96450 SYSTEMSAPPROACHFORBETTEREDUCATIONRESULTS     SABER WorkingPaperSeries Number9 March 2015 WhatMattersMostfor SchoolAutonomyandAccountability: AFrameworkPaper  SystemsApproachforBetterEducationResults SABERWorkingPaperSeries          WhatMattersMostfor SchoolAutonomyandAccountability: AFrameworkPaper          AngelaDemasandGustavoArcia  GlobalEngagementandKnowledgeTeam EducationGlobalPractice TheWorldBank   March30,2015   TableofContents Acronyms.........................................................................................................................................i ..........................................................................................................................ii  Acknowledgements iii Abstract........................................................................................................................................... I. Introduction............................................................................................................................. 1 Objectives................................................................................................................................1 DecentralizationandEducation.............................................................................................. 1 II. WhatareSchoolAutonomyandSchoolAccountability?........................................................ 2 III. ConceptualFramework........................................................................................................... 4 Evidence..................................................................................................................................8 TheThreeA’sandSABERSAA............................................................................................... 13 IV. WhatMattersMost?SAAPolicyGoals,PolicyActionsandEvidence...................................13 PolicyGoalOne:LevelofAutonomyinthePlanningandManagementoftheSchool Budget…….............................................................................................................................. 14 PolicyGoalTwo:LevelofAutonomyinPersonnelManagement........................................18 PolicyGoalThree:RoleoftheSchoolCouncilonSchoolGovernance................................22 PolicyGoalFour:SchoolandStudentAssessment.............................................................. 27 PolicyGoalFive:SchoolAccountability................................................................................. 30 V. ImplementingtheSABERFramework................................................................................... 34 SABERInstrumentandMethodology.................................................................................... 34 VI. Conclusions............................................................................................................................ 36 References....................................................................................................................................37 WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper i  Appendix1:SchoolAutonomyandAccountabilityPolicyGoalsandPolicyActions...................45 Appendix2:RubricforSABERSchoolAutonomyandAccountability.........................................46  Figure1.TheShortandLongRoutestoAccountability................................................................. 5 Figure2.The3A’sModelasaClosedͲloopSystem....................................................................... 7 Figure3.SBMimplementationͲyearsuntilfullimpactofintervention......................................11  Box1.WhatareSchoolAutonomyandAccountability?................................................................ 3 Box2.PathstoSchoolͲBasedManagement................................................................................... 6 Box3.ClosedͲloopsystemsandSBM............................................................................................. 6 ............................................................................................................ 7 Box4.ManagerialActivities  Table1.SelectedexperienceswithSBMinterventionsandtheirimpacts....................................9 Table2.PolicyGoal1:Policyactions,indicatorsandevidence................................................... 18 Table3.PolicyGoal2:Policyactions,indicatorsandevidence................................................... 21 Table4.PolicyGoal3:Policyactions,indicatorsandevidence................................................... 26 Table5.PolicyGoal4:Policyactions,indicatorsandevidence................................................... 30 Table6:PolicyGoal5:Policyactions,indicators,andevidence................................................. 33 WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper ii  Acronyms  AAA  Autonomy,Assessment,andAccountability BOS  BantuanOperasionalSekolah CMS  CommunityͲManagedSchools DCI  DataCollectionInstrument EDUCO EducaciónconParticipacióndelaCommunidad EMIS  EducationManagementInformationSystem ETP  ExtraTeacherProgram OECD  OrganisationforEconomicCoͲoperationandDevelopment PEC  ProgramaEscuelasdeCalidad PACEͲA PartnershipforAdvancingCommunityͲbasedEducationͲAfghanistan PISA  ProgramforInternationalStudentAssessment SAA  SchoolAutonomyandAccountability SIPs  SchoolImprovementPlans SBM  SchoolͲBasedManagement SABER  SystemsApproachforBetterEducationResults                 WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper i  Acknowledgements  This report was prepared by Angela Demas, Senior Education Specialist and Task Team Leader, Global Engagement and Knowledge Team; and Gustavo Arcia, Consultant, Global Engagement and Knowledge Team.Thanksinparticulargo tothemeeting chairs: Amit Dar,Director, Education,andHarry Patrinos, PracticeManager,fortheirguidanceandfeedbackonthecontentanddirectionofthepaper.Wewould like to thank the peer reviewers: Dandan Chen, Program Leader (ECCU3) and Tazeen Fasih, Senior Education Specialist (GEDDR), whose valuable feedback provided insights and helped to enhance the qualityofthepaper.ThereportalsobenefittedfromtheadviceandinputsfromLuisBenveniste,Practice Manager; Husein AbdulͲHamid, Donald Rey Baum, Mary Breeding, Marguerite Clarke, Emily Garner, Oni LuskͲStover,ClarkMatthews,HalseyRogers,andTakakoYuki.Wearealsogratefultothepriormembers oftheSABERSAAteam,particularlyKazuroShibuya,SeniorEducationSpecialist(secondedfromJICA)who contributedtothepilotingofinitialinstruments,therevisionofthepolicyactions,asetofrubricsandthe datacollectioninstrument basedonthisframeworkpaper.Inaddition,wewouldliketo thankRestituto Jr.Cardenas,ProgramAssistantandFahmaNur,SeniorProgramAssistantfortheirsupportinformatting andfinalizingthepaper.    WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper ii  Abstract  Thispaperprovidesanoverviewofwhatmattersmostforschoolautonomyandaccountability.Thefocus is on public schools at the primary and the secondary level. This paper begins by grounding School Autonomy and Accountability in its theoretical evidence base (impact evaluations, lessons learned from experience,andliterature reviews)and thendiscussesguidingprinciplesandtoolsforanalyzing country policy choices. The goal of this paper is to provide a framework for classifying and analyzing education systemsaroundtheworldaccordingtothefollowingfivepolicygoalsthatarecriticalforenablingeffective schoolautonomyandaccountability:(1)levelofautonomyintheplanningandmanagementoftheschool budget;(2)levelofautonomyinpersonnelmanagement;(3)roleofschoolcouncilsinschoolgovernance; (4)schoolandstudentassessment,and(5)accountabilitytostakeholders.Thispaperalsodiscusseshow country context matters to school autonomy and accountability and how balancing policy goals matters topolicymakingforimprovededucationqualityandlearningforall.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper iii  I. Introduction  Objectives Theobjectiveofthispaperistoprovideaframeworkforwhatmattersmostinfosteringschoolautonomy and accountability (SAA) and why this is important. The focus is on public schools at the primary and secondary levels. The paper also discusses School Autonomy and Accountability tools for assessing a country’s development of policies that provide an enabling environment for SAA. SABERͲSAA is one of the instruments that has been developed and tested under SABER, the Systems Approach to Better Education Results, initiative created by the World Bank as part of its education strategy (World Bank 2011b). The application of the policy intent and policy implementation instruments can be important tools for education system reform if they are used as instruments for planning and monitoring the enablingconditionsforimprovingsystemperformance.  This paper begins by providing a short background on decentralization and its relationship to the education sector through SAA.  It then provides the case for school autonomy and accountability and introducestheconceptualframeworkforSAA.Next,itgroundsSAAinitstheoreticalevidencebaseand discusses the guiding principles and tools for analyzing country policy choices. A goal of the paper is to provide a framework for classifying and analyzing education systems around the world according to the followingfivepolicygoalsthatarecriticalforenablingeffectiveschoolautonomyandaccountability:(1) level of autonomy in the planning and management of the school budget; (2) level of autonomy in personnel management; (3) role of school councils in school governance; (4) school and student assessment, and (5) accountability to stakeholders.  This paper also discusses how country context matterstoschoolautonomyandaccountabilityandhowbalancingpolicygoalsmatterstopolicymaking forimprovededucationqualityandlearningforall. DecentralizationandEducation In matters of governance, decentralization is seen as an appealing alternative to the centralized state given the range of benefits associated with this approach. It is regarded as a way to: (i) introduce more intergovernmental competition and checks and balances; (ii) make government more responsive and efficient in service delivery, (iii) diffuse social and political tensions and ensure local and political autonomy (Bardhan 2002). Decentralization can help ease decisionͲmaking bottlenecks that are caused bycentralgovernmentplanningandcontrolofimportanteconomicandsocialactivities.Itcanalsohelp simplify complex bureaucratic procedures and increase sensitivity to local conditions and needs, by placingmorecontrolatthelocallevelwhereneedsarebestknown.Withdecentralization,theimpactwill depend on the many factors related to design. Similar to other policy issues that are complicated, the outcomewilldependonamyriadofindividualpolitical,fiscal,andadministrativepoliciesandinstitutions as well as their interaction within a given country (Litvak and Seddon 1999; Bardhan 2002). At the same time, it is important to keep in mind that structures of local accountability may not be in place in developing countries and “capture” by local elites may frustrate the goal of quality and equitable public service delivery. To be effective, decentralization must attempt to change existing structures of power within communities, improve opportunities for participation and voice, and engage all citizens including thepoorordisadvantagedintheprocess(Bardhan2002).  