gain. For these reasons, successfulreform requires dedication of resources andeffort to forming purpose-driven networks oforganizations that can align and mobilize largeconstituencies in support of particular reformmeasures. A Rand Corporation study (Bodilly,Ikemoto, & Stockly, 2004) highlights the importance of technical assistance in supportingthe building of just such collaborations.

Experience in Guinea helps illustrate the abovepoint. In the mid-1990s Guinea wanted to redeploy surplus lower secondary school

Edu

cation

Ref

orm

Sup

port

Toda

y

12teachers to primary schools that lacked staff.Those secondary teachers who were workingonly a few hours a week and living comfortablyin small cities around the country were a natural constituency that would oppose justsuch a reform. The Ministry recognized theneed to mobilize another constituency thatwould lobby in favor of teacher redeployment.Parents and community members in villagesthat had schools but no teachers were activatedas a constituency in each district through amedia campaign and local dialogue sessions.Surplus secondary teachers could no longerdefend their situation while sitting across thetable from the people whose children weredenied access to primary school because of it.

NEW FOCUS ON THE INSTITUTIONAL

ENVIRONMENT FOR REFORM

The original ERS approach focused on interestgroups and pressures, but tended not to obviatethe internal logic of institutions as a source ofinertia. Schools and their communities need tooperate within an institutional environment thatcreates expectations, norms, incentives, anddisincentives for the different choices they maymake. Reform efforts need to pay attention tohow the institutional environment does or doesnot support the intended changes, and, moreimportant, reforms need to define specificstrategies for altering the institutional environ-ment so that implementation is actually feasible.The importance of infrastructures external toschools and their communities that can simultaneously challenge schools to do betterand nurture them in their efforts to improveare cited both by Fullan (2000) and theAnnenberg Institute for School Reform (SchoolCommunities that Work [SCtW]). Tangiblerewards for success and sanctions for failureare important but by themselves will not dramatically change behaviors. Elmore (1996)argues that the normative environment—theset of accepted behaviors and habits that arepart of the organizational culture of an educa-tion system—determines how people will reactto a system of rewards and sanctions betterthan any tangible benefit or loss promised bythe system. Reform strategies need to addresshow system infrastructure (whether within the

ministry of education or arranged through partnerships and affiliations with other organizations) is positioned to reinforce achanged normative environment and providethe ongoing supports that schools and communities need to succeed.

In discussing the challenges to implementingeducation reforms, Bodilly et al. (2004) pointout that changing behaviors within organizationsis one of the most difficult tasks to accomplish,and it is especially so when multiple levels ofgovernment are involved. For any reform to succeed, numerous individuals and groups inthe concerned institutions are expected tobehave differently. These groups “respond toand are driven by many varying incentives, rules,and regulations inherent in the infrastructure ofschools and schooling” (p. 17). Attempts toimplement reforms often lead to adaptationsthat include some changed behaviors consistentwith the reform goal, some that represent coping with the reforms, and some that activelydivert or subvert the intentions of the reform.Less than desirable outcomes occur becausethe supports that proposed changes requireare not put in place, and certainly are not institutionalized as supporting infrastructure.

Two examples, one negative and one positive,illustrate this point. In a study of schoolimprovement in four countries in East Africa,Heneveld and his colleagues (Heneveld,Ndidde, Rajonhson, & Swati, 2006) found thattextbooks were more available at the schoollevel than they had been in the past, but thebooks were still not used regularly to enhanceinstruction. They concluded that this wasbecause neither the supporting infrastructureof training for teachers on how to use booksnor regular supervision and support to reinforcethe use of textbooks in the classroom waspresent. In fact, Heneveld et al. found thatsupervision and support were infrequent,lacked any meaningful follow-up, were unfocusedand confused several different purposes, andwere not visibly associated with improvingschool outcomes. In contrast, in a study of ninecommunity-based complementary educationprograms, DeStefano, Hartwell, Moore, and

Education

Reform

Support

Today

13Balwanz (2006) found that the effectiveness ofthe community schools derived in part fromnetworks of well-trained teacher support personnel that visited schools at least once amonth. Teachers in these programs are giveninitial training and additional intensive supportduring their first year and continuing for several years

IMPLICATIONS FOR DONOR PROJECTS

Whether supporting policy-level reforms or initiating school-level improvements, if fundersare going to provide targeted assistance, thatassistance has to be strategic in addressingthe political and institutional obstacles outlinedabove if it is going to leverage impact beyondthe immediate scope of a project. This sectionhas argued the following: that scale and sustainability are still a problem, that the morerecent literature concurs that politics are para-mount, that changes in the donor environmentmake a reform support approach all the moreimportant, and that the basic tools and techniques proposed 10 years ago are still theright ones. This approach is enriched by payingattention to institutional inertia. We now turn tohow one takes all these factors into considera-tion in supporting reforms.

Edu

cation

Ref

orm

Sup

port

Toda

y

14

Education

Reform

Support

Today

153.Clearing Space, Filling Space,Building Reform SupportInfrastructure Scale-up and sustainability remain elusive for ing patterns and then devote resources, time,most education reform efforts because of intel- and effort to making the case for why changeslectual, political, and institutional resistance. in the existing circumstances would be better.Doing something about those limitations If improved quality requires more interactiverequires, first, recognizing that the existing teaching methodologies, then teachers, admin-arrangements in any education system are not istrators, students, and parents need to be dis-accidental. Making room in the existing political abused of their prevailing mental model that aeconomic landscape of the education sector— good classroom has children quietly attentiveclearing space—so that the substance of to the teacher’s every word. In the case of asound educational improvement can be intro- policy-level project focused on more broad-scaleduced—filling space—constitutes the opera- reform, the same intellectual space-clearingtional core of the Education Reform Support may be required. For example, to promote fund-approach. ing reform successfully, a project may need to

help build the case for why weighted allocationsCLEARING SPACE based on need would be more equitable than

uniform capitation grants (something that mayOne of the major problems in setting forth andgo against the prevailing mental model of whatsustaining a reform agenda is the existence ofconstitutes fairness).preconceived ideas, mental models, institution-

al traditions, and economic and social benefitsIn other cases, the space that needs to bederived from the existing arrangements. A keycleared may be more political than intellectual.step in supporting reform is thus to reduce theEveryone knows that more teachers are neededintellectual and political space taken up byin rural areas, but the teachers’ unions workexisting views and interests.hard to protect the interests of their memberswho prefer working in towns and cities. UnlessWhy does the U.S. school year still follow theanother political force can be mobilized torhythms of an agrarian society that has notoppose the interests of unions, space will notexisted for at least 70 years? Why are curricu-be cleared to allow reforms in teacher deploy-lum documents in almost all cultures written inment. Ministry officials who are concerneda common and impenetrable jargon? Why doabout unrest in the capital city will make suremost classrooms around the world have chil-that resources are spent on urban schools anddren sitting in rows copying notes off a black-university students. Only when the interests ofboard in the front of the room? To proposethose who are on the losing end of thatreforms to these well-entrenched views of whenresource allocation are voiced, organized, andschool is in session, what curriculum looksmobilized, will it be realistic to expect that poli-like, or how classrooms should be organized,cy will change. Clearing political space involveswe have to clear intellectual space. School-levelmobilizing the stakeholders and interests thatprojects that want to introduce and bring tostand to gain from proposed policy changes,scale innovations in education need to recog-and working strategically to overcome or neu-nize how those innovations depart from prevail-tralize the interests that oppose change.

Edu

cation

Ref

orm

Sup

port

Toda

y

16A third kind of space-clearing concerns theinstitutions involved in managing and imple-menting education. Directorates within the cen-tral ministry, regional and local educationoffices, and other officials within the systemhave all figured out how to operate to their ownbenefit within the existing set of institutionalarrangements and prevailing rules of the game.To change what they do, how they do it, andwith what degree of effectiveness and account-ability, requires clearing space in terms of theprevailing institutional norms, expectations,incentives, and relationships. Before we canexpect new managerial relationships andadministrative and organizational practices totake root, we first need to understand thedynamics that underlie the existing institutionalcontext. Grindle (2004) points out how almostall the reforms she reviewed in Latin America inthe 1990s “imposed at least short-term bur-dens on politically important groups that readilydefined themselves as losers”—for example,teachers’ unions that feared losing jobs andbenefits (p. 11). Bureaucrats within LatinAmerican ministries also saw themselves aslosers, fearing loss of jobs or responsibility orhaving to deal with genuine job performancestandards. Political party leaders who relied onpatronage as part of their power base also sawthemselves as losers if new standards were tobe implemented for how jobs were assignedand carried out. Furthermore, Grindle observesthat the institutions within which the potentiallosers (and winners) defend their positionshave “histories that shape values and behav-iors and that create ongoing incentives for con-flict and cooperation over time” (p. 14).

In addition, if other actors are to becomeinvolved as part of the institutional capacitymobilized to support reform, then space alsoneeds to be cleared in terms of how govern-mental and nongovernmental institutions(NGOs) interact. Can ministry officials partnerwith NGOs or private sector firms? Are theremechanisms within the bureaucracy for formal-izing relationships with community-based organ-izations? In Mali, in 1995, the Ministry ofEducation created an office and wrote specificpolicy to promote and support the establish-

ment of community schools. This created anew, official way for the government to interactwith NGOs in providing schooling. A coalition ofNGOs, with USAID support, worked to convinceMinistry officials that it was in their interest toofficially recognize and support these schoolsbecause they were reaching children that thepublic sector could not. Those NGOs clearedenough institutional space so that the newoffice and policies could be put in place.

“Clearing space” thus refers to the need toreduce the influence of traditional interests andviews. The tools one can use to carry out thisspace-clearing are listed in section 5, but, tosummarize, the task generally involves bringinghard evidence, and countervailing politicalpower, to clear away obstacles. It implies tryingto isolate narrow interests or traditional modesof belief both intellectually and politically andtrying to demonstrate the negative effect ofthese interests on the general good.

FILLING SPACE

In education, as in physics, nature abhors avacuum. If intellectual, political, or institutionalspace is cleared in the education landscape,something will fill that space. There is no lackof “solutions” waiting to be proffered—often bythe very political and institutional interests thatneeded to be cleared out in the first place.Educational quality needs to be improved? Theteachers’ union is always prepared to arguethat one needs to increase salaries to attractand retain high-quality teachers. The comput-ers-in-education lobby is ready to argue thatconnectivity to the Internet is the key. The business sector will press so that English ortourism studies are stuffed into an alreadyoverloaded curriculum. Because space getsfilled no matter what, projects that support sus-tainable reform need to invest in filling space inconjunction with their investments in clearingspace. But how do we know which space-fillingactivities to promote?

Space-filling activities should be judged primarilyby whether they are effective and whether theyachieve their effectiveness at reasonable cost.

Education

Reform

Support

Today

17Easterly (2006) argues that most developmentprojects fail to systematically evaluate whetherthe innovations they introduce actually producemeasurable impact. While the details of educa-tion sector reforms and innovations are obviouslycontext-specific, there is an identifiable set ofreforms that research and experience show ismost likely to contribute to greater educationaleffectiveness. Space needs to be filled in sucha way as to create clear objectives for theschooling system; provide sustained, adequatefinancing to schools to achieve those objec-tives; give principals and teachers sufficientautonomy to manage for results; and createaccountability systems to monitor and respondto results (Pritchett, 2004, p. 65).

ENSURING SCHOOLS HAVE THE RESOURCES AND

AUTHORITY THEY NEED

Space-filling activities that are most associatedwith tangible improvements in outcomes forstudents include what the Annenberg Institutefor School Reform (AISR) refers to as changesthat ensure schools have the power andresources that enable them to respond to theneeds of their students. School administratorsand teachers (and their communities) must havethe knowledge, authority, and organizationalcapacity to respond to their day-to-day challengesif reform is to be visible in the daily work ofschools (Earl, Watson, & Katz, 2003). Forexample, research on effective communityschools has highlighted the important roles thatlocal governance and decision making play inorganizing schools that more consistently providerural, poor children with better opportunities tolearn (DeStefano et al., 2006). Research onautonomous schools in Nicaragua found thatthose schools with greater local authority havebetter student outcomes (King and Ozler ascited in Winkler, 2006). Two cross-nationalstudies using Trends in International Mathematicsand Science Study (TIMSS) and Program forInternational Student Assessment (PISA) datafound that schools that manage their own budg-ets and recruit their own teachers have highertest scores than those that don’t (WoBmann andWoBmann & Fuchs as cited in Winkler, 2006).

SCHOOL-COMMUNITY COLLABORATION

A growing emphasis on local control is not limit-ed to delegating more power and resources toschools. Ample research now argues forincreased school-community collaboration as avital component of improved local governanceand decision making. Doherty and Abernathy(1998) state that improving relationshipsbetween schools and their communities is avital part of making any kind of lasting changein the learning environment. Dyakowska (2004)notes that professionals need to recognize thatparents can be considered experts in knowingwhat is best for their children. Warren (2005)emphasizes that community initiatives canimprove the social context of education so thatchildren come to school better able to learn.Mastro and Jalloh (2005) report there is grow-ing evidence that successful collaborationbetween school and community groups has ledto improved academic and social/emotionaloutcomes.

PUBLIC-PRIVATE PARTNERSHIPS

In addition to greater local control, purposefulpublic-private partnerships hold promise as effective space-filling strategies.Nongovernmental organizations have continuedto play increasingly significant roles in promoting,supporting and, carrying out educationalimprovements around the world (Shanti, 2001;Miller-Granvaux, Welmond, & Wolf, 2002). Thegrowth of charter schools in the United States;the success of large-scale nongovernmentalcomplementary education programs as diverseas the Bangladesh Rural AdvancementCommittee (BRAC) and the School for Life inGhana; and the increased role of nongovern-mental actors in shaping policy and supportingimplementation all demonstrate how the notionof education system capacity needs to beexpanded to include more than just the officialministerial education apparatus (Bray, 2000;Care International, 2003; Lake & Hill, 2005;Miller-Granvaux et al., 2002; Nath, Sylva, &Grimes, 1999).

Edu

cation

Ref

orm

Sup

port

Toda

y

18ACCOUNTABILITY FOR RESULTS

If schools are to have greater authority andcontrol over resources, then part of what fillsspace in the education landscape must also bemechanisms for increased accountability.Annenberg argues that accountability shouldnot be limited to an accounting of resources,but should focus on demonstrating that addi-tional resources and authority are being usedto produce better results for students. Schoolsand local districts need to know current andpast results so they can track how their strate-gies are (or are not) contributing to improve-ment. Local districts and their partners needdata systems that enable them to monitor stu-dent, school, and partner performance asmeasured against the results they expect. Areresources actually available and used asintended, are programs being implemented upto expectation, are results improving as fast ashoped? Making these kinds of data publiclyavailable requires communication techniquesand savvy (Sexton, 2004)—for example, know-ing or being able to figure out the best ways toshare information on finances, on programimplementation, and on student performanceso that the various groups can have equalunderstanding of the issues involved.

Making public the budget allocated to eachschool in Uganda—by requiring that it be postedon a wall at the school and having it publishedin the newspaper—greatly enhances each community’s capacity to hold the schoolaccountable for demonstrably using thoseresources to improve education. School performance appraisal meetings in Ghana bring communities and their schools togetherto discuss student achievement results on the national examination, culminating in jointagreements about what can be done toimprove student outcomes. A similar process of school self-assessment has become national policy in Namibia.

TIMELINESS AND EFFECTIVENESS OF SCHOOL

AND TEACHER SUPPORT

If schools and their communities are going tobe held accountable for results, then, concomi-tantly, the system needs to be accountable formaking sure the supports and interventionsthat can assist struggling schools and commu-nities are available and effective. Timelinessand effectiveness of support interventions arecritical system capacities for successful reformimplementation. DC VOICE, a community-basedreform support organization in Washington,D.C., organizes an annual survey to determineif schools are ready—in terms of staffing,materials, resources, and programming—fortheir students when they open in the fall (DCVOICE, 2005). This activity is an exercise inholding the school district publicly accountablefor making sure that schools have what theyneed to succeed.

Space needs to be filled with the infrastructurethat can support and sustain improvingschools. Research on community schoolsreveals that effective programs include regularteacher support and ongoing training. Frequentobservation and feedback regarding instruction,demonstration of specific pedagogical strate-gies, and organized meetings among teachersall contribute to improved school quality(DeStefano et al., 2006). Research also pointsto the importance of decentralized structures inmaking the education system better able torespond to and support schools and their com-munities. Galiani and Schargrodsky (cited inWinkler, 2006) found that provinces inArgentina with better management capacity atthe decentralized level saw improvements intest scores when authority was decentralized,while those with poor capacity saw test scoresdecrease. The 2004 World Development Report(World Bank, 2004), which examined how serv-ices like public education could be made moreresponsive to the needs of poor people, con-cluded that communities need closer, moredirect links to the authorities that make deci-sions regarding their children’s schools.

Education

Reform

Support

Today

19

FILLING SPACE AND THE CONDITIONS

FOR SUCCESS

Most school-level projects do invest in whatcan be characterized as space-filling activities,often ones that are consistent with theresearch summarized above. However, fillingspace requires not just introducing the innova-tions, but also structuring the way in which theyare introduced so that their impact can be eval-uated, the conditions required for successfulimplementation can be documented, and obsta-cles that may need to be “cleared” for thoseinnovations to be sustained can be identified.

This goes far beyond the usual institutionalcapacity issues—for example, district supervi-sors being trained to provide pedagogical sup-port to teachers. At issue is whether the pre-vailing institutional environment permits, pro-motes, and rewards the new job performancestandards—for example, on what basis are dis-trict supervisors evaluated, what traditionallyhas been their source of power and prestige intheir jobs, and what is the history of their rela-tionships to individual schools and teachers.Unless these other issues can be identifiedand understood, we cannot mount crediblestrategies for truly assembling the conditionsfor success needed to support teachers tryingout new pedagogy in their classrooms.

How do the space clearing and space fillingactivities that support reform get implementedand strategically coordinated? Whose job is itto see that space is cleared and that sound,cutting-edge innovations and improvements areintroduced?

REFORM SUPPORT INFRASTRUCTURE

Space is cleared and filled all the time in edu-cation systems. If the newly named ministerbelieves strongly in vocational and technicalhigh schools, he’ll clear space for investmentin building or equipping some of them by reduc-ing funds for other things. Numerous projectsfill all kinds of space with the solutions theyoffer—new textbooks, HIV/AIDS education,girls’ scholarships, construction of secondaryschools, Internet connectivity, and on and on.

San Juan Metropolitan Alliance forEducation (SJMAE): An Example of aMultifaceted Reform SupportInfrastructure

In forming the SJMAE, several organizations came together, eachbringing different capacities to the table.Sacred Heart University contributedexpertise in teaching and learning and institutional capacity for teacherprofessional development and support.ASPIRA, a nonprofit organizationsteeped in community organizing,brought to the collaborative provenmethodologies for training, mobilizing,and supporting parents and communi-ties. The College Board of Puerto Ricocontributed expertise in developingand managing assessments and usinga variety of data on student develop-ment. The Puerto Rico CommunityFoundation had ties to the local com-munity. This group of organizations allshared a common vision of schools aslearning organizations, school-commu-nity relationships as central to improv-ing education, and reform as aprocess grounded in what works at theschool and classroom level. TheAlliance formed relationships with 11schools and the district office inCataño, a poor San Juan neighbor-hood. It cleared intellectual space inthe local community by challenging theprevailing mental model of quality edu-cation and worked to fill space withschool-level innovations in teaching,school organization, communityengagement, and district supportactivities (based on personal knowl-edge of the authors).

Edu

cation

Ref

orm

Sup

port

Toda

y

20What is missing is a coordinated attempt toclear space for reforms that actually improveeducational opportunities and outcomes for the vast majority of students. And what is too often missing is a coherent set of space-fillingactivities that provide the supports needed toimplement those reforms thoroughly.

It is too much to expect that any education sys-tem on its own can clear and fill space in thecoordinated and strategic manner needed forspecific reforms to be developed, promulgated,implemented at scale, and sustained. As westated in our introduction, politics and institu-tional context dictate most of what can or can-not occur in an education system. Many of theactors within the system are affiliated with thepolitical interests benefiting from the statusquo arrangements. It is politically naïve to thinkthey will suddenly become advocates forreform. These actors are also part of the veryinstitutional context that needs to be changed.And the complex policy strategies needed tocarry out reforms far surpass the implementationcapabilities of most policy makers and educationsystem managers (Kronley & Handley, 2003). Ifa country is not effectively providing even basiceducation services, common sense suggeststhat the “meta” skills and attitudes needed tofix the policy problems (clear space, fill spacewith new ideas) are likely to be even scarcerthan implementation skills. Reform-minded governments and the donors supporting themthus have to “bootstrap” the ability to carry outreform by helping bring together networks ofactors who can start and carry on this work.

The original ERS series coined the term,“reform support infrastructure,” to describe thenetwork of organizations and individuals thatcan best help bring about successful reforms.The work of a reform support infrastructure iscomplex and spans a full range of competenciesand capacities, such as generating demand forspecific reforms, creating opportunities for andfacilitating dialogue among diverse stakeholders,organizing and mobilizing constituencies, engaging in strategic communication campaigns,developing specific technical “fixes” or models

of effective management and administration,presenting data and analysis in compelling formats, etc.

No single organization could possess this vastan array of capabilities. For this reason a viablereform support infrastructure, by necessity,depends on the collaboration and coordinationof a variety of organizations and individuals.What unifies a collection of actors is their commitment to a specific reform agenda. A network of public, private, and nonprofit entitiescould come together as a reform support infrastructure. The particular set of actorsneeded depends on both the nature of thereform challenge at any particular point as well as the sociopolitical context within whichreform is being pursued (Healey & DeStefano,1997). For example, the San Juan MetropolitanAlliance for Education was founded in 1994 bya diverse group of institutions in Puerto Ricowith a shared interest in supporting and sustaining meaningful changes in the learningexperiences of students (see the text box onthe previous page).

What a reform support infrastructure does atany point is totally context-specific, yet certaintypes of activities are fairly common. For example,if the need is to generate demand from a mobilized interest group to counterbalance the entrenched interests that prevent reform,then the reform support infrastructure couldconcentrate on using data and information in acommunication campaign to generate demand,and on using organizing and mobilizing tech-niques to gather constituencies to promote thenecessary reforms.

If what is needed is to broker agreement ornegotiate specific trade-offs among competingpolitical and institutional interests, then thereform support infrastructure could organizeand facilitate policy dialogue and deliberations.

If what is needed is ongoing help designing procedures and practices for implementation,then the reform support infrastructure couldmobilize technical expertise and implementationcapacity in the needed areas.

Education

Reform

Support

Today

21In each case, what is required is a flexible network of organizations and individuals capableof responding strategically to the emergingchallenges and opportunities associated with areform agenda. In reviewing a Ford Foundationinitiative aimed at supporting reform supportnetworks in eight U.S. cities, Bodilly, Chun,Ikemoto, & Stockly (2004) concluded that thesuccess of the reform support infrastructure ineach city depended on the amount of time,interaction, and trust among collaborative members; the legitimacy and authority of thelead organization (which was different in eachsite); the adept use of data to inform the group’stheory of action; the group’s commitment tocontinuous reflection and re-strategizing basedon hard evidence; and early attention to scaleand sustainability as key long-term goals.

The experience of District Community VoicesInformed for Education Change (DC VOICE)related in the box below illustrates the strategicactions of a reform support infrastructure.

The Strategic Actions of DistrictCommunity Voices Informed forEducation Change

DC VOICE brought together a numberof community organizations and indi-vidual activists to work initially in onecluster of elementary schools and amiddle school. The collaborative thenset its sights on the broader educationreform agenda, conducting researchand focus groups to identify criticalsystem-level factors that affect thequality of teaching and learning in D.C.classrooms. The framework they devel-oped, “Supports for Quality Teaching,”highlights needed reforms in howteachers are hired, developed, sup-ported, and evaluated. The frameworkwas used to clear intellectual andinstitutional space for thinking differ-ently about the human resources chal-lenges in the District. In addition,through partnerships with university-based researchers, data were gath-ered to analyze the experiences offirst-year teachers. Enlisting the com-munity-organizing and -mobilizing skillof DC ACORN, DC VOICE used thedata to clear political space by inform-ing and mobilizing parents and othercommunity members to advocate forbetter induction and mentoring for newteachers. DC VOICE testified beforethe school board and worked withschool district staff to fill the clearedspace with well-crafted induction andmentoring policies and programs(based on personal knowledge of theauthors).

Edu

cation

Ref

orm

Sup

port

Toda

y

22

Education

Reform

Support

Today

234.Education Reform Support Activities and ProjectsThe foregoing sections have noted the nonlinear and difficult to plan very rigidly. Whileimportance of supporting reform and the “big one can come up with discrete, projectizablepicture” aspects of reform support. From these activities (“tools in a toolbox”) to constitute asections it should be clear that supporting reform support process, the number and orderreform, as a donor activity, is different from of activities or tools one deploys is likely totraditional donor education projects and from change during the lifetime of a reform supportdonor approaches that assume reform is a effort; with certain tools in the toolbox beingsimple matter of laying down some policy deployed ahead of plan, and others relegatedsuggestions and conditions, and engaging in for later, depending on political opportunity, tim-a one-time formal discussion or workshop. ing, and what seems to be working. More thanImplementing projects that build schools, make traditional projects or one-shot dialogue ses-curricular changes, or train teachers is at least sions, ERS is opportunity- and event-driven. Ita known art, even if it is not a simple blueprint requires that donors have a well-stocked “tool-process. There are models to follow and box” and a deep awareness of the environ-

experienced staff in ment, so they can use many different tools in

reform support,

e any project, can

thought of as being

mposed of discrete

puts, processes,

d activities aimed

contributing to a

formed system…”

“…

lik

be

co

in

an

at

re

donor agencies who can the toolbox as opportunities permit or situa-mentor others who have tions demand, rather than according to a pre-never set up such set and linear plan. We realize this poses aprojects. And holding a challenge to donors’ planning. In the examplesworkshop or two to that follow we try to emphasize some of thesuggest a few policy nonlinear and opportunistic aspects of thechanges or discuss process, to highlight their importance and thussome conditions is also the need to rise to the challenge of beingsomething donors have responsive to opportunity, and to show thatexperience with. But with a little effort and creativity such changesimplementing reform can readily be made.

support activities in the manner discussedabove is relatively new ground. What does it The easiest way to think of “projectizable” ERStake? It might appear that supporting reform is activities, and the one followed in the ERSan arcane and purely intuitive “political” art. approach, is to adapt the classical corporateHowever, ERS proposes that reform support, decision support paradigm: data-gatheringlike any project, can be thought of as being leads to analysis, which leads to a rationalcomposed of discrete inputs, processes, and decision. In trying to support transparent publicactivities aimed at contributing to a reformed decision making, we alter this basic paradigmsystem, and, therefore, that donors can apply in two ways. First, the objective is not to reachtechnique to the messy process of supporting a single bottom-line rational decision, but toreform in a proficient manner. This section sug- promote deliberation concerning public policygests approaches and activities that can make trade-offs that leads, in a “searching”reform support itself a “projectizable” activity. approach, to better decisions at the margin.

The ERS approach, however, also turns theERS does retain some of the messiness of sequence around and suggests that it is policypolitical activity. The process is indeed inherently competition (dialogue, discussion, debate,

Edu

cation

Ref

orm

Sup

port

Toda

y

24social marketing) that leads to the demand for analysis, which, in turn, leads to the demand fordata gathering. It is the dissatisfaction, competition, and social tension created by policy debatethat lead to demands for more serious analysis; without such demand, serious analysis tends tofall on deaf ears. The ERS approach also augments this decision support paradigm by proposingthat donors need to help institutionalize within a country the ability to perform these dialogue,analysis, and data functions. One can then propose activities in any or (ideally) all of those areasto support a reform process. The text box below provides some brief examples of activities thatcan be supported relating to data, analysis, dialogue, and communication and building reform sup-port capacity. Further discussion and specific examples from donor projects follow.