Thereseemstobeaconsensussincethe1980s,thattoomuchcentralizationor,conversely,absolutelocal autonomy are both harmful and that it is necessary to put in place a better system of collaboration between the national, regional and local centers of decisionͲmaking.  For decentralizing education WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 1 systems, the process requires strong political commitment and leadership in order to succeed (McLean and King 1999). Countries around the world have been experimenting with some form of education decentralization.Ithasbecomecentraltoeducationpolicy.Initialevidenceindicatesthatdecentralization tosubnationalgovernmentsmaybeinsufficient,andinordertoimproveschoolsandlearning,increased autonomy for communities and school actors may be necessary (McLean and King 1999). A way to decentralize decisionͲmaking power in education from the central government to the school level is knownasschoolͲbasedmanagement(SBM)(Caldwell2005;Barrera,FasihandPatrinos2009).  Decentralizededucationcanhelpgetparentsandstudentsclosertotheprovidersofeducation,ensuring betteraccesstopedagogicalandmanagerialmethodsmoreintunewiththeirneeds.However,ifsuchan approach is taken to the limit, it may result in a fragmented education system where standards may be reduced and local community values may become too parochial to benefit society at large (Ritzen, van Domelen and de Vijlder 1997).  This paper discusses what matters most for school autonomy and accountabilityandproducinganenablingenvironmentfortheintendedoutcomes.  II. WhatareSchoolAutonomyandSchoolAccountability? Improved school management leads to better outcomes. Decentralization, school autonomy and community empowerment have been at the center of the education policy discussions for several decades.Wearebeginningtounderstandmoreandmorethroughagrowingbodyofevidencethathigher managementqualityisstronglyassociatedwithbettereducationaloutcomes(Bloometal.2014).Itleads tomoreefficientschoolsthathaveautonomytomakedecisionsonbudget,management,personnel,and everyday items that have an impact on their school environment and learning that is taking place. This includes changing the environment in which decisions about resource allocation are made, where effectiveschoolͲleveldecisionͲmakingcantakeplacebyschoolͲlevelagents.Italsomeansthatthosewho aretakingdecisionsareaccountabletohigherlevelsofauthorityatthedistrictandcentrallevelsbutalso tothegreaterschoolcommunitywhoall,tosomedegree,haveoversightroleswhethertheyarepolicyͲ makers,supervisorsorconsumersofeducationservices. School autonomy and accountability are key components of an education system that ensure educationalquality.Bytransferringcoremanagerialresponsibilitiestoschools,schoolautonomyfosters local accountability; helps reflect local priorities, values, and needs through increased participation of parents and the community; and gives teachers the opportunity to establish a personal commitment to students and their parents.  Increased school autonomy and improved accountability are necessary conditions for improved learning because they align teacher and parent incentives (Bruns, Filmer and Patrinos 2011). Studies have shown a clear causal link between school autonomy and efficiency in resourceuse(Barreraetal.2009).Viewedinthiscontext,schoolautonomyandaccountabilityshouldbe consideredessentialcomponentsofanoverallstrategyforimprovinglearningoutcomes. Benchmarking and monitoring indicators of school autonomy and accountability allows a country to rapidly assess its education system, thus setting the stage for improving policy planning and implementation. To be clear onwhatismeantbyschoolautonomyandaccountabilityinthispaperseedefinitionsinBox1.       WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 2   Box1.WhatareSchoolAutonomyandAccountability? School autonomy is a form of school management in which schools are given decisionͲmaking authority over their operations, including the hiring and firing of personnel, and the assessment of teachersandpedagogicalpractices.Schoolmanagementunderautonomymaygiveanimportantrole to the School Council, representing the interests of parents, in budget planning and approval, as well asavoice/voteinpersonneldecisions.ByincludingtheSchoolCouncilinschoolmanagement,school autonomyfostersaccountability(DiGropello2004,2006;Barrera,FasihandPatrinos2009). Initsbasicformaccountabilityisdefinedastheacceptanceofresponsibilityandbeinganswerablefor one’sactions.Inschoolmanagement,accountabilitymaytakeotheradditionalmeanings:(i)theactof compliancewiththerulesandregulationsofschoolgovernance;(ii)reportingtothosewithoversight authority over the school; and (iii) linking rewards and sanctions to expected results (Heim 1996; Rechebei2010).   To be effective, school autonomy must function on the basis of compatible incentives, taking into account national education policies including incentives for the implementation of those policies. Having more managerialresponsibilitiesattheschoollevelautomaticallyimpliesthataschoolmustalsobeaccountableto localstakeholdersaswellasnationalandlocalauthorities.Theempiricalevidencefromeducationsystemsin which schools enjoy managerial autonomy is that autonomy is beneficial for restoring the social contract betweenparentsandschoolsandinstrumentalinsettinginmotionpoliciestoimprovestudentlearning. The progression in school autonomy in the last two decades has led to the conceptualization of SchoolͲ BasedManagement(SBM)asaformofadecentralizededucationsysteminwhichschoolpersonnelare inchargeofmakingmostmanagerialdecisions,frequentlyinpartnershipwithparentsandthecommunity oftenthroughschoolcouncils1(Barrera,Fasih,andPatrinos2009).Morelocalcontrolhelpscreatebetter conditions for improving student learning in a sustainable way since it gives teachers and parents more opportunities to develop common goals, increase their mutual commitment to student learning, and promotemoreefficientuseofscarceschoolresources. Types of SchoolͲBased Management. In addition to the degree of devolved autonomy provided to the schoollevel,SBMmustdefinewhoisinvestedwiththedecisionͲmakingpowerattheschoollevel.There arefourSBMmodelstohelpusdefinethis(BarreraͲOsorioetal.2009):  x Administrative control:  Authority is devolved to the principal. Its aim is to make each school more accountable to the central district.  The benefits include increasing efficiency of expenditures on personnel and curriculum and making one person more accountable to the centralauthority. x Professional control:  Teachers hold the main decisionͲmaking authority.  This model aims to make better useofteachers’knowledgeofwhattheschoolneedsatthe classroomlevel.Itcan motivateteachersandleadtogreaterefficiencyandeffectivenessinteaching. x CommunityͲcontrol:  Parents or the community have major decisionͲmaking authority. Under thismodelitisassumedthatprincipalsandteachersbecomemoreresponsivetoparents’needs andthecurriculumcanreflectlocalneeds. 1Theterm“schoolcouncil”issynonymouswithseveralothertermsusedaroundtheworld,suchasschoolmanagement committee,parentcouncil,schoolcommittee,etc.Forconsistency,thispaperwilluseschoolcouncil. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 3  x Balanced control:  DecisionͲmaking authority is shared by the principal, teachers and parents. The aims are to take advantage of teachers’ knowledge of the school to improve school managementandtomakeschoolsmoreaccountabletoparents.  Existing models of SBM in real life generally blend the four models. SBM is not a set of predetermined policies and procedures, but a continuum of activities and policies put into place over time and with contextual sensitivity to improve the functioning of schools, allowing parents and teachers to focus on improvementsinlearning.Whilethereislittlehardevidencethatteacherqualitygrowsasadirectresult of SBM, it can be argued that increasing school accountability is a necessary condition for improving teacherquality.ImplementingSBMcanaugmentthesupportthatschoolcouncilsandparentsprovideto good teachers through various methods including salary and nonͲsalary incentives and establishing the necessaryconditionstoattractthebestteachers(Arciaetal.2011).Assuch,SBMcanfosteranewsocial contract between teachers and the community in which local cooperation and local accountability drive improvementsinprofessionalandpersonalperformancebyteachers(Patrinos2010).  III. ConceptualFramework  Whiletherehavebeenmanyschoolsofthoughtacrossthedifferentexperiencesinschoolautonomy,the principleofaccountabilitywasnotinitiallylinkedwithschoolautonomy(Eurydice2007).InthemidͲ1990s, theconceptofautonomywithaccountabilitybecameincreasinglyimportantandassumeddifferentforms in different countries. PISA results suggest that when autonomy and accountability are combined, they tend to be associated with better student performance (OECD2011).The experience of highͲperforming countries2onPISAindicatesthat: x Education systems in which schools have more autonomy over teaching content and student assessmenttendtoperformbetter. x Education systems in which schools have more autonomy over resource allocation and that publishtestresultsperformbetterthanschoolswithlessautonomy. x Education systems with standardized student assessment tend to do better than those without suchassessments. It was not until almost 10 years after the concept of linking autonomy with accountability started to emergethatthatTheWorldDevelopmentReport2004“Makingserviceworkforpoorpeople”introduced aconceptualframeworkfortheempowermentofcommunities.Thereporthighlightsthesignificanceof a “short” route of accountability that runs directly from users (e.g. citizens/clients/ communities) to frontlineserviceproviders(e.g.schools),inadditiontoanindirector“long”routeofaccountabilitywhere users hold service providers accountable through the state (Figure 1). SchoolͲbased management has beenreferredtoasaneffectivewaytoachievetheshortrouteofaccountabilityintheeducationsector.    2ExamplesofhighperformingcountriesthathaveimplementedschoolͲbasedmanagementpoliciesandframeworksinclude theNetherlands,Canada,andNewZealandamongothers. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 4  Figure1.TheShortandLongRoutestoAccountability  Source:AdaptedfromWDR2004 This framework illustrates two routes to accountability and applies to any system where the state or politicians set policy and rules – providers of services receive funding and have the mandate to deliver quality services – and the clients or citizens who receive services.   