Data

• Use of surveys, for example, on school quality, to demonstrate need for reform

• Use of existing data to drive discussions on equity of resource allocation

• Gathering political “data” through meetings and surveys to create political mapsof interest groups’ positions on key reform proposals

Analysis

• Cost-effectiveness analysis of various policy options, such as analysis of the costeffectiveness of more spending on salaries versus spending on materials

• Traditional internal efficiency (repetition, dropout, flow-through) and external(rates of return, labor market impact) analysis

• Inequality analysis on distribution of learning results, school access, or input andexpenditure distribution

• Cost projections and other simulation tools to underpin debates about the needto prioritize or to lobby for improved funding for the sector

• School effectiveness analysis to underpin debate about management reforms

Dialogue &Communication

• Forums for dialogue

• Policy presentations, debates, and deliberations

• Social marketing and mobilization campaigns

• Use of media

Networking &Reform SupportInfrastructure

• Encouraging and financing reform coalitions and networks

• Encouraging and financing analysis and advocacy NGOs by giving them perform-ance-oriented contracts rather than general purpose support

• Developing analytical and advocacy capacity in civil society and government itself

• Developing sustainable funding for advocacy and analysis institutions

Education

Reform

Support

Today

25DATA

Ultimately, empirical data must underpin theanalysis and promotion of reform ideas.However, promoting reform does not requirethat donors spend much energy and funding onthe creation and maintenance of traditional,supply-driven EMIS or recurrent, general-pur-pose school or household surveys. On the con-trary, there are good reasons to believe that,when there is no demand for data arising fromthe competition of policy ideas or the need tomanage the sector in a publicly accountablefashion, the data generated by EMIS systemswill tend not to be used, and the systems willtend not to be sustained (Crouch, 1997).Furthermore, countries already have manymore data than they are using. Data use is cur-rently far more constrained by demand and bythe lack of technical skill and imagination inusing what is already gathered, than it is by thesupply of data.

The sorts of “projectizable” data activities rec-ommended for ERS use existing data to feedinto dialogue processes, or, alternatively, gath-er original data but, again, for a specific policydiscussion purpose. This approach is very dif-ferent from usual EMIS or general-purposeschool (or household) survey sorts of activities.The recommended approach reasons first andforemost from the demand side: the data aregathered (or compiled if they already exist) toenlighten a specific policy dilemma. What followare three examples of different ERS data activi-ties from a variety of contexts. In each casethese activities were identified as useful inadvancing elements of the different reformagendas from different starting points, throughdifferent interactions with education officialsand the broader public, and with varying specif-ic tactics. What they have in common is anoverarching approach: data, used strategically,are an essential component of reform support.These examples are not meant to be copied,but we hope they and the others elaboratedthroughout this section of the paper will conveyhow ERS activities work, and demonstrate thatthey can accompany projects that include moretraditional education components.

Example 1. Data for policy-level dialogue inPeru. A case in point is a recent (2005 and2006) World Bank experience in Peru, wheresimple data on children’s reading were gath-ered to feed a policy dialogue process focusedon school quality and learning. Peru, which hadparticipated in PISA in 2000, carries out a sur-vey-based national learning assessment approx-imately every three years. These surveys areuseful, but it was felt that something more dra-matic and simple was needed to really focusattention. The World Bank team hit upon theidea of simply asking children to read in asmall but representative sample of 254 chil-dren in 22 schools, and recording (in bothaudio and video) the difference between a childwho can read fairly well and an average child(not reading very well, if at all), so that policymakers and opinion makers could see the dif-ference. The reading assessment was extreme-ly simple; it focused just on fluency (words readper minute in a simple paragraph) and compre-hension. A video was then created showing thedifference between children who can read andchildren who cannot read, pointing out that toomany children cannot read, and indicating thatthere are already quite a few schools wherechildren can read. These schools are character-ized by the use of standards and monitoring,teacher accountability and devotion, and sup-port to teachers on instructional techniques.Shown in live policy dialogue settings in fourlocalities to more than 2,000 persons, as wellas on prime-time TV, the video led to discus-sions on talk shows, editorials in the newspa-pers and magazines, and so on. A healthydebate on quality and learning standardsensued, and is continuing. This policy dialoguewas held during the presidential campaign, andone of the political parties adopted the issue ofearly-grade reading, including fluency goals(words read per minute) as an educationalstandard. The winning president adopted simi-lar standards and highlighted them in his inau-gural speech. These sorts of goals have alsobeen embodied as preferred performance indi-cators by the Ministry of Finance. While theidea all along was to use children’s reading asa simple and dramatic way to illustrate the

Edu

cation

Ref

orm

Sup

port

Toda

y

26need to improve policy around school quality,the whole notion of using a video, which endedup being extremely powerful, only arosebecause the analysts and World Bank staffobserved stakeholders’ powerful reaction whenthey heard simple audio recordings of somechildren’s halting reading compared to what aproficient child’s reading sounds like. Thus, theinnovation that turned out to be perhaps themost powerful one in the whole process, interms of generating debate, was an opportunis-tic rather than a planned one (see AlvaradoPerez, 2006).

Example 2. Data for local accountability inNamibia. Another case of using information tosupport reform, but this time at the schoollevel rather than at the policy level, is fromNamibia. A post-independence priority inNamibia was addressing the disparities of eco-nomic opportunity in society, particularly in theNorthern provinces, where the largely poor,black populations lived. USAID programs hadsupported education development for years,and by 2000 were highlighted by a project thatsupported key reforms built on a foundation ofactive learning, continuous assessment of stu-dent progress, and community-supportedschool improvement planning in four Northernprovinces. With support from the USAID proj-ect, some senior Ministry of Education officialsproposed a school self-assessment systemthat was designed by project staff working witheducation officers and tried in a few schoolswith parents, teachers, and principals startingin 2001. The project worked with theInspectors of Education, advisory teachers, andresource teachers who form the circuit supportteams in the regions to implement the system.These same teams then worked on revising theindicators and methodologies, thus acquiringfurther ownership of the system. Eventually,404 schools and more than 2,500 teacherswere participating in the School ImprovementProgram, which included implementation of theschool self-assessment process. The participat-ing schools used the system regularly to evalu-ate their effectiveness and improvementprogress. With the self-assessment, teachersare expected to reflect on their own practice

and use that reflection to participate in theoverall school improvement discussion.Namibia's school self-assessment system hasbecome a powerful means to include parentsand community members as well as teachers,principals, and local education officials in theprocess of defining and implementing schoolimprovement activities using empirical evi-dence. School Self-Assessment (SSA) hasenabled regional education offices to providesite-based teacher development and schoolimprovement support activities based on actualneeds, and it has been institutionalized as theMinistry's tool for operationalizing the newlyintroduced National Standards of SchoolEffectiveness at the local level.

It is interesting to note that this SSA was notpart of the original work plan in this project.Opportunity presented itself when a few region-al officers and Ministry officials took part in astudy tour to another country, where they sawsuch a system in operation (which was, in turn,an adaptation of the Scottish system) andbecame interested. The opportunity was seizedand an approach was collaboratively designedwith the Ministry. Furthermore, as use of thetool developed, and as local actors becameinvolved in a thorough adaptation rather thansimply copying a tool from somewhere else, itspurposes and uses diverged somewhat fromthe original design, and its interpretationbecame richer. The success of the SSA systemas an engine of system reform required a mixof initiatives, strategies, and circumstances.The Minister’s leadership was obviously a keyelement, but success depended on clearingspace and enabling active support at severallevels. Schools, teachers, and communitieshad many historical reasons for resisting hon-est reporting to central authorities. The strate-gy of using information to support reform wasitself a challenge. Much of the success depend-ed on establishing ways to use information atthe school level to address local problems, toprotect elements of school data to enable hon-est reporting, and to use the information in asupportive and nonjudgmental way (personalcommunication, Donna LeCzel, Chief of Party,BES3 Project, June 6, 2006; Carolyn Pugliese,

Education

Reform

Support

Today

27Program Associate, BES3 Project; see alsoMuhammed & LeCzel, 2003).

Example 3. Use of data to drive communityawareness and solve equity problems inNigeria. In various pre-selected local govern-ments, USAID provided support to improveequity in resource allocation via the use ofdata. A series of stakeholder forums was con-vened to discuss the level of funding allocatedto primary education and the equitable distribu-tion of existing resources among schools.Stakeholder analysis identified participants,who included representatives of parents,schools (teachers), local businesses, politicalparties, religious leaders, and local governmenadministration and elected representatives.Facilitated by local government educationadministrators, the forums took place over oneyear. Stakeholder representatives were firstasked to identify and prioritize the most impor-tant problems of the schools in their local gov-ernment. Information on those problems wasbased on existing school census surveys pre-sented back to the stakeholders at a subse-quent forum. Data were presented graphically,representing the distribution of resourcesamong schools, using indicators such aspupil:teacher ratios, pupil:classroom ratios,pupil:seat ratios, and books per child. However,feedback from stakeholders suggested discrep-ancies between the reality in the schools andthe data reported. It was decided therefore tocollect data collaboratively (parents and educa-tors) and present to stakeholders again beforethey were asked to use the information toestablish minimum criteria for each of theirproblems and develop action plans.

Two subsequent forums monitored implementa-tion compared to the action plan and then, inthe final forum, data were used to assessprogress. Important elements of the processwere that (a) district education officers led theprocess; (b) comparative data were used toengage the interest of politicians and parents;

t

(c) emphasis was placed on using informationto find local solutions to problems; and (d)problems referred to higher authorities werepursued by political and community stakehold-ers. USAID did not initially aim to improve situa-tions by adding resources. However, resourceconstraints limited progress in implementingaction plans developed by the stakeholders, soa one-for-one matching grant component wasintroduced, which leveraged funding from publicand private sector. By leveraging data, dia-logue, and the grants, the project enabled localresources to be allocated more efficiently andequitably. For example, the project helped to (a)increase resources from a variety of localsources; (b) reallocate physical resources(benches and teachers) among schools first,before providing additional resources, toensure that every school met minimum criteria;and (c) change state policy in response to localadvocacy for a specific reading hour in the cur-riculum. It should be noted that while theprocess was generally effective and wellreceived, it worked better in some areas thanin others, and this variation in itself is instruc-tive about the ERS approach. For example,changes tended to be more sustainable whenthere were changes in government (after elec-tions) in areas where dialogue involved not justthe incumbent politicians, but where the dia-logue was broader. The notion of involving non-incumbent politicians, also tried out in the Peruexample listed above, was highly innovative andsomewhat criticized at first (personal communi-cation, Alastair Rodd, June 10, 2006).

ANALYSIS

Raw data do not say much, without the value-added of an analytical process. Most traditionaldata gathering in education, however, assumesthat data are used for some vaguely specifiedfunction, called “management” (thus the “M” inEMIS), and therefore emphasizes the produc-tion of relatively raw data, which, not surprising,are not used for much of anything, except, per-haps, reporting to international organizations.1

1 At least in most education systems in most developing countries where donors work. Once data production is sufficiently agile and suffi-ciently based on actual management functions, and once there is sufficient demand for management decisions to be publicly defensible, rawdata use for immediate management applications does increase, and supply-side interventions make sense. But this typically begins to hap-pen only in upper-middle-income countries where most bilateral donors are absent or have a marginal presence.

Edu

cation

Ref

orm

Sup

port

Toda

y

28Analysis itself ideally has to be driven bydebates about issues that require reform; italso should not be supply-led. Thus, analysis ofschool finance, for example, should be drivenby a sharp policy concern over funding equity.Or analysis of school effectiveness has to bedone in response to sharp budget pressureleading to a need to increase efficiency. Ifanalysis is promoted solely out of a vague per-ception, or conventional wisdom, that it is“right” and “professional” to carry out analysisbefore making decisions, then the effort typi-cally will not find much echo or sustainability.What is also often lacking is rigorous analysisof the effectiveness of educational innovations.We expect projects to demonstrate the effec-tiveness of methodology, curriculum, or teachertraining reforms, but most of them fail to estab-lish the baseline facts about how effective theirinnovations actually are. Easterly (2006) callsfor donor-funded efforts to do more systematicexperimentation and evaluation to establishfact-based foundations from which to leveragebroader implementation and impact. The twoexamples below show how analysis can workfrom both the demand and supply sides.

Example 4. Demand-driven analysis in SouthAfrica. A case of demand-driven analysis, anddonor support for it, comes from South Africa.(This case also illustrates effective donor cre-ation of reform support infrastructure.) As iswell known, education funding in South Africaduring apartheid was extremely and purposelyinequitable. Not surprising, the new govern-ment, even before it came to power, wanted tochange this. In the mid-1990s, USAID providedtwo major forms of support to reform the fund-ing system. First, was analytical support torevamp the school funding and resourcing for-mulas. Second, was support to the training andcapacity development of (mostly) governmentofficials, some NGO officials, and some aca-demics, in the nature of the financial reforms,in both theory and implementation practice.The training was well coordinated with the poli-cy development, in the sense that the assis-tance providers working on the policy alsodeveloped the training materials and providedthe training itself. The capacity development

took the form of patient, long-term collaborativework between technical assistance providersand local counterparts over many years.Several years after the original reforms weredesigned, it was clear that they had some unin-tended consequences. In 2003 the nationalgovernment issued a sophisticated reanalysisof the situation and proposed certain “reformsof the reform,” such as requiring that povertytargeting be based on a national definition.This reanalysis did not require intense donorsupport, as the ability to carry out this analysishad been internalized to a large degree. In thiscase, the training, even though extremely inten-sive, was an add-on. The provider of technicalassistance, and the provider’s counterparts,were aware that demands of day-to-day activi-ties and problem solving were preventing knowl-edge transfer. Counterparts and the technicalassistance (TA) provider started an informal“club” on Friday afternoons to devote voluntarytime to capacity building. After a year, thiseffort was transmuted into a formal course,and this course—based as it was on a realneed, having been informally field-tested, andbeing offered by the same technical assistanceprovider who had worked on the policies (train-ing thus being tightly linked to need to imple-ment policy)—was quite successful. Beingaware of the environment, having a well-stocked toolbox, and having donors willing tobe flexible were all keys to success (based onpersonal knowledge of the authors; seeCrouch, 2006).

Example 5. Data and analysis to drive sectorvisioning in Pakistan. A visioning process anduse of a projection model in Pakistan show howdifferent kinds of analyses can support areform. A major policy problem in Pakistan isthe general lack of a shared vision of what anaffordable but desirable education systemmight look like. Accordingly, policies are bothunaligned and unrealistic, reforms do not worktoward a common end, and donor support proj-ects do not contribute much toward reform. Thereform support work undertaken to deal withthis situation unfolded in three phases. First,efforts were made to facilitate stakeholders’understanding of the need for a vision: to gen-

Education

Reform

Support

Today

29erate demand for one, given the predominantview that their existing policy frameworksamounted to a vision. Through informed discus-sions with stakeholders from around the coun-try, this understanding was facilitated, but notwithout tension. Stakeholders first tended tothink they had a vision, but the fact that it wasnot a practical one was not clear. This led tothe realization that the process could benefitfrom a costing model to support the entirevisioning effort. Second, these same stake-holders were then asked to delineate theirdepiction of the present-day system with a par-ticular emphasis on planning, management,professional development, and finance flows(knowing that these emphases would manifestissues germane to standards, performanceappraisal, accountability, and career ladders).With this understanding, these workshop partic-ipants were then asked to envision a system inwhich all of the problems and issues they justidentified no longer existed and to delineatethose elements of the system that were crucialto its eventual realization. For example, from agoal of well-trained teachers in every class-room, the discussion addressed the mecha-nisms needed to ensure well-trained teachersin every classroom, say career ladders to bringabout demand-driven professional developmentand accountability systems. This became thematerial for an emerging vision that was putforth in a discussion document distributed forcomment around the country. Feedback on thisdiscussion document was gathered in writtenform and in a series of stakeholder workshopsthat furthered the emerging vision. Here theprocess faced a key choice: how to elicit feed-back from thousands of people, yet keep itboth focused and manageable. A decision wasmade to present the emerging vision inchunks—classrooms, schools/communities,districts, etc.—and ask a number of key ques-tions about each chunk, the answers to whichwould help flesh out critical elements of thevision (and engender widespread ownership ofthose elements). Finally, the cost of the emerg-ing vision had to be assessed. It is one thingto envision a well-equipped classroom with aqualified teacher in it; it is quite another todetermine just what well-equipped classrooms

and qualified teachers cost and how this addsup to an affordable or unaffordable educationsector budget. A second round of workshopsfocused on cost. With the use of a computer-ized cost-projection model, the costs of variousscenarios were assessed, and an optimal solu-tion was finally realized vis-à-vis enrollmentgrowth, cost-saving measures, and quality-enhancing reforms (personal communication,Hank Healey, June 8, 2006.)

DIALOGUE AND COMMUNICATION

Perhaps the most important part of the ERSapproach is fostering communication, debate,and dialogue. As can be seen in the data andanalysis examples just presented, communica-tion and dialogue are often integral parts of thedata and analysis activities. In fact, communi-cation and dialogue arewhat create demand for “Perhaps the most

important part of

the ERS approach

fostering communic

debate, and dialog

is

ation,

ue.”

analysis and contribute tosustainable results.

Donor support to dialogueprocesses can take atleast two forms. A rela-tively common one is for the donor itself toengage in dialogue, as an actual actor. This isuseful, and sometimes is a good way to putsome issues quickly and proficiently on thedebate agenda, but is not really ideal. Worse,in some of these situations, “dialogue”becomes an imposition on a task manager’schecklist, and the donor will consider itself tohave “dialogued” if it holds a semipublic semi-nar or workshop. Alternatively, the “dialogue”may be just a discussion of policy conditionson a loan or grant. This sort of approach doesnot take to heart the notion that debate andseeking solutions are keys to sustainable andendogenous policy reform. It continues toassume that counterpart governments arerational, tend to arrive at “rational-comprehen-sive” solutions based on discrete bits of evi-dence, and such solutions are socially optimaland can lead to development. To put it inEasterly’s terms, this is a “planner’s” solution(Easterly, 2006). As noted, on occasion thiskind of approach is useful, because it can,

Edu

cation

Ref

orm

Sup

port

Toda

y

30indeed, put issues on the table quickly and pro-ficiently, but it is not the best way to proceed.

An alternative, which the ERS approachdefends, is to foster more genuine dialogue,where debate is encouraged, opposing actorsare brought to the table, and the donor eithersimply finances the dialogue process or trulysees itself as just one more actor. This doesnot mean that the donor has to shy away fromespousing a particular position or defending aparticular approach, and even engaging inintense social marketing of the approach. But itdoes mean that the donor has to see itself asfallible, abandon a “planner” mentality, andadopt a “debater” or “searcher” mentality. It isimportant to note that this ERS position is notbased on an emotional or politically correctviewpoint that suggests donors have to be“nice,” but, instead, on the empirically support-ed sense that large-scale solutions based onsupposedly rational analysis often do not work,and that sustainable, implementable solutionsthat do work (particularly for implementation-intensive sectors, such as social service deliv-ery) emerge from of a process of competitionas well as a process, searching or “feelingone’s way” (Easterly, 2006). Fostering this kindof debate is closely linked to the need to cre-ate an ongoing reform support infrastructure,as discussed below.

The four examples of policy dialogue presentedhere below highlight how strategically chosensupports, with astute attention to the politicsof the context for reform, can advance nationalprocesses of dialogue and shape the wayreform issues are defined and taken on. Inaddition to these kinds of dialogue activities,communication more broadly defined is alsointegral to sound ERS.

Example 6. Continental policy dialogue in LatinAmerica and the Caribbean. PREAL is a jointactivity of the Inter-American Dialogue inWashington and the Corporation forDevelopment Research (CINDE) in Santiago,

Chile, seeking, as its Spanish initials indicate2,to promote education reform in Latin America.PREAL promotes reform not only throughdebate, but also by tracking and reporting onprogress through its well-known “educationreport cards” on individual countries and coun-try groups, and by disseminating best prac-tices. PREAL’s approach relies on involving civilsociety and sociopolitical leaders chosen fortheir personal qualities and history, rather thanfor their official position.

Over more than a decade, PREAL has grownfrom a relatively limited effort to promote policydebate in six countries to a region-wide initia-tive aimed at building a broad and active con-stituency for education reform. It has becomethe leading non-government voice on educationin Latin America and a strong advocate forinvolving civil society leaders in the work ofeducation reform. Much of its work has beenexperimental, including pioneering the conceptof country education report cards in LatinAmerica, developing user-friendly publicationsthat address key policy issues, convincingrespected leaders from outside the educationsector to work for education reform, helpingnational advocacy groups engage local leadersin discussions on education policy, and estab-lishing region-wide, issue-specific workinggroups that identify and debate policy implica-tions in areas such as national standards,assessment, the teaching profession, andschool autonomy.

PREAL is an excellent example of a donor activ-ity that promotes real debate and discussion,while not shying away from its own advocacyand points of view. For example, PREAL pro-motes the notion that standards and assess-ment are key to educational improvement.While it would be foolhardy to attribute thecausality behind continental trends to a singleinstitution, the reality is that the drive for meas-urement and assessment in Latin America hasincreased tremendously in the last 10 years.While PREAL has been only one of the voices

2 Programa de Promoción de la Reforma Educativa en América Latina y el Caribe, or “Program for the Promotion of Education Reform in LatinAmerica and the Caribbean” in direct translation (note that the English name is not a direct translation of the Spanish name).

Education

Reform

Support

Today

31pressing this drive forward, it has been per-haps the most insistent, consistent, and profi-cient voice, at least in the countries where ithas been most active. Dissemination of PREALpoints of view takes place through an extensivepublications program, national and internationalseminars, its web page, study tours, and taskforces and working groups that encourage inter-national collaboration (created from personalcommunication, Jeffrey Puryear, Director,PREAL, June 2 2006, and the authors’ personalknowledge of the PREAL process; for moreinformation see www.preal.org and www.thedia-logue.org).

Example 7. Policy dialogue around finance andother issues in Guatemala. After the electionsin 2003, and following the failure of the previ-ous administration to deliver on high expecta-tions, USAID was committed to supporting thenew government in substantive reforms, andparticularly the Minister of Education, a dynam-ic leader from the private sector. USAID identi-fied two key issues—low levels of investmentin the social sectors (education and health) andpolicy instability from one administration to thenext. A bridging, one-year project in 2004 wasintended to support a national dialogue amongthe full spectrum of political actors inGuatemala and to develop a broad-based com-mitment to increased investment. The initialdesign sought to use the project as a neutralentity to convene all stakeholders in an empiri-cal assessment of the problem and an explo-ration of solutions that would have supportfrom all elements of society. This effort toimplement the core philosophy of the ERSapproach immediately ran into resistance fromthe Ministry, which was concerned about inclu-sion of political opponents and saw theMinistry of Education (MOE) as the only legiti-mate convener of such a dialogue. The projectwas modified to support a national visioningprocess, led by the Ministry but including arange of institutions. This decision had signifi-cant implications for the project: it establisheda critical enabling relationship with the keystakeholder, but at the same time, the donorand project lost control of the schedule andstructure of the process. The project became a

technical secretariat for a national policy dia-logue, conducting technical analyses andresearch in response to working group con-cerns. The USAID approach uses an explicitanalysis and communication methodology towork with stakeholders, identify audiences,identify behaviors to be targeted and the fac-tors that affect those behaviors, and developcommunication strategies that respond tothose factors (see USAID, 2006). The resultwas a country-led process with a high degree ofownership, but with implementation delays fromthe USAID project management standpoint.

Success of the project led to a Ministry requestfor follow-on support for social investment poli-cy dialogue, with two components—improvingthe financial efficiency of the education sector,and expanding the broader societal dialogue. Inthis stage of USAID support, the project hasfaced greater challenges. The national politicalenvironment began to focus on future elec-tions, the Minister’s leadership was challengedby opponents (and congress), and resistance tobroad-based dialogue (and perceived empower-ment of political opponents) increased. Theproject is currently working to balance the inter-ests of maintaining key ownership and supportof the Ministry with the longer-term USAID inter-est in assuring sustainability of the reformsthrough a broad-based social compact (person-al communication, John Gillies, June 2006).

Example 8. Postwar dialogue in El Salvador.Through a negotiated settlement, El Salvadoremerged in the early 1990s from a long periodof civil war. During the civil war, education hadbeen neglected, though interesting experimentswere carried out, particularly in the guerrilla-controlled areas. These experiments tended toemphasize self-help, community involvement inschooling, and the spontaneous creation of“communities of practice” among the teachersin the guerrilla-held zones. While not necessari-ly providing higher-quality education than thatoffered by traditional schools, both parties sawthese schools as a promising way to increaseenrollment and involve communities. The newright-wing government called on USAID to sup-port a process of dialogue that involved all key

Edu

cation

Ref

orm

Sup

port

Toda

y

32segments of society, including those associat-ed with the left. Ten integrated teams wereformed, each of which was to draft a report onone of 10 critical technical areas in the educa-tion sector. An advisory committee (mini-parlia-ment) of about 50 representatives from some30 organizations was created to vet thereports. This committee met with the technicalteams every week for 12 weeks. The processsucceeded in generating an approach to reformthat was legitimate across a fairly broad spec-trum of society. Reasons for the success of theprocess included a) a focus on the feasibility ofsolutions as opposed to fundamental values; ba focus on process itself and the use of goodprocess management based on clear rules andprocedures; and c) selection of representativesof various organizations for their personal char-acteristics, such as the ability to stay focusedon solutions. Agreements were reached thatlaid the foundation for an approach to sus-tained reform based on community-controlledschools (based on Schiefelbein, 2006).

Example 9. Political maneuvering to supportreform in Brazil. In the mid-1990s, Brazil insti-tuted some of the most far-reaching educationreforms undertaken in developing countries inrecent decades. These reforms have had majorimpact, particularly on access. The primarycompletion rate went up by 20 percentagepoints between the mid-1990s and 2001, andsecondary enrollment improved greatly.3 Therewere many large-scale changes within the edu-cation sector, but certain financial reformswere key to driving improvements in accessand equity. These financial reforms had twoparticular features. First, the funding wasreformed so as to be driven by formulas thatrewarded enrollment, as opposed to ad hoc ortradition-based financial allocations, or, worse,allocations based on each municipality’s ownrevenue-raising capacity from its own sources(as was partially the case before the reforms).Second, the funding greatly reduced under-spending in poor regions by creating state-levelequalization funds that pooled all within-statefunding, thus preventing poor municipalities

)

from under-spending, and by using federalfunds to create a floor of per-student spendingwithin states and municipalities. At the schoollevel, transfers were set by an enrollment-driv-en formula, but schools had to set up parent-teacher associations (PTAs) and bank accountsbefore they received funding. Thus, funding wasused to drive governance changes at theschool level. The number of PTAs increasedfrom 11,000 to 60,000, so incentives, itseems, do work.

These reforms were not easy and required con-siderable political maneuvering and mobilizationof public opinion. At the parliamentary level,the then-Minister pushed the financial reformsthrough largely by stealth rather than by debateand dialogue; for example, it is not clear thatlegislators were totally aware of the implica-tions of what they were approving. This gotthings done quickly, but some analysts think itmade subsequent changes more difficult. Itshould also be noted that financial reformstend to be more cut-and-dried than, say,reforms in teaching or governance processes.Reforms that are less implementation-inten-sive, and mostly require only a few decisions,may be easier to carry out via quick politicsand maneuvering than via open dialogue.However, in many of the reforms, the Ministerdid mobilize public opinion and debate, basedon data analysis, adroitly. He increased thefunding and capability of the data systems andused the data to draw attention to inequalityand other issues in dialogue with the public.The Ministry began to display the newlyacquired data and assessment information inuseful formats that created a new transparencywith civil society. The Minister participated inradio talk shows, created a publicity campaignentitled “Wake Up Brazil! It’s Time for School,”and used a thirty minute slot during the federalgovernment’s nightly newscast to relay educa-tion news and respond to citizen letters. Heengaged and empowered civil society to holdtheir local and state governments accountablefor quality education, even creating a Federal 1-800 line that citizens and PTAs could use to

3 While the key financial reforms were in primary education, these had a ripple effect on secondary education.

Education

Reform

Support

Today

33report local misuse of education policy. Using acombination of political acumen and mobiliza-tion of opinion and dialogue, the Minister andhis team were able to pursue reforms with pro-found and far-reaching impacts unusual in thedeveloping, or, indeed, developed world (basedon personal communication, Audrey-marieSchuh Moore, September 2006; various pre-sentations by Paulo Renato Souza; and discus-sions with Brazilian analysts).

BUILDING REFORM SUPPORT INFRASTRUCTURE

All of the activities discussed above can helpsupport a given reform, or a given package ofreforms, but they do not institutionalize a country’s capacity to debate and design reformin an ongoing way. Designing a given package

of reforms is, of course, ultimate goal of

or assistance and

acity building shoul

to help create the

astructure whereby

a-based analysis,

cy dialogue, and

tegic communicatio

be ongoing and

ptive…”

“The

don

cap

be

infr

dat

poli

stra

can

ada

better than nothing.But designing onepackage of reforms isdstill too much like a“planner’s” solutionand not enough like a“searching” solution.The ultimate goal ofdonor assistance and

n capacity buildingshould be to help createthe infrastructurewhereby data-based

analysis, policy dialogue, and strategic communication can be ongoing and adaptive,so that the system can continue to carry outgradual improvements creatively. This point ofview is based on the notion that policy reformitself has to be sustainable and on the awareness that every policy reform raises new problems to be solved.