The traditional or long route of accountability happens when citizens can formally “voice” their concerns through voting for politicians who most closely are aligned with their ideologies and promise to provide the funding and services that thecitizenswant(compact).  Theshorterrouteaffordsclientsthepowertomorefrequentlyprovidefeedbacktoproviderstoletthem know how they are doing and to hold them accountable for good quality services. For education, the short route allows for voice and inputs on decisionͲmaking at the school level for direct clients who are parentsandstudents.DecisionͲmakingattheschoollevelisimportantandinvolvesavarietyofactivities. TheempiricalevidencefromSBMshowsthatitcantakemanyformsorcombinemanyactivities(Barrera etal.2009)withdifferingdegreesofsuccess(seeBox2).  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 5  Box2.PathstoSchoolͲBasedManagement InmanycountriestheimplementationofSBMhasincreasedstudentenrollment,studentandteacher attendance,andparentinvolvement.However,theempiricalevidencefromLatinAmericashowsvery fewcasesinwhichSBMhasmadeasignificantdifferenceinlearningoutcomes(Patrinos2011),while in Europe there is substantial evidence showing a positive impact of school autonomy on learning (Eurydice2007).TwoapproachestoSBMͲthegrassrootsapproachtakeninLatinAmerica,incontexts wheretheinstitutionalstructurewasweakorservicedeliverywashamperedduetointernalconflict, andtheoperationalefficiencyapproachtakeninEurope,whereinstitutionswerestrongerͲcoincide in applying managerial principles to promote better education quality, but they are driven by two differentmodesofaccountabilitytoparentsandthecommunity.IntheLatinAmericanmodel,schools areheldaccountablethroughparticipatoryschoolͲbasedmanagement(DiGropello2004)whileinthe European model accountability is based on trust in schools and their teachers (Arcia, Patrinos, Porta andMacdonald2011).Ineithercase,schoolautonomyhasbeguntotransformtraditionaleducation fromasystembasedonprocessesandinputsintoonedrivenbyresults(Hood2001).   When do SBM components become critical for learning?  When a school or a school system does not functionproperly,itcanbeasubstantialbarriertosuccess.Themanagerialcomponentofaschoolsystem isanecessarybutinsufficientconditionforlearning.Onecanfixsomemanagerialcomponentsandobtain no results or alter other components and get good results. The combination of components crucial for success is still under study, but the evidence to date points to a set of variables that foster managerial autonomy, the assessment of results, and the use of the assessment to promote accountability among all stakeholders (Bruns, Filmer and Patrinos 2011). When these three components are in balance with eachother,theyforma“closedͲloopsystem”(seeBox3).Visually,itistheclosingofacircleofthethree interrelatedcomponents.   Box3.ClosedͲloopsystemsandSBM Theinterrelationsbetweenautonomy,assessment,andaccountabilitycanbe comparedtoa“closedͲloopsystem”,oroneinwhichfeedbackconstantly informsoutput.InaclosedͲloopsystem,datadoesnotflowoneway;instead,it returnstopartsofthesystemtoprovidenewinformationthatdynamically influencesresults.InthecaseofSBM,assessment,forexample,bothenablesthe autonomyofschoolcouncilstomakeinformeddecisionsaboutschoolquality andalsoallowsforaccountabilityatahigherlevel,whichcanmeasureresultsattheschoolleveland providesupportasnecessary.InaclosedͲloopsystem,allelementsinbalancearecriticalto achievingsuccess(KaplanandNorton2008). WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 6  Defining a managerial system that can achieve closure is conceptually important for school based management, since it transforms its components from a list of managerial activities (Box 4), to a set of interconnected variables that work together to improve system performance. Unless SBM activities contribute to system closure, they are just a collection of isolated managerial decisions. As components of a managerial system, SBM Box4.ManagerialActivities activities may behave as mediating variables: they produce an enabling environment for teachers and students, allowing for 9 Budgeting,salaries pedagogical variables, school inputs, and personal effort to work as 9 Hiring,transfers intended. 9 Curriculum 9 Infrastructure If an SBM system is unable to close the loop, are partial solutions 9 Schoolgrants effective? Yes, schools can still function but their degree of 9 Schoolcalendar effectivenessandefficiencywouldbelowerthanifthesystemcloses 9 Monitoring& theloop.Inthisregard,SBMcanachieveclosureoftheloopwhenit Evaluation allows enough autonomy to make informed decisions, evaluate its 9 Dissemination resultsandusethatinformationtoholdsomeoneaccountable. Representationally this is captured in the “Three A’s Model.” SBM can achieve balance as a closedͲloop system when autonomy, student assessment, and accountability, are operationally interrelated through the functions of their school councils, the policies for improving teacherquality,andEducationManagementInformationSystems(EMIS)(seeFigure2).  Figure2.The3A’sModelasaClosedͲloopSystem  Source:AdaptedfromArciaetal.2011     WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 7  School councils are crucial for implementing school autonomy because they serve as representatives of the school clients: parents and students. As such, the school council can be a resource to school managementintheprocessoftailoringschoolservices(curricula,teachingmaterials,schoolcalendar,and teacher selection) to the needs of students. A more active role of school councils in school governance canmakeschoolautonomymoreeffective. Schoolassessmentsarethevehiclesusedbyschoolstodeterminetheirneedsforchangesinpedagogical practicesandtodeterminethetrainingneedsofteachers.Themainobjectiveofanyassessmentsystem is to monitor learning, which in turn is linked to teacher quality. So for SBM to be a closedͲloop system, schoolandstudentassessmentwouldneedtolinktoteacherperformanceandteacherquality. Finally,anEMISisintegraltoaccountabilitybecauseitisthemechanisminplacetoreportonperformance indicatorsattheschoolandsystemlevels.AnEMISenforcesaccountabilitytotheextentthatitisfeddata of good quality and it is used to produce reports that are informative to parents and society about the performanceoftheeducationsector.Insummary,theinterrelationbetweenAutonomy,Assessment,and Accountability(AAA)mustbemadeoperationalbyreinforcingtherolesofschoolcouncils,policiesaimed atimprovingteacherquality,andtheoperationofanEMIS.Otherwise,thereisariskthattheAAAmodel maynotreachtheoptimalstatusasaclosedͲloopsystem. Inmanagerialterms,itisclearthatthepointofcontactbetweenautonomousschoolsandtheirclientsis primarilythroughtheschoolcouncil(Corrales2006).Similarly,schoolassessmentsarethevehiclesused by schools to determine their needs for changes in pedagogical practices and to determine the training needsoftheirteachers.Bothpedagogicalchangesandteachertrainingaredeterminantfactorsofteacher quality(Vegas2001).Finally,theroleofEMISonaccountabilityiswellestablishedandmakesiteasierto report on indicators of internal efficiency and on standardized test scores (Bruns, Filmer, and Patrinos 2011).  Evidence Wehavediscussedwhatschoolautonomyandaccountabilityareandthatwhentheyareappliedtogether they can be beneficial for improving school efficiency, effectiveness and learning outcomes. Empirical evidence from countries that have implemented school autonomy and accountability suggests that a certainsetofpoliciesandpracticesareeffectiveinfosteringmanagerialautonomy,assessmentofresults, andtheuseofassessmentstopromoteaccountability(Table1).   WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 8  Table1.SelectedexperienceswithSBMinterventionsandtheirimpacts Program/ Country Authors Intervention Results/Findings Project Educacióncon Communityassociationsresponsible Increasesreadingscores Jimenez& Participación foradministeringfunds,hiring/firing &decreases Sawada ElSalvador dela teachers,monitoring&maintaining absenteeism(1999); 1999, Communidad infrastructure. Increasesretention 2003,2014 (EDUCO) (2003,2014) a)Schoolgrantforallschools(school Positiveeffecton committeedevelopSIP),b)Training learningoutcomes; forschoolcommitteeinplanning, increasedtestscoresin Bantuan budgetingandsupportingeducation language0.51standard Operasional quality.c)Democraticelectionof deviationsandmathby Sekolah(BOS) Pradhanet Indonesia schoolcommitteemembers,d) 0.46SD.School (School al.2011. Linkagebetweenschoolcommittee committee'slinkage Operational andvillagecounciltoenableschool withvillagecouncilsand Assistance) committeetomobilizecommunity havingelectedschool support. committeemembers madepositiveimpact. ExtraTeacher Trainingschoolcommitteesto Higherstudenttest Duflo, ProgramͲPeer monitorteacherperformanceand scores,lowerteacher Dupas,and Kenya Effects,PTR, committeeͲbasedhiringofteachers. absenteeism,small Kremer. andTeacher decreaseinstudent 2007 Incentives dropout Murnane, Grantsprovidedtopublicschoolsto Reduceddropoutrate; Willett,and implement5Ͳyearschool noeffectonrepetition Programa Cardenas. improvementplans(SIPs)thatschool Escuelasde 2006. staffandcommunitydesign.Parent Calidad(PEC) associationspurchasingsuppliesand Decreaseddropout, (Quality Skoufias& carryouttheplans.Trainingfor failure,andrepetition School Shapiro. schoolprincipals. rates.Nomeasurable Program) 2006. impactonoutcomesin indigenousschools. Smallgrantstoparentassociations Increasedparticipation Mexico Gertler, (AGEs);AGEsmanagegrants(civil ofparentsinmonitoring Apoyoala Patrinos, works,schoolequipment,materials schoolperformanceand Gestion andRubioͲ forstudents,pedagogicaltraining, decisionͲmaking. Escolar Codina.  and performance Ͳ based  incentives Decreasedgradefailure (AGE)(Support 2006. forteachers. andrepetition.Positive toSchool impactontestscores Management LopezͲ  Program) Calva& Increasedtestscores. Espinosa, 2006. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 9  Program/ Country Authors Intervention Results/Findings Project a)Incentivegrantsforcommunities (1)ReductioninoutͲofͲ thattakeovermanagementof schoolchildrenand governmentͲfundedschools;block repetitionrate,(2) Nepal Chaudhury grantstiedtoperformancefor Increasedequity Community and Nepal schools;b)ScholarshipforoutͲof (disadvantagedcastes Support Parajuli. schoolchildrenfrompoor performedbetter),(3) Project 2010. households,c)Capacitybuildingfor Increasedstudent SMC,trainingforteachersin performance(higher instructionalplanning&delivery. avg.inTIMSSscience) a)IntroductionofSBM;schools Small,overallpositive develop(SIP)withparentsandthe effectonaverage communityusingstudent schoolͲleveltestscores. achievementandlearningneeds data.AnnualImplementationPlan Cristina Third (AIP)atbeginningofschoolyearand Ling,Nidhi Elementary reportcardsharedwiththe Philippines Khattri, Education communityatendoftheschool Shreyasi Project(TEEP) year,b)Trainingofheadteachersin Jha.2010. implementationofSIPandAIP,c) Schoolgrantsformaintenance, training,curriculumdevelopment, textbooksandoperatingexpenses basedonAIP. Schoolgrant(onlyinthefirstyear), Positiveimpactson comprehensiveschoolmanagementͲ teacherandstudent WholeSchool Blimpoand The trainingprogramtoprincipals, absenteeism,butno Development Evans. Gambia teachers,andrepresentativesofthe impactonstudent (WSD) 2011. community;provisionofschool learning managementmanual.   