Activities that might help institutionalize reformsupport can be divided in two groups: those oriented toward government and those orientedtoward institutions outside government.

The advantage of supporting government unitsto carry out ongoing policy reform is that suchunits have access to a tax base, so, in principle,they might be sustainable. The conundrum, ofcourse, is that if there was real governmental

interest in supporting proper analysis, thedonor intervention would not be needed tobegin with. Thus, the very need for donor supportor intervention is itself prima facie evidence ofgovernmental failure, or at least governmentalweakness, and suggests that the donor inter-vention may well not be sustained after directsupport ceases. Government actors fail to carryout proper analysis not primarily because theyare technically weak, but because analysis isnot a priority or results in too many inconvenienttruths. The observed incapacity to carry outanalysis and proper debate in the educationsector is, to a large degree, a consequence oflack of political interest in objective analysis,and not due to an inherent lack of access tobrainpower or skill. And the common donor reasoning that “once governments see whatgood things data and analysis can do, they willbe enthused and will sustain the effort” is usually wrong, as has been proven by the constant failure of governments to sustainEMIS and policy units. Unsatisfied with the lackof interest and sustainability of data and policyunits in government, the donors have turned toNGOs or think tanks that work in the publicinterest and have gone beyond service deliveryand into policy analysis and debate.

But this approach is also problematic, becauseprivate willingness to support policy reformunits, beyond the duration of the donor’s funding,is also hard to find. Weak states usually coexistwith weak private sectors and civil societies,where the notion of policy-oriented and sustainedphilanthropy (such as endowing a think tank) isforeign indeed. Service delivery or corporatesocial responsibility kinds of philanthropy canbe found, but policy-based philanthropy is rarein poor countries. Generally, the poorer thecountry, where good policy is most sorely needed, the weaker the state, and the weakerthe private sector and civil society at providingfunds to nongovernmental policy reform units.Is there no way out?

Partly the answer is, indeed, no. Donors maysimply be called to support this kind of activityfor a long time in the weaker countries. It ismost likely wishful thinking to expect policy

Edu

cation

Ref

orm

Sup

port

Toda

y

34activities (government or otherwise) to becomeself-sustaining in most very poor countries inthe short term. (Note that local consultingfirms, which carry out studies for a fee, orcarry out advocacy processes also for a fee,can indeed become sustainable as institutions,but the source of financing is still, typically,donor money—donors continue to underwritepolicy work, even if they do so through contractsrather than through institutional support grants.)

But countries vary a great deal, and one canfind privately based policy reform work even inrelatively poor, small countries, and no activityat all in bigger, richer countries. Ultimately,deciding whether to support government or non-government institutions as “reformist” agentshas to rely on a case-by-case analysis of eachsituation. If government is receptive and truly reform-minded, most likely investing ingovernment units is a good idea, as there issome hope of sustainability through a tax base.If government is not very receptive, and the private sector is enlightened enough to supportpolicy reform ideas that go beyond the usualcorporate philanthropy and problem-fixingapproaches, then donor partnerships with theprivate sector can be fruitful. If neither conditioncan be found, donors may find themselves inthe situation of directly funding NGOs that cantake on research and advocacy roles, and thismay need to be carried out for a long time. Ifthere is sufficient funding, some mix of theseapproaches is, of course, recommended. NGOanalytical and data-based pressure can keepgovernment analytical units honest, for example,so even where governments are receptive andreform-minded, supporting NGOs makes sense.The basis of good decisions is competition of ideas.

Example 10. Institutionalization of reform sup-port via private sector in El Salvador, weak pri-vate sector support in Peru. In El Salvador,FEPADE (Fundación Empresarial para elDesarrollo Educativo—EntrepreneurialFoundation for Educational Development), creat-ed in 1987, got much of its early support fromUSAID and tended to engage only in very tradi-tional philanthropy and service delivery or nar-

row policy issues (around vocational educationissues). It has evolved, however, into a remark-ably multifaceted institution with some of themost visionary policy activities of any privateeducation group in Latin America. FEPADEdivides its activities into “point” actions (tradi-tional philanthropy and privately funded servicedelivery, such as adopt-a-school schemes,donations, corporate social responsibility); “pro-grammatic” actions (such as supporting school-level, public sector pilot projects that teachschools how to implement existing policyreforms); and “strategic” interventions (such asdirect policy debate and analysis, mobilizationof private sector opinion on reform issues, sup-port to the public sector on policy analysis; seeFreund, 2005). In contrast, in a much biggercountry, Peru, private sector and civil societysupport to public education policy rarely tran-scends philanthropic or corporate socialresponsibility interventions, though things maybe changing recently. Thus, it is very hard topredict where the private sector might pitch inwith ongoing support to public educationdebates. Donors can test the waters by sup-porting initial activities (say, via matchinggrants) that may be in the service deliveryarea, and then gradually encourage a shifttoward policy-oriented activities with full privatesupport.

In other situations, direct support of govern-ment and NGO policy reform infrastructures isa worthwhile investment.

Example 11. “Distributed” institutionalizationor reform support capacity in South Africa. InSouth Africa, for example, USAID and otherdonors, such as the Netherlands, invested incapacity development of governmental officials,academic actors, and NGO actors, thus creat-ing a robust reform support infrastructure overa period of about 10 years. After providing poli-cy advice assistance on post-apartheid finan-cial reforms, USAID supported capacity devel-opment in implementing these reforms at twolevels. First, at the provincial head office level,USAID helped train government officials tounderstand and be able to implement some ofthe key reforms. Second, USAID helped school-

Education

Reform

Support

Today

35and district-level staff understand and applymany of the reforms. Other donors providedsimilar assistance; in fact, often more than onedonor assisted the same units. Both USAIDand the Netherlands embassy supported theEducation Foundation, an NGO that specializedin policy debate and analysis and providingdata digests, among other services. After theend of apartheid, many of these NGOs col-lapsed, as donor funding turned toward govern-ment itself. But, even though some eventuallyceased to exist, these NGOs provided usefulservices for 10 to 15 years or so, helping seethe country through a remarkable transforma-tion. In that sense their legacy and contributionhas certainly been sustained, even if as institu-tions they may not have been. Others havemade a successful transition to contract-basedproviders of analytical services to governmentor the private sector, as opposed to being sup-ported by donors’ institutional grants. WhileSouth Africa may be unique in the sharpness ofthe contradictions it faced and exemplified,much of the same sorts of issues can be foundin all countries, albeit in a less dramatic way(based on personal knowledge of the authors).

The need to institutionalize reform support isnot peculiar to developing countries. The UnitedStates, for example, has a similar need. Inspite of the fact that routine management ofeducation usually works better in the UnitedStates than it does in most developing coun-tries, countries like the United States have sim-ilar problems in sustaining reform. The follow-ing example documents efforts to develop areform support infrastructure in a domesticU.S. context.

Example 12. Institutionalization of reform sup-port in the United States: San Juan andWashington, D.C. The San Juan MetropolitanAlliance for Education (SJMAE) in Puerto Rico isa reform support network that was initiallyseeded and supported by the Ford Foundation.SJMAE was founded in 1994 and initiallyworked with the 11 schools in the Cataño dis-trict in San Juan. The Alliance provided school-level support for reforms in classroom practiceand school organization and management, but

linked those supports to efforts to reshapepolicies regarding how district-level supportneeded to be structured. After working toestablish a record of success in Cataño, theSJMAE knew it had to turn its attention to theissues of scale and sustainability: how couldthey affect education on a broader scale inPuerto Rico, and what policy innovations couldensure the sustainability of what was happen-ing in Cataño? The Alliance’s leaders woulddialogue regularly with policy makers, but amore systematic assessment of the policyimplications of their work was called for. Astudy conducted in 2001–2002 drew lessonsfrom the Cataño experience for island-wide poli-cy concerning the role of district offices in sup-porting school improvement. That study servedas the basis for facilitated dialogue among allPuerto Rico’s superintendents and theDepartment of Education. As a result, thePuerto Rican Department of Education and theW. K. Kellogg Foundation committed sufficientresources to the Alliance’s work to allow theprogram to expand to six other districts. Basedon its success at linking school-level interven-tions to policy reform, the Alliance formed theInstitute for Education Policy and CommunityDevelopment in 2005. The Institute evaluateseffective grassroots initiatives to discover theirpolicy and institutional implications for thePuerto Rican education system. A private uni-versity, several community-based organizations,the College Board of Puerto Rico, the PuertoRican Community Foundation, and theConservatory of Music and Museum of Art con-stitute a reform support infrastructure that hasgrown out of the SJMAE’s original work support-ing school-level change and has evolved toencompass a variety of strategies for promot-ing, supporting, and sustaining educationreform in Puerto Rico (based on personalknowledge of the authors).

In the spring of 1998 a group of teachers, par-ents, and community activists started meetingto explore ways to work collaboratively on edu-cation reform in the District of Columbia. With aplanning grant from the Ford Foundation,District Community Voices Organized andInformed for Change in Education (DC VOICE)

Edu

cation

Ref

orm

Sup

port

Toda

y

36came together as a reform support network ofnumerous organizations and individuals workingto improve education in Washington, D.C. Theysaw a more informed and better organized pub-lic as essential to generating greater demandand accountability for improved education inthe city. They also realized that many organiza-tions and activists already worked on educationissues in Washington, so that better collabora-tion among those entities would be most pro-ductive (as opposed to one more NGOs com-peting for funding). As part of how it promotescollaboration and ensures the relevance of itsreform support work, DC VOICE annually bringstogether its numerous members and partnersto identify key issues the collaborative networkwill take on or support. And it carries out itsdata, analysis, communication, and community-organizing work through its various con-stituents. DC VOICE attracts funding from vari-ous foundation and corporate sources to fulfillits mission of promoting a more informed andactive public voice for quality education (basedon personal knowledge of the authors).

PROJECTIZING DATA, ANALYSIS, DIALOGUE, AND

SUPPORT TO A REFORM SUPPORT

INFRASTRUCTURE

The institutions or groups that implement thesevarious activities can form a formal or informalreform support infrastructure. If donors have aconscious and planned approach to supportinga broad front of reform agendas through anarray of actors among whom relationships areactively promoted, then the Reform SupportInfrastructure that emerges will tend to bemore formal and self-aware.

A particularly effective way to “projectize”reform support is to set up projects that havesome school-level components that consciouslytry to implement, on the ground, the policiesthat the policy-level components are helping todevelop. In this way, policy-level projects can“walk down the ladder” from policy reform toschool implementation. Thus, for example, adonor project might have a policy-level compo-nent that could design school financing policiesto discriminate in favor of the poor (by creating

formulas in which poor students carry a heavierweight), while a school-level component couldbe training district and school officials on howto implement these formulas. However, itshould be noted that this ideal approach (twin-ning the same policy-level and school-levelissues in continued donor activity) can be doneonly in a country where there is considerablecontinuity of donor approach. The reason forthis is that there is often a lag of two to threeyears between the point at which a donor canassist with a policy development and the pointat which a counterpart government begins toseriously try to implement the policy in ques-tion and thus requires school-level piloting ofthe policy. Alternatively, if such continuity doesnot exist, a project can support implementationof policies that were set in place two to threeyears ago, while taking advantage of the good-will this generates to put in place advisory anddialogue services that assist the governmentwith ongoing issues that may not require imple-mentation for another two to three years.

Furthermore, school-level interventions cancontribute to reform support by documentingthe conditions needed to assure the success ofimprovement initiatives. A project can thenwork with education sector authorities to definethe changes in the policy and institutional envi-ronments that are needed if those conditionsfor success are going to prevail. In this manner,a fairly traditional school-level pilot project canidentify reform issues—changes in the law, ingovernance policy, in administrative structures,in relationships—around which data, analysis,and communication activities could be mobi-lized. This how a school-level project can walkup the ladder to policy-level reforms (which ERSactivities could then help address).

Education

Reform

Support

Today

375.Limitations (and Advantages) of the ERS ApproachIn this section we recognize that there are limitations to the ERS approach and discuss ways toovercome them. Evidently, though, in some countries the limitations are harder to overcome than in others. The table below shows situations where the limitations are likely to be very difficult to overcome and where, therefore, more traditional “input” projects are likely to remain the optimaltype of donor assistance, or where serious adaptations may be needed for an ERS approach to work.

Type of Country Problems Rendering ERS Reasons and Possible ReactionsInappropriate or Less Likely to Succeed

Emerging from disaster or conflict Need for immediate relief; policy reform is a long-run process. Thereis likely to be impatience with long-term approaches. Nonetheless,countries in transition are often more amenable to new polices. ERS could be a fruitful way to invest in such situations, but theinvestment would have to be proportionately small.

Government institutions that are The best approach most likely is, again, an inputs-based project,nontransparent and hostile to open since policy-based work will typically not progress in such settings.discussion, evidence-based policy Presumably there is a foreign policy reason to be involved in such asuggestions, or innovation in country, since countries where these conditions prevail are not likelygeneral, in a context of generally to benefit from assistance oriented toward economic and socialpoor delivery of services and development. On the other hand, only government institutions aremanagement hostile to dialogue, but there is a reasonable civil society or

community life, and government will not prevent donors from workingat that level, then a good place to start is by reinforcing civil society,including parents’ groups at the “micro” level.

Counterpart government strong A policy-based approach can still work in such settings, but the donorand resilient, but has own ideas, agency has to decide to “trust” the overall direction of policy andin context of good or improving provide advisors and support to help implement overall policy. Theregovernance and service delivery— will be reduced scope for supporting policy reform ideas that runcounterparts therefore not very contrary to government’s ideas in particular areas, since overall open to discussion about policy policy and service delivery is improving. Such countries are often indirections; tendency may be specific a strong bargaining position with respect to donors, particularly if, into one ministry, depending on team addition to improving governance overall, they are of foreign policy in place or geopolitical importance to the donor country’s government.

In addition to the context-related caveats mentioned above, there are disadvantages to the ERSapproach as well as the advantages. We raise these disadvantages and advantages for three reasons. First, because such awareness can prevent the disappointment that might ensue if one isexcessively optimistic at the outset. Second, and more important, because being aware of theproblems can help mitigate them. Third, because colleagues and leaders in donor agencies may beaware of the problems but may not be as aware of the advantages. In short, knowing the limitationsand advantages of the approach is an operational consideration, because it can affect how oneimplements ERS, and a “salesmanship” consideration, because it affects how one defends theapproach to one’s colleagues in a donor agency.