One key factor to keep in mind is that it takes time to achieve results through SBM. Evidence from the USA based on 232 studies examining schoolͲbased management shows that improvements do not become evident until about five years of implementation and not until the eighth year are they substantiallysignificant(Bormanetal.2003)(Figure3).   WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 10  Figure3.SBMimplementationͲyearsuntilfullimpactofintervention 0.6 0.50 0.5 AdjustedEffectSize 0.39 0.4 0.3 0.25 0.23 0.2 0.17 0.15 0.14 0.13 0.1 0 1Year 2Year 3Year 4Year 5Year 6Year 7Year 8Ͳ14'Years YearsofImplementation  Source:Bormanetal.2003  Another factor to keep in mind is that is that there are different depths to what is meant by devolving autonomy and accountability to the school level and different country contexts usually influence the design and composition of SAA reforms.  Often school grants are a popular way to provide resources directly to schools, but while they are a mechanism to transfer funds, that alone does not transfer autonomyordecisionͲmakingonhowthosefundsarespent,nordoesitimplyaccountability.Someforms ofSBMincludecommunityschoolswhicharethegrassrootsmethodofchoiceinCentralAmerica.Other formsofSBMthattaketheconceptsfurtherareautonomousschoolslikeAustralia’sIndependentPublic Schools,USCharterschoolsandUKAcademies.  There are some regional variations in context and emphasis of higher objectives of SAA. In studies from Africa for example, SBM reforms or community participation were often emphasized to meet the increasingdemandstoaccesssecondaryeducationandqualityofeducation.InEuropeandCentralAsia, theyfocusedonSAAinordertoimproveefficiencyandqualityofservicedeliveryinanenvironmentofa decliningschoolͲagepopulation.InEastAsiaandthePacific,thereisaninterestinSAAaspartofoverall decentralizationofgovernanceandanincreaseinparticipationatlocallevels(Takeda,DemasandShibuya 2014).  Typically,indevelopingcountriesseekingtoprovidebetteraccesstoeducation,thefirstkindsofresults experiencedasaproductofSAAreformsincludedecreaseinabsenteeismofstudentsandteachers.This may or may not help increase learning outcomes. Without effective oversight and capacity to take on additional responsibilities at the local level, school councils or parent associations may not be able to understandschoolandstudentresults enoughto knowthattheirschoolsare underperformingandthat school authorities or governments should be held accountable (Mansuri and Rao 2013; Hanushek et al. 2013).Thesystemremainsoutofbalance.  AformofcommunityschoolSBMwhereparticipationisintroducedtosolveaprincipalͲagentproblemis El Salvador’s EDUCO3 schools, which started in 1995 to address a gap in provision of education service from the central government after the civil war. The objective was to provide access to school for the poorest and most isolated rural communities. Other goals included supporting community participation ineducation;improvingthequalityofpreandprimaryschooling;andimprovingschoolͲlevelmanagement 3CommunityManagedSchoolsProgram WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 11  administrationbyallowingcommunitiestoidentifyandmanageschoolpriorities.Communityassociations weretrainedandgivenautonomyforadministeringfunds,hiringandfiringteachersandmonitoringand maintaininginfrastructure.Asexpectedthefirstimpactevaluationconfirmeddecreasesinabsenteeism. However,italsoconfirmedthatbyenhancingcommunityandparentalinvolvementinEDUCOschools,it improvedstudentreading scoresin comparisontotraditionalschoolswhichmayhavelongͲtermeffects on achievement (Jimenez and Sawada 1999). Other evaluations have shown significant increases in retentionrates(JimenezandSawada2003;2014).Theincreasedcapacityofparentsthroughtrainingand their involvement in management and oversight in some critical areas helped to foster autonomy with accountability.  Whilethe EDUCOSchoolsdidregistersomegainsforthe poorin termsofreadingscores,in generalthe CentralAmericanexperiencewithSBMisbetterknownforgainsininternalefficiencyindicatorsandless soinlearningoutcomes.ThismaybebecausethecontextualnatureoftheseSBMprogramsemphasized more the administrative and community participation aspects of reform and less on school and student assessmentthatunderthecircumstanceswerebettermanagedbythecentraltechnicalauthorityrather than at the school level.  Such an approach would not fix deficiencies in teacher knowledge and, by inference, increase learning.  In isolation, SAA activities may improve the performance of process variables,suchasschoolattendance,butmayyieldinconsistentresultsintermsofgainsintestscores.If SAAisconsideredasaschoollevelsystemthatincludestheperiodicassessmentofteachersandstudents and the incentives for improving teacher quality, then the impact of SAA activities related to increased autonomyandaccountabilitymayyieldimprovedlearningmoreconsistentlythanatpresent.  AsystemthathasdecentralizedonlysomeautonomytoschoolsisthecaseofMexico.ThePECProgram (Programa de Escuela con Calidad), which increased responsibility of parents by involving them in the management of school grants made the most difference in lowering repetition and failure rates in comparison to control schools (Skoufias and Shapiro 2006; Gertler et al. 2006 ).  Targeted training to parents in School Improvement Planning (SIP) and monitoring also significantly increased Spanish and math scores (LopezͲCalva and Espinosa 2006; Arcia, Kattan, Patrinos and RiveraͲOlvera 2013). A similar outcomewasreportedin preliminaryresultsofanimpactevaluationofSAA programsinNiger(Kunieda 2014).  Targeted training of school management committees in establishing learning goals for their schoolsandsupportingandmonitoringthemthroughtheSIPregisteredsignificantimprovementsintest scores.Withoutthistargetedtraining,theSAAprograms,whileuseful,didnotrecordsignificantincreases inlearningoutcomesattheschoolsthatparticipated.  OntheSBMspectrum,schoolsthataremanagedasautonomousgovernmentschoolsprobablyhavethe most freedom. They are government funded but operate with substantial independence and many are ownedandmanagedbytheirowngoverningbody.Astudycomparingautonomousgovernmentschools, privateschoolsandtraditionalpublicschools,foundthatdifferencesintheinstitutionalenvironmenthave a particularly important effect on the way schools are managed (Bloom et al. 2014). The autonomous governmentschoolsgarneredsignificantlyhighermanagementscoresthanpublicandprivateschoolsand thosehighermanagementscoresarepositivelycorrelatedwithbetterstudentoutcomes.Thisheldtrue for the OECD countries and Brazil.  An example is the UK Academies that came about during a 1988 EducationReformthatpromotedautonomousschools.Whentheprogramwasevaluatednineyearslater there was a significantly large achievement gain (0.25 standard deviation improvement in pass rates on standardizedexams)atschoolswhichoptedintotheprogramincomparisontothosethatdidnot(Clark 2009).Thisgaininlearningoutcomesrepresentsaboutoneyearofschooling.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 12  Some researchers argue that SAA reforms work better in developed countries in contrast to developing countriesmostlyduetolowcapacityoflocalparticipants(Hanushek,LinkandWoessman2013).Knowing the context can help to avoid these pitfalls. SAA reforms can take many shapes and forms and this is usuallyin response to the country’s political economy, education goals, performance issues, and history just to name a few contextual factors that may influence the policies and design of intended SAA measures.SAAreformsjustbecomeamenuofactivitiesiftheyarenotconnectedandbalancedviathe three A’s – autonomy, accountability and assessment. This is what makes the difference. Even here, however, there are no proven combinations and as education systems evolve, the mix of interventions andpoliciesmustadjusttostayinalignmentandachieveormaintainaclosedͲloopamongthethreeA’s.  TheThreeA’sandSABERSAA TheThreeA’sframeworkservesasthearchitectureoftheSABERSAAtool.Availableresearchsuggested fivemainpolicygoalsthatschoolautonomyandaccountabilityshouldmeetinordertoenableaclosedͲ loop system where autonomy, assessment, and accountability reinforce each other in order to produce anenablingmanagerialenvironmentthatpromotesbetterlearningoutcomes.Thefivemainpolicygoals thatarederivedfromthismodelandthatmatterforsuccessinSAAarethefollowing:  1. Levelofautonomyinplanningandmanagementoftheschoolbudget. 2. Levelofautonomyinpersonnelmanagement 3. Roleoftheschoolcouncilinschoolgovernance(participation) 4. Schoolandstudentassessment 5. Accountability  Thefirstpolicygoalfocusesonthedegreeofautonomythatschoolshaveinplanningandmanagingtheir budgets.  This is desirable because it can increase efficiency of financial resources and give schools flexibilityonplanningandexecution.Thesecondpolicygoalfocusesonthedegreeofautonomyaschool hasinpersonnelmanagementincludingprincipals,teachersandnonͲteachingstaff.Thethirdgoalfocuses on participation in school governance, and it is where parents can exercise real power as clients of the education system. The fourth focuses on the regularity of measuring student learning with the intent to use results to inform stakeholders and make adjustments (managerial, pedagogical and personnel). Finally,thefifthfocusesonusinginformationtopromoteaccountabilityandreinforcebettermanagement of financial, operational and learning outcomes. The five policy goals are broken down by 24 correspondingpolicyactions.Eachofthesepolicyactionsissupportedbyaseriesofquestionsthathelp usunderstandifpolicies/laws/andmanualsenablingtheseactivitiesexistandhowwelldevelopedthey are.  In this next section of the paper we will present the supporting evidence for what matters in school autonomyandaccountabilityaccordingtothefivepolicygoalsandtheirpolicyactions.  IV. WhatMattersMost?SAAPolicyGoals,PolicyActionsandEvidence  Understanding that schoolͲbased management (SBM) activities are part of a system significantly alters theconceptualizationofSBMindicatorsofautonomyandaccountability,sincenowtheyshouldbelinked in a way that achieves a closedͲloop system. The 3As framework helps diagnose the status of SBM in a givencountry,wheresomesubcomponentsofthesystemmaybeabsentorinnascentformwhileother subcomponents may already be wellͲfunctioning. The fact that some subcomponents may be at earlier WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 13  stages may simply be a reflection of the political economy of SBM in a given country, where particular social,legalorinstitutionalbarriersmaytakemoretimetoovercome.  x Thefirsttwopolicygoalsareaboutautonomyattheschoollevel,especiallytheauthorityofschool level stakeholders over school resources – budget and personnel. By giving authority to schools and school councils, where parents participate, SBM incorporates local incentives into the planningandresourceallocationprocess. x The third policy goal focuses on community participation and the role of the school council in schoolgovernance.Thisnotonlyincludesthecommunity’sabilitytohaveavoiceoroverseekey school governance functions, but also how well supported those communities are to define and understand their roles, build capacity, execute the school plan and do so in a transparent and inclusivemanner. x The fourth policy goal is about the routine use of school and student assessment results to continuallyreflectandmakepedagogical,operationalandpersonneladjustmentsforthepurpose ofimprovingperformance.Inthiscase,thesharingandanalysisofresultsatthevariouslevelsof theeducationsystemandtothepublicisimportantsothatallstakeholderscantakethenecessary actions. x Thefifthpolicygoalrelatestoschoolaccountability,whichiskeyforimprovingeducationquality and service delivery. This goal includes the policies that enable stakeholders to receive comprehensible information on their schools, provide oversight, comply with regulations, link rewardsandsanctions,andcreatefeedbackloops.  These policy goals reinforce each other. Improvements in their implementation have an impact on the performanceofotherpolicies.ThisisthesystemicnatureofSBMthatisassessedbySABERͲSAA.  The system’s approach to Autonomy, Assessment, and Accountability suggests that their related policy actions move along a continuum of strength, with some areas becoming stronger before others, but keeping in mind that attaining a balance among the three As helps to achieve a necessary closedͲloop system and reinforces relationships between the three areas to better support successful school level outcomes. By examining policy actions in terms of strength, one could anticipate the pace and depth of theSBMreforms.  PolicyGoalOne:LevelofAutonomyinthePlanningandManagementoftheSchoolBudget  This policy goal focuses on the degree of autonomy that schools have in planning and managing their budgets. There is a strong positive relationship between school autonomy and student performance (Barrera, Fasih, and Patrinos 2009). When fiscal transfers given to schools are under local control, authorityrestsattheschoollevel.Acombinationoflocalauthorityfigures—schoolprincipals,teachers, communities, parents—are then involved in monitoring those central budgetary allocations. School autonomy in the planning and management of the school budget is considered desirable because it can increase the efficiency of financial resources, give schools more flexibility in budget management, and giveparentstheopportunitytohavemorevoiceonbudgetplanningandexecution.Greaterautonomyat theschoollevelhelpsschoolsfightforcentralresourcessince theycanuse theindicatorsofassessment torenderaccountsofstudentperformanceandintheprocessusemoralsuasiontogetincreasedfunding fromthecentrallevel.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 14  Budgetary autonomy includes giving schools responsibility for negotiating and setting the salaries of its teaching and nonͲteaching staff and using monetary and nonͲmonetary bonuses as rewards for good performance. In centralized systems, teachers are paid directly by the Ministry of Education or the Ministry of Finance under union or civil service agreements. As a result, in centralized systems schools have less influence over teacher performance because they have no financial leverage over teachers. Inversely, if a school negotiates teachers’ salaries, as private schools routinely do, it may be able to motivateteachersdirectlywithrewardsforajobwelldone. Based on the review of the evidence on school autonomy and accountability policies, the SABERͲSAA framework has identified five policy actions that education systems can use to reach this goal of school autonomyinplanningandmanagementofthebudget. PolicyGoal1 PolicyActions Levelofautonomyintheplanning 1A.Legalauthorityovermanagementoftheoperational and management of the school budget budget. 1B.LegalauthorityoverthemanagementofnonͲteachingstaff salaries 1C.Legalauthorityovermanagementofteachersalaries 1D.Legalauthoritytoraiseadditionalfundsfortheschool 1E.Collaborativebudgetplanning  1A. Legal authority over management of the operational budget.  Most countries whose students performwellininternationalstudentachievementtestsgivetheirlocalauthoritiesandschoolssubstantial autonomyoverallocatingandmanagingresourcesandadaptingandimplementingeducationalcontent, orboth(Barerra,Fasih,andPatrinos2009).Inastudyofeightcountries4covering1,800schools,results show that schools with higher management scores are positively correlated with better pupil outcomes (Bloom, Lemos, Sadun, and Van Reenen 2014).  In particular, autonomous government schools (e.g. charter schools in the US and academies in the UK) have significantly higher management scores than regular government schoolsand private schools. The difference in management between both types of schools is closely linked to the strength of governance or having a strong accountability for student performancetoanoutsidebody,degreeofschoolleadership,andalongͲtermstrategyfortheschool.It doesnotappeartobeinfluencedbydifferencesinstudentcomposition,geographiccharacteristics,basic demographics,orcharacteristicsofschoolprincipals.  There are several ways for funds to be transferred to schools. They include: central allocation to local government, direct transfer to schools, block/school grants, and formulaͲfinancing, among others.  To someextent,themethodoftransferdependsonhowmuchdecisionͲmakingauthorityovermanagement ofthebudgetisdevolvedtothelocalorschoollevel.Evenattheschoollevel,someSBMprogramstransfer authority only to school principals or teachers, while others mandate parental and community participationthroughalegallyestablishedbody,likeaschoolcouncil(Barrera,Fasih,andPatrinos2009). Autonomyinthemanagementofthebudgetattheschoollevelisbeneficialforschooloperationsandfor accountabilitysincefundscanbeallocatedtorelevantareasinneedandmonitoredbylocalstakeholders.  Controlled experiments on school grants in The Gambia (Blimpo and Evans 2011) and Indonesia (World Bank 2011a), where decisionͲmaking authority was devolved to the school council to manage those grants, have demonstrated a variety of benefits. In The Gambia, schools that received a grant and 4Brazil,Canada,Germany,India,Italy,Sweden,UnitedKingdom,andUSA WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 15  managementtrainingforschoolstaffandparentsshowedlowerratesofteacherabsenteeismandhigher levelsofstudentattendance.  Indonesia’sBOSprogramsetupandempoweredschoolcommitteeswiththeauthoritytoplanandmake decisions over nonͲsalary operational expenditures. It gave schools block grants based on a perͲstudent formula and provided management training to school committees that were elected by the community into their positions. The block grants were used to pay for more student support activities and to hire more staff, with results showing a significant increase in teacher attendance and student scores in languageandmathematics(WorldBank2011a).  1B and 1C. Legal authority over management of nonͲteaching staff and teacher salaries. The enforcement of schoolͲbased management activities is a necessary but insufficient condition for improving learning. The interaction between teachers and students determines to a large extent what students learn. The quality of teaching and importance of teachers in student learning is confirmed throughvariousstudies(VegasandUmansky2005,WorldBank2004).Evidencefromthesestudiesholds thatSBMcanfosterteachereffectivenessifitaddressespersonalincentives,suchasifwagesaredefined locally which we address here, if hiring and firing decisions are made locally (see Policy Goal 2), and if teacherincentivesareunderstoodbetteratthelocallevel.  Autonomyincludesgivingschoolsresponsibilityfornegotiatingandsettingthesalariesofitsteachingand nonͲteaching staff and using monetary and nonͲmonetary bonuses as rewards for good performance. Moving the authority to determine salaries closer to the school level can be controversial and is more likelytobegoodpolicyinsituationswherecentralmanagementofteachersisnotworkingwellanddoes notshowprospectsofimproving.Authorityovermanagementofteachersalarieshasbeenlegallygranted tolocalauthoritiesinsomecountries(forexampleBulgaria,HongKong,andKazakhstan).InBulgaria,the governmentinstitutedschoolautonomyreformsin2007Ͳ08inanefforttoshiftawayfromasystemwhere centralgovernmentmanagedinputsandlackedoutcomemeasures.Inthenewsystem,schoolprincipals manageallfundsallocatedtotheschoolanddetermineanindividualteacher’sremuneration.Thereforms also introduced differentiated teacher pay based on teachers’ performance and effort. Under the new system, the principal makes the pay determination based on a centrally defined framework and specific criteria determined at the school level.  Additionally, while current legislation neither requires nor discourages the use of student assessment data for differentiating teachers’ pay, principals are increasinglyusingstudentassessmenttestresultsforthatpurpose(WorldBank2010).  Whiletherehavebeenstudiesfrombothdevelopedanddevelopingcountriesthatshowbenefitsofpay for performance, more field evidence is needed. However, if a school negotiates teachers’ bonuses or salaries, as private schools routinely do, it may be able to motivate teachers directly with rewards for a jobwelldone.  Nepalisacaseinpoint,wherecommunityͲmanagedschools(CMS)havevariouslegalrightswhenitcomes to managing teachers, including the right to link teacher salaries to school performance5. Nepal first introduced CMS in 1951 to fill a void in services in the education sector.  However, in 1971, the governmenttookbackfiscalandmanagerialresponsibility.Intheory,teacherswereaccountabletoastate that was far removed from the school and in reality the only substantial relationship between state and teacherwasthesalarypayment.In2001,duetooverwhelmingpublicdissatisfaction,Nepal’sgovernment 5CMShavethelegalrighttotransferregular(governmentͲrecruited)teachersbacktothedistrictheadquartersandtodirectly hireandfirecommunityͲrecruitedteachersaswell. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 16  decided to return schools to community management on a gradual and voluntary basis.  The impact evaluationonNepal’sprogramrevealsthatdevolvingmanagerialresponsibilitiestocommunitieshashad a significant impact on outcomes related to access and equity, and for system efficiency measured by grade progression and repetition, the treatment effects were both substantial6 (Chaudhury and Parajuli 2010).  