Edu

cation

Ref

orm

Sup

port

Toda

y

38The table below shows the limitations and, in each case, suggestions for dealing with them. It alsoshows the inherent advantages. Note that while the two first columns can be read across fromeach other (the concepts in the same rows match), the last column cannot.

Limitations and Advantages of an ERS Approach

Limitations of an Means to overcome Advantages of an ERS Approach Limitations ERS Approach

The process is not quick— Redefine outcomes so they are similar to It is the most highly leveragedit may take some time to those in a governance, citizen participation, form of donor input into asee results. or local development project—seek allies country’s development

and models in democracy and governance, strategy. For the few hundredlocal governance, or local economic thousand dollars of investmentdevelopment sectors.4 needed to provide advice

about and advocacy for a policy change (e.g., altering a school finance formula sothat it benefits the poor), adonor can “leverage” millionsor billions of dollars of thehost country’s own spending

The process is not easy—there are fewer blueprintsthan for standard, input-enhancing projects.

Select the most experienced implementers.Select an implementation approach thatallows the funding provider to seek themost experienced guidance before singlingout a particular implementer.

It is easy to get misled and Invest a lot of time in the initial foray intoin favor of the poor, or infavor of rural areas.

“played with” by various the policy issues, in getting to know theinterest groups. actors and their inter-relationships before

committing to guides through the local It is likely to lead to the mostpolitics and policy circles. sustainable and replicable

results, because the hostgovernment’s systems andapproaches change. Theresults are “owned” by thenational authorities and aredemand-driven.

To generate reforms, citizenshave to be mobilized toexpress demands, for example, for more qualityeducation, or school nutrition.Getting policy change throughcitizen demand will fosterincreased accountability inimplementation, because citizens will learn to expresstheir demands and will con-

The process is risky. If one defines the impact as an actual policy change, particularly a policy changethat goes in a particular direction (e.g., a decision to decentralize school purchases), there is a significant chance of having no output.

Many mechanisms can be used to minimizethis risk. One can redefine the impactmeasure as learning how to analyze,debate, and make policy. Alternatively, onecan accept the risk as inherent in theprocess and spread the risk by backingmany policy changes, not just a single one.One can ally oneself with other donors topressure for the same change; somedonors may use conditionality, but condi-tionality may lead to superficial adoption ofpolicy changes. Finally, one can make thereform activity only a small portion of anoverall project that also includes traditionalinput and process activities such asteacher training or textbook provision.

Host-country governments, One may need to accept that the tinue to use this knowledgeparticularly in poorer approach is not suitable in all countries. beyond the policy changecountries, may not under- (See section on preconditions for success.) process and into the policystand the approach and Alternatively, the policy-based approach implementation process.may prefer traditional may be only one part of a larger project,projects of a more most of whose components are of a more“philanthropic” nature. traditional kind.They may oppose anapproach that tries toenhance the transparency of the policy process andempower citizens.

4 For suggestions on this, see USAID (2003). For a description of a combined education and governance program, in Ghana, see the descriptionof the Strengthened Decentralized Local Governance project, at http://www.usaid.gov/policy/budget/cbj2005/afr/pdf/641-005.pdf. For adescription of typical governance indicators, see USAID (1998). Note in particular sections 5 and 6 on civil society and governance, respectively. These are easy to adapt to the education sector.

Education

Reform

Support

Today

39An important consideration below is the possibility of redefining education goals as governance goals. At least two prominentdonors, namely USAID and the U.K.’sDepartment for International Development(DfID), are making governance the centerpieceof their overall development approach, and are already defining projects whose sectoralemphases are governance-related. Using governance goals even in sectoral activities is thus likely to gain increasing legitimacy and may be a structural trend.

To summarize the table above, one can saythat the advantages and disadvantages of theERS approach, as opposed to a traditionalschool-based approach, are the same as theadvantages and disadvantages of truly develop-mental activities as opposed to philanthropy orgiveaways. This is the well-known trade-offbetween immediate impact or relief on the onehand and sustainability or replication on theother. Naturally, there is a whole spectrum inthis matter, not just a trade-off between twopolar opposites. “Giving away” textbooks ormoney is the least developmental possibility,but has the most immediate impact on relief;providing funding for textbook writing or text-book procurement management advisors movesin the direction of sustainability; and reformingbudgets and policies (e.g., by ensuring thatteacher salaries do not eat up 90 percent ofthe budget), so that countries provide them-selves with the expertise on how to write text-books, is the most leveraged and sustainableapproach.

The “Advantages” column in the table to theleft notes that ERS—or any policy-basedapproach—is the most highly leveraged donoractivity available. This suggests a simple way toavoid the risk that one may simply have noimpact, or very little impact. What allows a policy-based approach to bring so much lever-age is that it is inexpensive—the input beingpurchased might be one or two advisors forone or two person-years and some support fora policy dialogue process. This usually costsvery little, compared to massive school build-ing, teacher training, or almost any other commodity or training-based assistance. Thelow cost of policy inputs means that one canattempt to change more than one policy, thusspreading one’s risk, or that one can have arelatively small policy-based component in aproject that has larger, more traditional input-or process-based components. In that sensethe policy-based component of a set of activitiesis similar to one’s more speculative investmentsin a balanced portfolio: high risk, yet high payoff. Thus, a donor might invest in either avariety of such ventures, spreading the risk, orin some high-risk ventures but in more secureones as well.

Edu

cation

Ref

orm

Sup

port

Toda

y

40

Education

Reform

Support

Today

416.Justifying an ERS ApproachThus far the ERS approach has been justifiedfrom a “developmental” or “theoretical” pointof view. Given what we know about develop-ment, supporting high-leverage reform activi-ties, as opposed to simply donating commodi-ties or undertaking pilot experiments, is the“right” thing to do. And, given what we knowabout reform, it is hard to imagine designingreform support interventions that do not takeinto account the politics and institutional con-text in the ways described above. However, werecognize that within international developmentagencies, operational staff are challenged to“justify” their programs and project designs inmore operational or bureaucratic terms.Precisely because of the political nature ofERS, and because the management of political-ly driven processes requires greater flexibility,adaptability, and timely, strategic decisions,development agency staff will need to lobby forthe room to support the kinds of activities wehave described (it is ironic that they will needto clear space in their own agencies for thiswork). Confronted by the press for fast results,donor agency officials might well ask them-selves how reform support activities can bejustified bureaucratically in the agencies wherethey work. In fact, we have voiced this specificconcern.5 This section therefore looks at theparticular issue of how agency staff can selldecision makers on using resources for reformsupport activities.

In explaining or justifying reform support projectactivities, or a reform support component with-in a more traditional approach, there are twobasic choices. One is to try to justify theapproach in terms of the most recent trends inone’s own donor agency. Because these trendscan change quickly, this approach has its dan-gers. But this approach can also be of someuse, particularly if it can be demonstrated that

more recent trends are not really inconsistentwith longer-term trends. Another approach is tofocus directly on longer-term trends, ideally notjust in one’s own agency, but in donor agenciesin general. These longer-term trends in thinkingare fairly sustained within given agencies andare surprisingly similar across agencies thatwould otherwise appear to have differentapproaches.

Starting with the longer-term trends, one cannote that USAID’s most current education poli-cy statement (USAID, 2005) still makes sys-temic reform USAID’s premier approach toimproving education, as did agency policy state-ments from 10 years ago:

Systemic reform is the foundation of USAID’swork to strengthen education systems. Overthe past decade, the Agency has developedand refined a sector-support approach thatemphasizes support to sector reforms andcapacity-building efforts developed and led byhost-country governments. This approachhelps ensure sustainability and focuses onwhat children are actually learning in theclassroom or nonformal educational settings.

USAID missions work with host governmentsto adopt and implement policies that increaseeducational efficiency, promote equitableaccess, and raise educational quality...

An important aim is to help host countriesdevelop their own capacity to design appropri-ate policies on an ongoing basis, managehuman and financial effectively, and build sus-tainable capacity. This includes developingeffective policy analysis units within educationministries. (USAID, 2005, pp. 8 -9).

5 Participant comments and feedback from the USAID Education Overview Course in, Hagerstown, Maryland, October 2–13, 2005.

Edu

cation

Ref

orm

Sup

port

Toda

y

42Similar concerns can be observed in otherdonors’ approaches. For example, the WorldBank noted the centrality of the politics ofreform for sustainable education change as farback as 1995, and probably earlier (WorldBank, 1995, p. 137). In the World Bank’s popu-lar “core” course on education reform, the poli-tics of reform receives significant attention,using an approach similar to ERS’s. Finally,“reform” items that emphasize accountabilityand sustainability have been a growing compo-nent of World Bank loans over the past decadeor so (see Crouch & Fasih, 2004).

While USAID does not currently sponsor budgetsupport, sector-wide approaches, other donorsdo so, and USAID has done so in the past, withthe non-project approach of the 1990s. Asnoted above, a current incarnation of thismodality, fairly popular with European donors inparticular, is the Sector Wide Approach (SWAP),which tends to pool donor resources and workthough countries’ own procurement systems,as a way to promote systemic reforms. SWAPhas much in common with ERS: it is a process-oriented approach that emphasizes policy dia-logue (Association for the Development ofEducation in Africa [ADEA], 2002), and thattrades off donors’ prerogatives to choose par-ticular projects (ideas, geographical areas)against the privilege of gaining a seat at thetable where fundamental government policiesare discussed. Reform support could be con-sidered a technique additional to SWAPs, orthat could support SWAPs. Unlike SWAPs, how-ever, reform support does not necessarily usefunding as a form of leverage. Nor does reformsupport require a comprehensive, sector-widereform program and/or project. Reform supportactivities can enhance the policy impact andsustainability of more traditional input-basedprojects and provide well-targeted supports asimplementation-focused companions to a sec-tor-wide reform program.

Shorter-run trends are not mutually exclusivewith longer-term trends. In spite of pressure for results, the most recent trends emphasize the importance of sustainability and local ownership:

We've often created parallel systems of serv-ice delivery that have allowed governments toshirk their responsibility and shifted citizens’expectations from their own governments tothe international donors. But despite thenoblest of intentions, outsiders cannot withsustainability, secure citizens' health andsafety, educate a critical mass, or create theconditions needed for economic growth, all ofwhich are necessary for development, and allof which are the responsibilities of govern-ment… We know that in “developing” coun-tries, where government accountability may belacking, we must address issues of gover-nance and democracy even as we supportprograms in health, education, and povertyalleviation. (Tobias, 2006)

And this is not taking place only at USAID, so itis not likely just a short-term trend. DfID is alsoputting governance at the center of its wholeapproach.

In short, while one could focus on recenttrends to justify a concern with reform and sus-tainability, it seems that an interest in sustain-ability, and hence reform, is a deeper andlonger-term trend that can be tapped to makethe internal case for reform support activities.

As noted by many observers, there is an appar-ent contradiction between a drive for measura-ble, relatively fast results, and a simultaneousinterest in sustainable reforms. But there is anaspect to this contradiction that is not oftennoticed and probably resolves the contradiction(which is why we note that this contradiction ismore apparent than real). The dissatisfactionwith the lack of results is not necessarily dis-satisfaction with all processes that take time todeliver ultimate yield. The pressure for betterresults might arise from high-level officialsobserving projects that should seek short-termimpact, because they are typical pilot projects,but that fail to deliver those effects. Many edu-cation projects, for example, are nominally ori-ented toward “quality,” but they have no goalsexpressed in terms of learning outcomes or anyother palpable quality outcome. Unfortunatelythis has been much more common in education

Education

Reform

Support

Today

43than in, say, the health sector, where outcomesin terms of reductions in infant mortality, forexample, are commonly used to describe adonor’s efforts. It seems reasonable to expecttraditional projects that emphasize inputs liketeacher training or textbook distribution to actu-ally contribute to measurable improvements inlearning within three to five years.

An important hypothesis, thus, is that donoragency leadership is not necessarily impatientwith approaches aimed at sustainability, but is,indeed, and reasonably, impatient with deliveryprojects that do not deliver measurable out-comes. This creates the impression that donoragency leadership would be impatient with any-thing that takes a long time to deliver results,even reform approaches that work. This maynot be the case. After all, donor agenciesengage with projects in governance issues,which take a long time to demonstrably affect“people-level” indicators such as infant mortali-ty. By the same token, though, activities suchas ERS ought to have goals definable in termsof the reforms themselves—goals similar tothose used in governance projects—so theycan deliver something measurable within a rea-sonable planning horizon.

Justifying and explaining reform support activities, then, requires some mixture of thefollowing:

• Reference to longer-term trends and stand-ing policy and, if desired, noting that emerg-ing thinking is not inconsistent with thelonger-term trends. These trends refer to theneed for reform support or ERS-likeapproaches to generate ownership and sustainability.

• Noting that reform support activities providethe highest leverage. Investing a few hundredthousand dollars in policy advice, analysis,or strategic communication that can helpshift local spending of millions of dollars in a particular direction (toward the poor, or inimproved systems of accountability to citizens,or to deliver actual results definable at thechild level) is a good investment.

• Ensuring that reform support approaches, iftried, are addressed at a few policies,instead of just one, in case there is a failureat the one policy one tries to promote. Thisis also necessary because the policy envi-ronment in developing countries is unpre-dictable, and one has to be ready to strikeat opportunities. A timely memorandum orwell-written but relatively casual positionpaper can influence policy making in waysthat a planned policy advice process mightnot be able to, if the memorandum or posi-tion paper is presented at a critical time.

• Including reform support activities as small-er components of more traditional projectsaimed at immediate results. Preciselybecause reform support activities can beinexpensive, compared to providing infra-structure, teacher training, or vehicles, theycan be an effective though small componentof larger, more traditional projects. ERSactivities can connect school-level pilots ordemonstrations projects to the broader poli-cy agenda by documenting and communicat-ing the results obtained and the policy andinstitutional conditions that support thoseresults.

• “Bringing forward” the results of reform sup-port activities by justifying them in terms ofoutputs or outcomes borrowed from the fieldof governance. This may require an explicitalliance with the governance interests in adonor agency, either at headquarters or fieldmission level.