In El Salvador’s EDUCO program, teachers’ wages were set by the local Educational Community Association (ACE) every year and were linked to their performance. This was in contrast to traditional schoolswhereteachersalariesfollowedafixedͲwagesystem.Theincentiveprovidedbyannualpayraises contingentonanindividualteacher’sperformancehadapositiveeffectonEDUCOschools.Inparticular, the effect of a teacher’s experience had a positive effect on a student’s continuation in school and the appropriate compensationforteachers hadanimportant effectonteachereffort(Jimenez andSawada, 1999&2003).  1D. Legal authority to raise additional funds for the school.  Greater community and parental involvement in school affairs can sometimes lead to the school receiving more private donations and grants in addition to funding that the school receives from the national government or from local taxes (WorldBank2007a).Theissuehereisthelegalauthoritytoseekfundsoutsideofthecentralgovernment budgetratherthantheneedtoraiseadditionalfundsfromparents.Thereisevidencethatuserfeesdeter the very poor from attending schools; however, there is also evidence indicating that fees empower parents to demand accountability (Bold et al 2013). This policy action assesses when schools can seek contributionsfromavarietyofpotentialfinancialsourcesincludingcityandlocalgovernments,domestic and international NGOs, and private donors. When combined with parent participation, this legal mandate reduces the chance of parents being pressured into giving more money to the school beyond what they want to contribute voluntarily.  Ensuring that local contributions preserve financial equity amongparentsrequiresthatgovernmentsadoptastrongcompensatorypolicyindistributingfundsacross regions. Under such conditions, local financing and modest user charges can boost performance by allowing parents and the community to exert greater control over school operations (King and Cordeiro 2005).Furthermore,iftherighttoraiseadditionalfundsisapproachedwithequityinmind,itcanimprove efficiencywithoutworseninginequality(KingandCordeiro2005).  1E. Collaborative Budget Planning and Preparation.  Parent collaboration in budget planning and preparationhasbeenrecognizedasapositiveinfluenceontransparencyandaccountabilityinthebudget preparationprocessattheschoollevel.Thisiscalledparticipatorybudgeting.Participatorybudgetingisa decisionͲmaking process through which citizens deliberate and negotiate over the distribution of public resources(Wampler2007). Traditionally, the preparation stages of the budget process are driven by agencies with some basic guidance on budget constraints and priorities from elected officials.   Agencies tend to prepare their budgetbasedonpreviousallocations,fosteringahighdegreeofinflexibilityintheallocationofresources (Moynihan2007).Participatorybudgetingmakesthebudgetingprocessmoretransparentandresponsive to citizens’ needs, empowering marginalized groups, making the budget more proͲpoor, and reducing corruption(MansuriandRao2013).  6Gradeprogressionisestimatedata15.6percentagepointincreaseandforrepetitiontherewasa10.6percentagepoint decrease. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 17  One major example of participatory budgeting is from Porto Alegre, the capital of the Brazilian state of RioGrandedoSul(MansuriandRao2013).Involvingcitizensinthediscussionsofbudgetplanninghelped improve public services, such as expanded access to basic services including sewerage systems, water, schooling, housing, and paved roads.  This approach can also be applied to service delivery in the education sector. If schools, including parents and community members, can participate in the school budgetprocesses,thentransparencyandaccountabilitywillbeimprovedwithpossiblegainsinresource efficiency. When the school budget is proposed by the school level, it can better reflect the needs of schoolsandleadtomoreefficientresourceutilizationandhighersatisfactionofschoolͲlevelstakeholders.  Table2.PolicyGoal1:Policyactions,indicatorsandevidence  PolicyActions DetailsofPolicyAction Evidence 1A Legalauthorityover Atwhatlevelistheauthority? Barrera,Fasih,Patrinos2009. managementofthe CanschoolsmanagenonͲsalary Bloom,Lemos,SadunandReenen operationalbudget expenditure?Underguidelines? 2014.BlimpoandEvans2011. Inconsultationw/community? WorldBank2011a.  1B Legalauthorityover Whohaslegalauthorityover VegasandUmansky2005.WB themanagementof nonͲteachingstaffsalary?Is 2004.WB2010.Chaudhuryand nonͲteachingstaff thereapayscalewith Parajuli2010.Jimenezand salaries guidelines?Whosetsthepay Sawada1999&2003. scale?Anyconsultation process? 1C Legalauthorityover Whohaslegalauthorityover Togetherwith1B(seeabove) themanagementof teachingstaffsalary?Istherea teachersalaries payscalewithguidelines?Who setsthepayscale?Any consultationprocess? 1D Legalauthorityto Canschoolsraiseadditional WorldBank2007.Boldetal2013. raiseadditionalfunds fundsandfromwhatsources? KingandCordeiro,2005. fortheschool 1E Collaborativebudget Cantheschoollevelmakea Wampler2007.Moynihan2007. planning proposalforthebudget?Will MansuriandRao,2013. higherlevelsofauthority considertheproposal,useitas areference,orasthemain guidetotransferresources?  PolicyGoalTwo:LevelofAutonomyinPersonnelManagement  This policy goal assesses policy intent in the management of school personnel, which includes the principal, teachers, and nonͲteaching staff. Appointing and deploying principals and teachers can be centralizedattheleveloftheMinistryofEducationoritcanbetheresponsibilityofregionalormunicipal governments.Infullydecentralizededucationsystems,schoolscanhaveautonomyinteacherhiringand removal decisions for their particular schools usually within acceptable standards set centrally or by the civilservicerules.Thisgivesaclearsignaltoteachersthattheschoolcouncilhasvoiceortheauthorityto weigh in on school personnel, which could cover hiring, overseeing salaries and performance, and WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 18  provisionofadditionalsupport.Thisrealizationshouldmaketeachersmoreresponsivetoparents’needs. PersonneldecisionͲmakingattheschoollevelregardingprincipalsislesscommonandisusuallyinstituted whenthereisveryweakcentralcapacitytodeliverandmanageeducationservices.  Based on the review of the evidence on school autonomy and accountability policies, the SABERͲSAA frameworkhasidentifiedthreepolicyactionsthateducationsystemscanusetoreachthegoalofschool autonomyinpersonnelmanagement. PolicyGoal2 PolicyActions The level of autonomy in 2A.Autonomyinteacherappointmentanddeploymentdecisions personnelmanagement 2B.AutonomyinnonͲteachingstaffappointmentanddeployment decisions 2C.Autonomyinschoolprincipalappointmentanddeployment decisions  2A.Autonomyinteacherappointmentanddeploymentdecisions.ThereareSBMprogramsthatcanbe described as “weak” which do not typically involve personnel decisions and there are those that are considered “strong” which do provide personnel autonomy to local stakeholders. Programs that can significantlyimproveoutcomes,especiallyforpoorandlowperformingschools,empowerschoolcouncils and/orparentstohaveaneffectonpersonneldecisions.Therearedifferentdegreesofeffectsize,butit isrecognized thata criticalaspectofSBM effectivenessismanagementofteachers,including theability toestablish incentivesand tofire nonͲperformingteachers(Brunsetal.2011).Astudyfrom Koreafinds that in providing schools with greater autonomy in hiring and staffing decisions, their principals and teachersfacestrongerincentivestodelivergoodstudentperformance.Greaterautonomycombinedwith keeping school principals and teachers accountable can be effective in improving student outcomes (Hahn,Wang,andYang2014).  Thecountrycontextfordeterminingthedegreeofautonomyinteacherappointmentanddeploymentat the local level is also critical.  When the ability of the central government to deliver services (such as quality education) is weak, policies allowing for strong personnel autonomy at the school level can be effectiveinensuringresponsiveandsatisfactorydeliveryofeducationtocommunities.Whencentralor regionalgovernmentauthorityismoreefficientinservicedeliveryandabletomeetqualitystandardsand beresponsive,thereislessneedforstrongpersonnelautonomyattheschoollevel.  Weak ability is often seen in postͲconflict countries where service delivery is lacking and there is little contact or accountability between the central authority and what is happening at the local school. For example,inCentralAmericainthe1990sandearly2000s,severalcountriesexperimentedwithvariations of El Salvador’s EDUCO model (PROHECO, PRONADE, etc.). EDUCO was first developed as a response to rural communities in El Salvador that were not adequately funded, managed or supported from the centrallevelduringandshortlyafterthecivilwar.Lackofaccesstopublicschoolsduringthecivilwaralso led many rural communities to organize their own schools by the community members themselves and request direct funding from the central government (Di Gropello 2006; Gillies, Crouch and Florez 2010). Given the crisis, communities were given wide autonomy, including hiring, replacing, and dismissing teachers according to their school’s needs. In the short term, these programs worked well and were successfulinincreasingenrollmentandreducingdropoutrates.Afghanistan’sPartnershipforAdvancing CommunityͲbasedEducation(PACEͲA)alsoallowsforlocallyrecruitedvillageteacherswhoareprovided witheducationalmaterialsandtraining.Resultsshowtheprogramsignificantlyincreasesenrollmentand WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 19  test scores among all children, eliminates the 21 percentage point gender disparity in enrollment, and dramaticallyreducesthedisparityintestscores(BurdeandLinden2012).  Even without a postͲconflict situation, allowing school councils to hire teachers on renewable contracts outside the civil service system and to influence their tenure may have reinforcing impacts on student learning when combined with local autonomy (Bruns, Filmer and Patrinos 2011). Kenya’s Extra Teacher Program(ETP)showedthatclasssizereductioncombinedwithhiringlocalteachersonshorttermcontract and increasing parental oversight led to significant increases in test scores (Duflo, Dupas, and Kremer 2007). Some Eastern European countries (e.g. Kazakhstan) have decentralized the authority to hire and replace teachers to the school principal. Spain and the United Kingdom are also examples of countries thathavemainstreamedtheprovisionofautonomytotheschooltohireandfireteachers.  Asthegovernment’sabilitytodelivergoodqualityservicesinanequitablewayincreases,itmaybecome less necessary for parents or school councils to have the highest degree of autonomy in the hiring and firingofteachers,yetitisstillimportantforthemtohaveavoiceandaffectchangeasmaybewarranted to ensure learning outcomes and responsiveness at the local level. This is evidenced by many European countries(Eurydice2007)thatmaybeconsideredtohaveweakpersonnelautonomyattheschoollevel.  