Edu

cation

Ref

orm

Sup

port

Toda

y

44

Education

Reform

Support

Today

457.Jump-Starting Education ReformSupport (What Can Be Done Now?)Section 4 provides examples of activities thatcan make up an education reform support project; section 5 explains that there are certaincircumstances in which an ERS approach mightnot be a good idea, or might need importantadaptations; and section 6 suggests ways tojustify and explain the importance of such interventions in operational or bureaucraticterms. As a matter of practicality, though,smaller reform support interventions can becarried out quickly, within the confines of anexisting project or when smaller amounts offunding are available for quick-start activities.These typically require less complex planningand less justification and can yield quickimpact or lay experimental foundations for amore serious and well-thought-out reform support intervention. Following are six particularly promising possibilities:

1. A project can inject fresh data, gatheredinexpensively, on schooling quality (see the discussion of reading in Peru in example 1from section 4) into policy dialogue sessionsthat foster dramatic awareness of the poorquality of schooling. For example, most policymakers have seldom heard a poor or rural childread and have little idea of how poor the qualityof schooling is, in terms of simple metrics suchas children’s reading ability, compared to thequality of schooling their own, middle- or upper-middle-class children receive. They also havelittle idea how unequal their systems are—forexample, Angola’s education system todayappears more unequal than South Africa’s didbefore the end of apartheid. Such dramaticcomparisons and illustrations can be useful.Simple data exercises (whether with freshempirical data or powerful analysis of existingdata) can be used to focus attention on

resource inequality or poor and unequal results,and can underpin policy reforms oriented atquality and equity. Policy dialogue impact canprobably be had for as little as US$100,000using these sorts of techniques.

2. Policy developments in most countries takeplace on a calendar that has nothing to do withdonors’ analytical or dialogue plans. At certainpoints in the policy process, there will be a sortof climax in the debate, and discussion arounda particular topic (the introduction of free secondary education in Uganda at present, forexample) will become super-heated. These sortsof climaxes in the policy process can take placein the lead-up to a new piece of legislation, inan election campaign, or in the early days of anew government. An intervention at thosepoints can be leveraged many more times (andhence more cost effectively in terms of policyimpact) than a constant barrage of analysesand forums that obey the donor’s plannedcycles or a policy unit’s supply-side analyticalofferings. Being ready with a very proficientanalysis and position paper that can be offeredto a government within weeks, in the middle ofa policy decision climax, is a good way to haveimpact way out of proportion to the cost of theactivity, if the paper and policy dialogue sessionsorganized around it are constructive in spirit,extremely well argued, and use good marketingtechnique. This kind of effort is not expensive,but it can enable entrée into further policydebate. It requires constant watchfulness andtimely intervention as well as some preexistingconnection and access to data. This kind ofintervention can take place within the contextof a more traditional project, with a small policyreform add-on.

Edu

cation

Ref

orm

Sup

port

Toda

y

463. Even if one cannot have immediate and dra-matic input into policy making, existing projectscan be used to lay the groundwork for futureimpact. Using existing projects to create grass-roots knowledge and pressure on simple edu-cation quality and equity indicators (see Nigeriaexample in section 4 above) can develop expe-riences that could be taken to scale in laterprojects that focus primarily on quality andequity of policy reform, if the smaller activitiesseem to work.

4. An existing school-level project can also pur-sue data, analysis, and communication thataddress some of the policy issues that itlearns about from its on-the-ground experience.Data can be gathered that show how the condi-tions for success need to be changed at theschool level to maximize the impact of, say,teacher training, by managing the school day ina way that makes it possible for teachers to dojoint planning and/or to observe each other’slessons. These conditions for success oftenhave policy or institutional implications that willrequire communication and dialogue to clearspace, mobilize supporters, and advocate fornecessary reforms. In this manner a pilotactivity can have broader impact, not by repli-cating its activities, but by altering the larger-scale environment so its activities can beapplied more widely.

5. Projects can also supply, on an opportunisticand demand-led basis, training to countriesthat are serious about reform. The WorldBank’s core course on education reform (nor-mally offered once a year, but offered twice in2006 in response to high demand) is now over-subscribed, which attests to both the quality ofthe course and the growing interest in reform.Collaboration with the World Bank on thiscourse, perhaps offered at a country level, andperhaps with more emphasis on the politicaleconomy aspects, might be a good way toplace some reform ideas on the agenda morefirmly.

A sixth possibility worth considering concernsthe challenge of working in fragile states orthose emerging from conflict or a disaster. In

these cases, external funding focuses on help-ing to rebuild a completely collapsed educationsystem, as in southern Sudan, Afghanistan, orSierra Leone. Schools need to be rebuilt, mas-sive numbers of teachers need to be found andhired, curriculum needs to be redefined, etc.So much attention focused on basic inputs,however, could actually be treated as an oppor-tunity to help rethink how the education systemcan best support the fundamental challenge ofassuring effective schooling. Documentingwhich supports are needed to help schoolsreestablish themselves and thrive in strife-tornareas, or investing in learning from the commu-nity-based or NGO-supported education activi-ties that may have grown up during the periodof conflict or in its immediate aftermath, iftaken with an ERS approach, can form thefoundation of a broad national dialogue abouthow to rebuild an education system. Instead ofreflexively recreating the standard image of aneducation sector, well-timed analysis, communi-cation, and facilitated dialogue could helprethink which system supports are actuallyneeded and how central, regional, and localministerial entities should or should not bereestablished.

Education

Reform

Support

Today

478.ConclusionsEven if the developed countries were to commit It is predicated on the notion that while fosteringas much of their budgets to foreign aid as even policy change is difficult, it is not simply a the most idealistic among us dare hope, aid matter of political intuition, luck, a sort of will be at best a fraction of countries’ total mysterious wisdom, or arcane art. It is—atspending on their own education sectors.6 With least partly—a matter of technique and craftsuch meager resources, and if resources were that, like all technique and craft, can be largelyall donors had to work with, only a trivial dent codified, practiced, and learned. This document,could be made in the developing world’s educa- like the earlier ERS series, tries to codify sometional backlogs. One step in the right direction, of this knowledge and attempts to inspire byaway from the thought that simple resource providing a number of cases that exemplify howtransfers will make much of a difference, is to others have approached the problem and havepromote improvements via demonstration proj- solved it. The requirements are having a variedects and the transfer of technical innovations. toolbox, deploying the tools in a harmoniousBut if these projects enter sterile policy environ- and logical way, being nimble and sensitive toments, then the receptivity is poor, pilot proj- the policy environment, and being able to adaptects remain at the pilot stage, and little is to conditions. And, in all of this, the foremostscaled up or sustained. Parallel implementa- point to remember (which perhaps best sum-tion processes are set up that do not do marizes the whole ERS approach) is that theenough to build up capacity. This has been right to criticize is earned by engagement.7

noted again and again, even with the mostpromising technologies, such as interactive The recent trends toward greater coordinationradio instruction, or certain teaching tech- of external assistance, articulated comprehen-niques. Donors thus seek to improve the poli- sive strategies and plans for achieving thecy environment, as a way of both ensuring that Millennium Development Goals, and pooledcountries use their own funds well, and to help funding to support countries that demonstrateimprove the scalability of innovations the coun- sound governance indicate an encouragingtries introduce as pilot projects. But this poli- increase in donor engagement in supportingcy-level work may, in turn, affect people’s lives country ownership of and commitment to edu-only indirectly. Furthermore, fostering policy cation reforms. However, ERS is founded notimprovement is a very difficult art, requiring just on greater funding agency engagement andmore than setting conditions on loans, provid- policy skill, but even more on increaseding technical advice, or “donor harmonization. engagement of a variety of actors within a

country—the individuals and institutions thatThe ERS approach provides a way for donors to ultimately will determine whether policies gen-bridge the gap between pilot projects and policy uinely reflect local will, and whether they canchange, and to work with counterparts within be implemented and, indeed, translated intohost countries to bring about policy change. tangible improvements in education.

6 Given commonly accepted perceptions of donor dependence and large proportions of budgets in some countries, due to the share of educa-tion coming from donor support, this may seem a strange thing to say. But this impression is created by the large donor role in countries thatare simultaneously very poor and “donor darlings.” If one extends the analysis to all developing countries (i.e., all low- and middle-income coun-tries), total current spending on education is somewhere around US$400 billion per year (estimated from the World Bank’s EdStats service). Ifforeign assistance were to increase somewhere between 100 percent and 200 percent of current levels (current levels being about US$8 bil-lion per year, estimated from DAC online data at http://www1.oecd.org/scripts/cde/DoQuery.asp)—and this is surely on the outside range ofoptimism—the assistance would cover about 5 percent of total education spending. Thus, what countries do with their own money matters agreat deal. The picture changes in the very poorest countries, where donor input can, indeed, make a material and direct difference, at themargin, if well managed, sustained, and replicable.

7 The authors acknowledge George Taylor, Chief of Party, CBE Project, Cambodia, who coined this phrase.

Edu

cation

Ref

orm

Sup

port

Toda

y

48

Education

Reform

Support

Today

49ANNEX A

In 1997, the United States Agency forInternational Development published the series,Education Reform Support (ERS). The six-volumeseries presents an integrated approach to the challenge of scaling up and sustaining education reform in developing countries, with a particular emphasis on Africa. It includes:

Volume One: Overview and BibliographyVolume Two: Foundations of the ApproachVolume Three: A Framework for Making It

HappenVolume Four: Tools and TechniquesVolume Five: Strategy Development and

Project DesignVolume Six: Evaluating Education Reform

Support

ERS aims to integrate traditional public policyanalysis (using information and analytical techniques) with public policy dialogue, advocacy,awareness, and political “salesmanship” (usingcommunication techniques). ERS seeks toinvoke these mechanisms as a means ofimproving the policy-making process so that itis 1) much richer in the use of information andanalysis; 2) more competitive, transparent, and accountable; and 3) more open to broad stakeholder participation.

The ERS approach consists of 1) an operationalframework for getting things done and aprocess for maneuvering strategically withinthat framework; 2) a set of analytical and policy-dialectical tools that are the substance of thatmaneuvering; and 3) a set of suggestions for designing (typically) donor-funded ERS activities. The ultimate aim is to build thenational institutional capacity to apply thisapproach, helping to establish and nurture apermanent “reform support infrastructure.”

Volume One of the series provides an overviewof the entire series, an extensive bibliography,and an explanation of some of the jargon thatis used to explain the ERS approach. The

overview includes 1) background and justifica-tion to place ERS within the context of currentdevelopment thinking; 2) a definition of ERS; 3)how to make it happen; 4) available “tools andtechniques”; and 5) how to define fundableactivities that could become ERS projects. Eachof the areas covered in the overview is takenup in greater detail in the remaining volumes.

Volume Two introduces the problem and establishes the justification and basis for theapproach in terms of past donor activities inthe sector, and it critiques from both “left” and“right” perspectives. This volume also lays outsome of the main lessons learned to establisha basis for the procedures and strategies thatare described in the remaining volumes. Thevolume concludes that, at least in regard todonors’ responsibilities, the aim should be to1) enhance system-wide reform, and 2) developa new type of education project that embodiesa “modernization” or “reformist” agenda(accountability, client orientation, targetedfinancing, competitive access to public fundingfor education provision, etc.) An annex reviewssome issues surrounding conditionality as amechanism for effecting policy change.

Volume Three provides an operational perspectiveon how to support reform activities. This volumeis a practical guide to making Education ReformSupport happen. In particular, it describes arange of implementable, fundable activities toestablish the structures, mechanisms, andmind-sets needed to facilitate learning-driven,ongoing development of an education system.

Volume Four discusses in depth the analyticaland communication tools and techniques thatcan be used to initiate and motivate reform,and places these tools and techniques in thecontext of past and ongoing donor activities.This volume provides considerable space totwo key concepts—systematization and policycommunication. Systematization is describedas a multifaceted strategy that works on several aspects of ERS in combination and

Edu

cation

Ref

orm

Sup

port

Toda

y

50systematically. Policy communication covers therange of communication settings and techniques,from dialogue, to policy marketing, to massmarketing, to negotiation, to advocacy.

Volume Five lays out the specific steps forlaunching a major level of support to an ERSprocess. This volume concentrates on thedevelopment of (typically donor-funded) projectsto carry out all of the operational and technicalactivities discussed in previous volumes. The volume concludes that the essence of asuccessful process of ERS is constant”restrategizing” and redesign, based on monitoring how the process is going. Annexesinclude instruments for 1) determining countryreadiness for participatory policy reform sup-port, 2) assessing stakeholder institutionalproblems; 3) analyzing interest-group pressure;and 4) identifying qualified long-term technicalassistants.

Volume Six presents ideas for how to monitorand evaluate an ERS process. This volumeexplores how to construct a framework that permits objectively verifiable measures of theimpact of policy reform support activities in away that does not subvert the basic intentionsof ERS efforts to build local institutional capacityand promote indigenous demand for reform.

Education

Reform

Support

Today

51BIBLIOGRAPHY

Abadzi, Helen, Crouch, L., Echegaray, M., Pasco, C., & Sampe, J. (2005, June). Monitoring basicskills acquisition through rapid learning assessments: A case study from Peru. Prospects, XXXV(2), 137–156.

Alvarado Pérez, Betty. 2006. “Success in Implementing Policy Dialogue in Peru.” EdData II Working Paper Series, No.1.

Association for the Development of Education in Africa (ADEA). (2002). Sector-wide approaches: Towards a comprehensive, integrated vision of development. ADEA Newsletter, 14(4), 3–5.

Bauer, S. C., & Bogotch, I. E. (2001). Analysis of the relationship among site council resources, council practices, and outcomes. Journal of School Leadership, 11(2), 98–119.

Bodilly, S. J., Chun, J., Ikemoto, G., & Stockly, S. (2004). Challenges and potential of a collaborativeapproach to education reform. Santa Monica, CA: Rand Corporation.

Bray, M. (2000). Community partnerships in education: Dimensions, variations and implications. Education for All: Thematic Studies. Paris: UNESCO.

Bryk, A. S., Sebring, P. B., Kerbow, D., Rollow, S., & Easton, J. Q. (1998). Charting Chicago schoolreform: Democratic localism as a lever for change. Boulder, CO: Westview Press.

Care International. (2003). Reaching underserved populations with basic education in deprivedareas of Ghana: Emerging best practices. Ghana: CARE International.

Clemens, M. (2004). The long walk to school: International education goals in historical perspective.Working Paper #37. Washington, DC: Center for Global Development.

Coburn, C. E. (2003). Rethinking scale: Moving beyond numbers to deep and lasting change.Educational Researcher, 32(6), 3–12.

Contreras, M. E., & Talavera Simoni, M. L. (2003). The Bolivian education reform 1992–2002:Case studies in large-scale education reform. Country Studies, Education Reform and Management Publication Series, 2(1). Washington, DC: Education Section, Human Development Department, World Bank.

Corrales, J. (1999). The politics of education reform: Bolstering the supply and demand; overcominginstitutional blocks. The Education Reform and Management Publication Series, 2(1).Washington, DC: World Bank.

Crouch, L. (1997). Sustainable EMIS: Who is accountable? In D. Chapman & L. Mahlck (eds.),From planning to action: Government initiatives for improving school-level practice. Paris:UNESCO, International Institute for Educational Planning.

Crouch, L. (2006). Book review of Report to the Minister: Review of the financing, resourcing andcosts of education in public schools. Educational Studies, 40(1).

Crouch, L., & DeStefano, J. (1997). Education Reform Support. Volume Five: Strategy developmentand project design. Washington, DC: Office of Sustainable Development, Bureau for Africa,USAID.

Crouch, L., & Fasih, T. (2004). Education after the WDR 2004: An inventory of accessible implementation ideas. Unpublished draft.

Edu

cation

Ref

orm

Sup

port

Toda

y

52Crouch, L., & Healey, F. H. (1997). Education Reform Support. Volume One: Overview and

bibliography. Washington, DC: Office of Sustainable Development, Bureau for Africa, USAID.

Crouch, L., Healey, F. H., & DeStefano, J. (1997). Education Reform Support. Volume Two:Foundations of education sector reform support. Washington, DC: Office of SustainableDevelopment, Bureau for Africa, USAID.

Crouch, L., Schwartz, B., Healey, F. H., & DeStefano, J. (1997). Education Reform Support. Volume Four: Tools and techniques. Washington, DC: Office of Sustainable Development, Bureaufor Africa, USAID.

DC VOICE. (2005). Starting off right: A school system in transition. Washington, DC: Author.

DeStefano, J., Hartwell, A., Moore, A., & Balwanz, D. (2006). Effective schools for disadvantagedand underserved populations. Prepared by USAID’s Education Quality Improvement Project forADEA’s Biennial Meeting, Libreville, Gabon.

DeStefano, J., & Crouch, L. (1997). Education Reform Support. Volume Six: Evaluating EducationReform Support. Washington, DC: Office of Sustainable Development, Bureau for Africa, USAID.

Devarajan, S., & Kanbur, R. (2005). A framework for scaling up poverty reduction, with illustrationsfrom South Asia. Ithaca, NY: Department of Economics, Cornell University.

Doherty, K., & Abernathy, S. (1998). Turning around low-performing schools: A guide for state andlocal leaders. Washington, DC: U.S. Department of Education.

Dyakowska, J. (2004). Parents create better schools: Advocacy, collaboration and partnershipsbetween educators, administrators and parents to build capacity for diversity. Vancouver:University of British Columbia.

Earl, L., Watson, N., & Katz, S. (2003). Large-scale education reform: Life cycles and implicationsfor sustainability. Reading, UK: CfBT Education Trust.

Easterly, W. (2006). White man’s burden: Why the West’s efforts to aid the rest have done somuch ill and so little good. New York: Penguin Press.

Elmore, R. (1996). Getting to scale with good educational practice. Harvard Educational Review,66(1), 1-26.

Elias, M. J., Zins, J. E., Graczyk, P. A., & Weissberg, R. P. (2003). Implementation, sustainabilityand scaling up of social-emotional and academic innovations in public schools. SchoolPsychology Review, 32(3), 303–319.

Fast Track Initiative (2006). Education for All – Fast Track Initiative: Accelerating progress towardsquality universal primary education. Washington, DC: World Bank.

Freund, R. (2005, April). FEPADE. PowerPoint presentation delivered at PREAL’s SeminarioInternacional: Buenas Prácticas para Mejorar la Educación en América Latina, San Salvador.

Fullan, M. (2000). The three stories of education reform. Phi Delta Kappan, 81(8), 518–584.

Fullan, M. (2003). Core principles as a means of deepening large scale reform. Paper prepared for the Department for Education and Skills, Britain.

Fullan, M., Bertani, A., & Quinn, J. (2004). New lessons for districtwide reform. EducationalLeadership, 61(7), 42–46.

Fullan, M., Rolheiser, C., Mascall, B., & Edge, K. (2002). Accomplishing large-scale reform: Atri-level proposition. Cambridge: TERC.

Education

Reform

Support

Today

53Gamage, D. (2003, March). School-based management leads to shared responsibility and quality in

education. Paper presented at the 46th annual conference of the Council for the InternationalExchange of Scholars, New Orleans, LA.

Gershberg, A. I., & Shatkin, G. (2002). Empowering parents and building communities: The role ofschool-based councils in educational governance and accountability. Working Paper #2002-004). New York: The New School University, Community Development Research Center.

Glennan, T. K., Bodilly, S. J., Galegher, J., & Kerri, K. A. (2004). Expanding the reach of educationreforms: Perspectives from leaders in the scale-up of educational interventions. Santa Monica,CA: Rand Corporation.

Grindle, M. (2004). Despite the odds: The contentious politics of education reform. Princeton, NJ,and Oxford: Princeton University Press.

Hargreaves, A., & Goodson, I. (2006). Educational change over time? The sustainability and non-sustainability of three decades of secondary school change and continuity. EducationalAdministration Quarterly, 42(1), 3–41.

Healey, F. H., & DeStefano, J. (1997). Education Reform Support. Volume Three: A framework formaking it happen. Washington, DC: Office of Sustainable Development, Bureau for Africa,USAID.

Heneveld, W., Nidde, A., Rajonhson, L., & Swati, F. (2006, March). Synthesis report: Local studieson the quality of primary education in four countries. Prepared by the Association for theDevelopment of Education in Africa (ADEA) for its biennial meeting, Libreville, Gabon.

Hess, A. G. (1999). Community participation or control? From New York to Chicago. Theory intoPractice, 3(4), 217–224.

Hess, F. M. (2004a). Status quo versus common sense. Education Week, 23(31), 43, 56.

Hess, F. M. (2004b). Common sense school reform. New York: Palgrave Macmillan.

Inter-American Development Bank and David Rockefeller Center. (2005). The politics of policies:Economic and social progress in Latin America. Washington DC: Inter-American Development Bank.

Jagannathan, S. (2001). The role of non-governmental organizations in primary education: A studyof six NGOs in India. Working Research Paper No. 2530. Washington, DC: World Bank.

Jehl, J., Blank, M., & McCloud, M. (2001, July/August). Lessons in collaboration: Bringing togethereducators and community builders. Shelterforce Online.

Johnson, P. E., & Logan, J. (2000). Efficacy and productivity: The future of school-baseddecision-making councils in Kentucky. Journal of School Leadership, 10(4), 311–331.

Kaufman, R., & Nelson, J. (Eds.). (2004). Crucial needs, weak incentives: Social sector reform,democratization and globalization in Latin America. Baltimore, MD: Woodrow Wilson CenterPress/John Hopkins University Press.

Krisnamoorthi, S. R. (2000). Making local school councils work: The implementation of local schoolcouncils in Chicago public elementary schools. Journal of Law and Education, 29(3), 285–314.

Kronley, R. A., & Handley, C. (2003). Reforming relationships: School districts, external organizations,and systemic change. Providence, RI: School Communities that Work.

Lake, R. J., & Hill, P. T. (Eds.). (2005). Hopes, fears and reality: A balanced look at charter schoolsin 2005. Seattle, WA: National Charter School Research Project.

Edu

cation

Ref

orm

Sup

port

Toda

y

54Mastro, E., & Jalloh, M. G. (2005). Enhancing service through effective school/community

collaboration. Practice Matters. Ithaca, NY: ACT for Youth Upstate Center for Excellence.

McDermott, K. A. (2000). Barriers to large-scale success of models for urban school reform.Educational Evaluation and Policy Analysis, 22(1), 83–93.

Miller-Granvaux, Y., Welmond, W., & Wolf, J. (2002). Evolving partnerships: The role of NGOs in basic education in Africa. Washington, DC: USAID.

Moe, T. M. (2003). Reform blockers. Education Next, 3(2), 56-61.

Moore, D. R., & Hanson, M. (2001). School system leaders propose ineffective strategies that are contradicted by test results and research. Chicago: Designs for Change.

Moore, D. R., & Merritt, G. (2002). Chicago’s local school councils: What the research says.Chicago: Designs for Change.

Moulton, J., Mundy, K., Welmond, M., & Williams. J. (2001). Education reforms in sub-SaharanAfrica: Paradigm lost? Westport, CT: Greenwood Press.

Muhammed, L., & LeCzel, D. (2003). School self assessment in Northern Namibia. Paper presentedat the American Evaluation Association 2003 conference, Reno, NV.

Nath, S. R., Sylva, K., & Grimes, J. (1999). Raising basic education levels in rural Bangladesh: Theimpact of a non-formal education programme. International Review of Education, 45(1), 5–26.

Pankratz, R. S., & Petrosko, J. M. (Eds.). (2000). All children can learn: Lessons from the Kentuckyreform experience. San Francisco: Jossey-Bass.

Pritchett, Lant. (2004). Towards a new consensus for addressing the global challenge of lack ofeducation. Working Paper #43. Washington, DC: Center for Global Development.

Ryan, S., Bryk, A. S., Lopez, G., Williams, K. P., Hall, K., & Luppescu, S. (1997). Charting reform:LCSs—Local leadership at work. Chicago: Consortium on Chicago School Research.

Samoff, J., & Sebatane, E. M. (with Martial Dembele). (2001, October). Scaling up by focusingdown: Creating space to expand education reform. Commissioned by the Association for theDevelopment of Education in Africa (ADEA) for its biennial meeting, Arusha.

Schiefelbein, E. (2006, May). Lessons from the education reforms in El Salvador, 1993–2005.Presented at the USAID Latin American Education Conference, Miami, FL.

Schmoker, M. (2004). Tipping point: From feckless reform to substantive instructional improvement.Phi Delta Kappan, 85(6), 424–432.

Sexton, R. F. (2004). Mobilizing citizens for better schools. New York: Teachers College Press.

School Communities that Work. (2002). School communities that work for results and equity.Providence, RI: Annenberg Institute for School Reform

Shanti, J. (2001) The Role Of Nongovernmental Organizations in Primary Education: A Study of SixNGOs in India. Policy Research Working Paper Series No. 2530. Washington, DC: World Bank.

Siegel, D. E., & Fruchter, N. (2002, April). An implementation study of performance-driven budgetingin the New York City public schools. Paper presented at 2002 annual meeting of AmericanEducational Research Association, New York, NY.

Smith, H. (2005). Ownership and capacity: Do current donor approaches help or hinder theachievement of international and national targets for education? International Journal ofEducational Development, 25(4), 445–455.

Education

Reform

Support

Today

55Southard, M., Muldoon, J., Porter, P., & Hood, D. (1997, March). School councils’ perceptions of

site-based decision making: Challenges and benefits. Paper presented at the annual meeting ofthe American Educational Research Association, Chicago, IL.

Stringfield, S. (2002). Big change questions: Is large-scale educational reform possible? Journal ofEducational Change, 3(1): 63–73.

Swedish International Development Agency (SIDA). (2002). SIDA’s policy for sector programme support and provisional guidelines.

Tobias, R. (2006). Remarks by Ambassador Randall L. Tobias, Director of U.S. Foreign Assistanceand USAID Administrator: ACVFA and the new approach to U.S. foreign assistance. The AdvisoryCommittee on Voluntary Foreign Aid (second public meeting for 2006). Washington, DC: TheGeorge Washington University.

Uemura, M. (1999). Community participation in education: What do we know? Effective schoolsand teachers and the knowledge management system, HDNED. Washington, DC: World Bank.

UNESCO Institute for Statistics. (2006). Children out of school. Measuring exclusion from secondary education. Montreal.

USAID. (1998) Handbook of Democracy and Governance Program Indicators. Center for Democracyand Governance, Technical Publication Series

USAID. (2005). Education strategy. Improving lives through learning. USAID Issue Paper No.1.Washington, DC.

USAID. (2006). Highlights of LAC mission reports. Unpublished. LAC Bureau, Education Office.Washington, DC.

USAID. (2003). “Approaching Education from a Good Governance Perspective. USAID ResourceGuide for Joint DG/Education Programs.” Office of Democracy and Governance.

Uvin, P., Jain, P. S., & Brown, D. L. (2000). Think large and act small: Toward a new paradigm forNGO scaling up. World Development, 28(8), 1409–1419.

Walker, E. M. (2000). Decentralization and participatory decision-making: Implementingschool-based management in the Abbott districts. Research Brief, 1(1).

Warren, M. R. (2005). Communities and schools: A new vision of urban education reform. HarvardEducational Review, 75(2), 133–139.

Weiss, S. (2001). School-based management: Rhetoric vs. reality. The Progress of EducationReform, 2(5).

Winkler, D. (2006). Identifying the impact of education decentralization on the quality of education draft). Education Quality Improvement Project 2, USAID.

Wohlstetter, P., & Briggs, K. L. (2001). Key elements of a successful school-based managementstrategy. School reform in Chicago: Lessons and opportunities. Chicago: Chicago Community Trust.

World Bank. (1995). Priorities and strategies for education: A World Bank review. Washington, DC.

World Bank. (2004). World development report 2004: Making services work for poor people.Washington, DC.

If you have questions, please contact:

USAIDPatrick CollinsCTO/EGAT/EDUSAID Washington1300 Pennsylvania Ave., NWWashington, DC 20532Tel: 202-712-4151Email: [email protected]

AEDJohn GilliesEQUIP2 Project Director1825 Connecticut Ave., NWWashington, DC 20009Tel: 202-884-8000Email: [email protected]: www.equip123.net

This paper was made possible by the generous support of the Americanpeople through the United StatesAgency for International Development(USAID) under Cooperative AgreementNo. GDG-A-00-03-00008-00. The contents are the responsibility of theAcademy for Educational Development(AED) through the Educational QualityImprovement Program 2 (EQUIP2) anddo not necessarily reflect the views ofUSAID or the United States Government.

Documents

Education Reform Support Today - epdc.org Reform... · Many of the examples we could not have compiled without the help of ... existing data to support debate or gather ... Education