In some school systems in Europe (e.g. Finland), where teachers are contracted under collective agreementorundercivilservicerules,thedegreeofautonomyoverpersonneladministrationislimited. In these countries, the central and local governments have in place a very stringent selection process in which most of the initial effort goes into selecting the best teacher candidates (Eurydice 2007). In these cases, one could argue that the role of school autonomy on personnel management is done at the very beginningthroughtheselectionprocessundertakenbylocalgovernments.  Under any circumstance, it is important to keep in mind that reforms that involve teacher incentives whetherindirectlythroughincreasingpersonnelautonomyattheschoollevelordirectlythroughteacher pay, are among the most politically sensitive. Creating coalitions for reform is critical and using informationtocommunicatethegoalsandbenefitsofthereformsisonewaytobuildunderstandingand buyͲinfromthevariouslocalstakeholdersincludingteachersandunions.  2B.AutonomyinnonͲteachingstaffappointmentanddeploymentdecisions.Personnelautonomythat maybeless politicallysensitive butstill allowstheschoolcounciltomakedecisionsaffecting thequality oftheirschoolistheareaofnonͲteachingstaff.Thisincludesadministration,support,andmaintenance staff. Decisions about recruiting, disciplining and dismissing nonͲteaching staff are taken within schools inmanyEuropeancountries(Eurydice,2007).Oftenitisthoseattheschoollevelwhowillknowbestthe needs of the school’s learning environment and if empowered, they can affect timely decisions that will supporttheschool’soutcomes.Bulgariahasbeensteadilyadvancingitsschoolautonomyreformswhich likeotherEasternEuropeancountrieshavebeenputinplacetogaingreaterefficiencyandqualityinthe educationsystem(Zafeirakou2004).Thesegoalshavebeenachievedinpartbydelegatingmorepowers in personnel autonomy to school principals who are in charge of hiring and firing of nonͲteaching staff (World Bank, 2010; Eurydice, 2007). The indirect effect is that better efficiency leaves more resources dedicatedtoimprovingoutcomes.  2C. Autonomy in school principal appointment and deployment decisions. School principals are instructional leaders as well as administrators who work with the local community and local education offices. An effective principal can be critical for the success of SBM and improving school and student performance. There is a small but statistically significant indirect effect of school leadership on school WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 20  achievement(HallingerandHeck,1996).Therefore,howschoolprincipalsareappointedanddeployedis animportantdecisionͲmakingpower.Thelevelofpersonnelautonomydecentralizedtoanintermediate leveloreventothelocallevelthroughtheschoolcouncil,givesmorecontrolorinfluencetolocalactors in determining a key leader at the school and holding that leader responsible as well as providing the principalsupportinreachingschoolgoals.  TheschoolautonomyreforminNicaraguawhichallowedschoolcouncilstohireandfireschoolprincipals had a positive effect on student test scores.  King and Özler (1998) studied the effects of Nicaragua’s school autonomy on student test scores in mathematics and Spanish. The authors found that de facto autonomy begun in 1990 had positive effects on student promotion and student achievement in math andlanguageinprimaryschoolandinlanguageinsecondaryschool.ThepositiveimpactsoflocallyͲhired principalspersistedformorethanadecade,untiltheautonomyprogramended(Arcia,PortaandLaguna 2004).  Two other examples are from Brazil and Chile, which implemented aggressive reforms that included student assessment and accountability and resulted in improved learning outcomes.  Both countries introduced a more transparent and competitive process for hiring principals using merit as the criterion insteadofpoliticalconnections.ManystatesinBrazilnowhiretheirprincipals,whoarerequiredtohave passedaschoolmanagementprogram,throughschoolelections(ElacquaandAlves2014).  Table3.PolicyGoal2:Policyactions,indicatorsandevidence  PolicyAction DetailsofPolicyAction Evidence 2A Autonomyinteacher Whohaslegalauthorityto Bruns,Filmer,Patrinos2011.Hahn appointmentand appoint,deployandtransfer Wang,andYang2014.DiGropello, deploymentdecisions teachers?Arethesedecisions 2006.Gilliesetal2010.Burdeand madeundercivilservant Linden,2012.Duflo,Dupas,and agreements?Whohasfinal Kremer,2007.Eurydice,2007. review? 2B AutonomyinnonͲ Whohasthelegalauthorityto Eurydice,2007.Zafeirakou,2004. teachingstaff appointanddeploynonͲ WBͲBulgariaSAReforms,2010. appointmentand teachingstaff?Docivilservant deploymentdecisions agreementsregulate appointments? 2C Autonomyinschool Whoappointsanddeploys HallingerandHeck,1996.Kingand principal principals?Whoevaluates Ozler,1998.Arcia,Portaand appointmentand principals'performance?How Laguna,2004.ElacquaandAlves, deploymentdecisions areprincipals'tenure,transfer, 2014. andremovaldetermined?Is thereaconsultationprocess?   WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 21  PolicyGoalThree:RoleoftheSchoolCouncilonSchoolGovernance In the context of school governance, the main objectives of school autonomy and accountability are to increaselocalstakeholderparticipation7inschoolmanagementactivities,promotelocaldecisionͲmaking, andallowforlocalvoiceandoversightattheschoollevel.Theparticipationoftheschoolcouncilinschool managementislinkedtoschoolautonomybecausetheschoolcouncilisacomponentofthelocalschool management team, which also links the school council to accountability through its part in the budget approvalandsupervisionprocess. There are several reasons why local stakeholder participation in school governance is important. First, parents have a natural incentive to improve their children’s education and they have a sound understandingoftheneedsoftheirchildrenandthelocalschool.Giventheopportunitythroughtheright mixofenablingpoliciesandsupport,parentandcommunityparticipationthrougharepresentativebody liketheschoolcouncilcanpositivelyinfluencetheeducationalenvironmentandoutcomesthroughcloser monitoringofschoolpersonnel,inclusiveandbettermanagementofresources,schoolevaluations,anda closer match between the school’s needs and its policies. Second, a wider participation of parent and communitymembersinschoolmanagementfostersmanagerialtransparencyandreducesopportunities for corruption (BarreraͲOsorio, Fasih, and Patrinos 2009). Third, participation of the school council can develop a higher sense of local ownership in schools. Finally, it is important to note that change management studies also have provided evidence that bringing stakeholders together to plan and implement meaningful activities also contributes to behavioral change in institutions, including schools. Collectiveschoolplanningactivitiescanprovideamutualvisionandsharedaccountabilityofwhatparents and school staff can commit in terms of support to the school.  These processes provide an enabling environmentforbettergovernance. Several studies by various development partners assert the importance of parent and community participationinschoolmanagement.ImpactevaluationssupportedbytheWorldBankhaveshownthat parent and community participation in school management have positive effects in increasing access to education (Di Gropello 2006, Chaudhury and Parajuli 2010), improving internal efficiency such as repetition and dropout (Jimenez and Sawada 2003, 2014; Skoufias and Shapiro 2006; Gertler, Patrinos, and RubioͲCodina 2006, Benveniste and Marshall 2004) and facilitating teacher and student attendance (Evans and Blimpo 2011; Jesse et al, 2010). However, there are also results that have not shown statisticallysignificantandpositiveeffectsontheseoutcomes.Intermsofstudentlearningoutcomes,the evidence on parent and community participation in school management is mixed. Schools with autonomous decisionͲmaking do not necessarily produce higher test scores; however, schools with greaterparentinvolvementthatarebetterequippeddohavebetterschooloutcomes(Gunnarssonetal 2009). For developing countries in particular, Contreras (2015) suggests that, developing institutional arrangements that allow parents to participate in academic content decisions at the school level has a greater impact on student learning than autonomous decisionͲmaking without enabling parent participation. There are various levels and formality of parent and community involvement in school governance and school activities. In West Africa, it was noted that Parent Associations or Parent Teacher Associations (PTAs)lackedthecapacitytomakerealchangeandbesustainable(BurkinaFaso,Senegal)(Shibuya2014). Theneedforamoreinstitutionalizedbodybeyondparentassociationshasbeenrecognizedasnecessary 7Localstakeholderscanincludetheschoolprincipal,teachers,parents,communitymembersand,insomecases,students. Theirrecognized,legalparticipationisusuallythroughaninstitutionalbody,namelyaschoolcouncil/parentcouncil/school managementcommittee/parentteacherassociation,etc. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 22  toallowparentandcommunityvoicestobebetterrepresented,andthishasresultedintheestablishment of such entities referred to as school councils, school management committees, school committees, schoolboardsandsoon.Theunderlyingpremiseforinstitutionalizedsupportisthatrelevantprocedures willbenewtothecommunitiesandthattheywouldrequiresupportinlegalandadministrativematters. Without the legal status, school councils cannot open bank accounts, receive financial transfers, and perform functions on the government’s behalf (Di Gropello, 2006). These constraints are a reflection of the limitations of PTAs or parent associations to manage school resources. The school councils are supposedtoensurethatschoolresourcesareusedmoreeffectively(LugazandDeGrauwe2010). Ingeneral,SBMprogramsdevolveauthoritytotheschoolleveloverthefollowingactivities:budgetissues such as allocating and overseeing budget; personnel management such as hiring and firing teachers; pedagogical issues such as selecting textbooks/curriculum; maintenance and infrastructure; and monitoringandevaluation(BarreraͲOsorio,Fasih,andPatrinos2009).Authorityoverbudget,personnel, and curriculum is stressed as crucial for those at the school level (Odden and Wohlstetter 1995; Wohlstetter 1995).  The election of school council members and their representation of parents and communitymembersinatransparentmannerareregardedasingredientsforsuccessfulSBMprograms. Based on review of the evidence on school autonomy and accountability policies, the SABERͲSAA framework has identified five policy actions that education systems can use to reach this goal of school councilparticipationinschoolgovernance. PolicyGoal3 PolicyAction RoleoftheSchoolCouncilin 3A.Participationoftheschoolcouncilinbudgetpreparation SchoolGovernance 3B.Participationinfinancialoversight 3C.Participationinpersonnelmanagement 3D.Communityparticipationinschoolactivities 3E.Communityparticipationinlearninginputs 3F.Transparencyincommunityparticipation  3A & B. Participation of the school council in budget preparation and financial oversight. The participation of the school council in school finances has several links to autonomy and accountability. First,itallowsfortheschooltorenderfinancialaccountstoparents,whichfostersefficiencyinresource use;itallowsforschoolcouncilstoseekadditionalfundsfortheschoolfromnonͲgovernmentalsources; itallowsforschoolstoarticulatethefinancialimplicationsoftheirworkplansinawaythatschoolcouncils canunderstand,anditcanhelpschoolcouncilsdemandmoreresourcesorbetterfinancialaccountability at other levels of government. In most cases, provision of a school grant or direct transfer of financial resources from the central government via regional or local governments to schools requires a schoolͲ levelbodytoprepareabudgetplan(proposalforexpenditure)andtooverseethebudget.  Evidence from Uganda and Mexico demonstrates the potential for positive effects of school council participation in budget preparation and financial oversight.  The Uganda study shows that a School ManagementCommittee’ssatisfactionwithschoolͲlevelplanningandbudgetingprocessesincreasesthe chances of higher student pass rates (Mulindwa, Habyarimana, and Bunjo 2013).  In Mexico’s Quality Schools Program (PEC), Parent Councils (Padres de Familia) are active in the preparation of the school budget. They prepare a plan for improving the school’s quality and then receive a fiveͲyear grant to implementtheactivitiesbasedontheschoolplan.ThePECgrantisallocatedtodifferentareasofschool operationsdependingonthepreferencesoftheparentcouncilandtheprincipalworkingtogether(SABER CR, World Bank, 2013b). Community participation in PEC schools significantly decreased the dropout, WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 23  failure, and repetition rates (Skoufias and Shapiro, 2006). Financial oversight is also a common function ofschoolcouncilsandisanimportantcomponentforaccountabilityandefficiency.Financialoversightby school councils was one of the common functions performed by school councils in Central America (Sawada,2003;DiGropelloandMarshall2005;DiGropello2006). 3C. Participation of the school council in personnel oversight. Personnel oversight is one of the key authoritiesthatschoolcouncilscanusetoholdteachersaccountable.Itmayincludepayingstaffsalaries, establishing incentives for teaching staff, hiring and firing teachers and administrative staff, supervising andevaluatingteachers,andfundingteachertrainingorotherformsofsupport(WorldBank2007). EvidencefromCentralAmerica(Sawada,2003;DiGropelloandMarshall2009;DiGropello2006)suggests that communities delegated with greater autonomy will do a better job of maximizing existing capacity byelicitingmoreeffortfromteachers.InKenya,providingschoolcommitteeswithfundstohireanextra teacher on a shortͲterm contract had a positive effect on learning. School committees were responsible for hiring the contract teacher and were free to replace or keep the original contract teacher based on performance(Duflo,Dupas,andKremer2007). Delegating personnel management authority to the school council has been shown to be beneficial for reducing teacher absenteeism and for inducing teachers to work harder. At the same time, as studies show, this could change the power dynamics at the school level (World Bank 2007b). Decentralized teachermanagementmaybeopposedbyteachersand/orteacherunionsbecauseitshiftsthebalanceof power in favor of parents and the school. Thus, necessary measures should be taken to deal with the politicaleconomyofschoolmanagementandtheschoolstakeholdersparticularlyforchangesinthearea ofpersonnelmanagementauthority. 3D.Participationoftheschoolcouncilinschoolactivities.Oneofthecoreobjectivesofdecentralization istoexpandtheinvolvementofactorswhodonotbelongtothecommunityofeducationprofessionals, in particular elected local authorities and community or parent representatives (Lugaz and De Grauwe 2010).Participationofparentsandcommunitymembersinschoolactivitiesthroughtheschoolcouncilis crucial to gain their support for school activities and to foster accountability. A commonly used tool to allow community participation in school activities is a school improvement/development/action plan (SIP). This tool can promote schoolͲlevel planning to reflect local voices and needs and form a basis for monitoringandevaluationofschoolactivities,whichisessentialforschoolͲleveldecisionͲmaking.Having aschoolplanthatfocusesonstudents’performance,staffinglevels,andinstructionalmaterialprovision increases chances for children’s pass rate as evidenced in Uganda for Grade 7 passes (Mulindwa et al., 2013).  In contrast, poor understanding of roles by parents and community members can be a major constraint forSBMandmayhavealowimpactonschooloutcomes(LugazandDeGrauwe2010;Banerjee,Banerji, Duflo, Glennerster, and Khemani 2008). This is why it is important for school councils to have the regulationsandthetoolstheyneedtocarryouttheirmissions,suchastheSIPtoolthatchannelsshared missionsandvisionsintofeasibleactivitiesinaparticipatoryway.Toassistschoolcouncilstohaveaclear understanding of their roles and how to develop school improvement plans, the availability of detailed instructions, guides, and manuals will be one of the enabling conditions to facilitate stakeholder engagementinschoolactivities.ThesetypesofinterventionsinMadagascarresultedinanincreaseintest scores and a four percentage point increase in student attendance (Jesse, Tan, Lassibille, and Nguyen 2010).Mexico’suseofsimplepostersandsimplestoryboardsassistedparents(literateandilliterate)to learnthebasicsaboutparentcouncilsandperformtheirdutieseffectively(Arciaetal2013).InSenegal, thegreaterfunctionalityofschoolmanagementcommittees(measuredbyimplementationofprocedural WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 24  policiessuchasholdinggeneralassembliesandsupportingvariablessuchastrainingandmonitoringvisits) isrelatedto betterlearningachievement,as measured bythe passrateoftheprimarygraduation exam (Yuki,IgeiandDemas,2015). 3E.Communityparticipationonlearninginputs.IthasbeenemphasizedthatschoolͲbasedmanagement should lead to improvement in teaching and learning. A study about schools and school districts that implemented SBM in the US, Canada and Australia, showed that people at the school site must have genuine authority over curriculum in addition to budget and personnel authorities as conditions for schoolͲbased management to improve school performance (Odden and Wohlstetter 1995; Wohlstetter 1995). According to Caldwell (2005), recent studies have highlighted the importance of local decisionͲ making regarding learning and teaching that meet the needs of students and take into account local priorities.OECDPISAdatashowthatatthecountrylevel,thegreaterthenumberofschoolsthathavethe responsibilitytodefineandelaboratetheircurriculaandassessments,thebettertheperformanceofthe entireschoolsystem,eventhoughitisnotalwaysthecaseforanindividualschool(OECD,2011).  ItcanbeobservedthatdevelopedcountriestendtohavedevolveddecisionͲmakingonlearninginputsto the school level (World Bank, 2007, 2009).  For instance, New Zealand delegated several pedagogical functions such as setting classroom hours by subject, selecting some textbook/curriculum, and determiningthemethodofinstruction.Incontrast,onlyafewdevelopingcountrieshavedevolvedthese functionstotheschoollevel.ElSalvadorandTheGambiaaretwoexamplesofdevelopingcountriesthat delegatedsomepedagogicalfunctionstoschoolcouncils,butingeneral,mostdevelopingcountrieshave only made partial advances in delegating pedagogical authority at the school level.  For example, in Guatemala, local authority is given only over the school calendar; whereas in Madagascar and Kenya, schoolscanselectsometextbooksormakesomecurriculumadjustments(WorldBank,2007;Barreraet al2009). There could be two reasons that account for less delegation of control over learning inputs. First, these pedagogical issues still tend to be controlled by the central level because of uniformity and standardization. Second, mostly it is not school councils but teachers who deal with pedagogical issues. Thisoftenpreventsschoolcouncilsfrominterveningwithpedagogicalissuesattheschoollevel.However, itisstillcriticaltoknowwhetherschoolcouncilsareinvolvedintheprocessofselectingortakingdecisions onlearninginputsthatcandirectlyaffectstudentlearningoutcomes. 3F.Transparencyincommunityparticipation.Ensuringtransparencyinselectingschoolcouncilmembers isacriticalentrypointforschoolͲbasedmanagementwithaccountability.Withouttransparency,itcannot be assumed that school council members are representing the voices of parents and the whole community. Democratic election is considered a way to ensure transparency in selecting school council members.Electingschoolcouncilmemberscanalso beaway toidentifyleaderswho,byvirtueoftheir socialpreferences,arelesswillingtocapturecontributionstopublicgoodsthanmostothersinthevillage, includingexternallyappointedvillageexecutiveofficers.InthecaseofTanzania,suchrelationshipsdonot existinrandomlydrawnresidentsorexternallyappointedvillage executiveofficers,whichsuggeststhat eitherelectoralselection,orthepowerofelectionstotransformthepreferencesofincumbents,produce leaders with proͲsocial preferences (Lierl 2014).  An impact evaluation of Indonesia’s school grant program,BatuanOperasionalSekolah (BOS)showsthatatreatmentgroupwiththe democratic election of school council members and the linkage with village representative councils, in combination with committeetraining,hadsubstantialpositiveeffectsonstudentlearning(Pradhan,M.,etal,2011).There are risks of “elite capture”, or that a few key stakeholders such as the school principal or a particular representative on the school management committee dominate decisionͲmaking powers (Mansuri and Rao2013;LugazandDe Grauwe2010;Nielsen,H.D 2007).However,suchriskcanbe mitigatedbyusing WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 25  termlimitsforschoolcouncilmembersaswellasopenandscheduledelectionstoreducedominanceby particularstakeholders.  Inaddition,holdinggeneralassembliesensuresthatschoolcouncilsareheldaccountabletoparentsand thewholecommunityfortheresultsofschoolimprovementplansandtheuseoftheschoolbudget.The impactevaluationoftheNepalCommunitySchoolSupportProjectshowspositiveeffectsonreductionin the proportion of outͲofͲschool children, and an increase in grade progression in primary grades. According to its Implementation Completion Report, the social audit conducted by a committee comprised of the chairperson of the PTA and two other parents was the most significant tool for monitoringprojectimplementation.Thesocialauditreportwastobepresentedtoandendorsedbythe parents’assembly.Thus,generalassemblyshouldbereinforcedbySBMpoliciesasaplatformtoholdthe schoolcouncilandschoolaccountabletoparentsandthewholecommunityfortheirperformance. Table4.PolicyGoal3:Policyactions,indicatorsandevidence  PolicyAction DetailsofPolicyAction Evidence 3A Participationofthe Whopreparestheschoolbudget? Mulindwa,Habyarimana,Bunjo, schoolcouncilin Doestheschoolcouncilhavea 2013.MexicoPECCountryReport, BudgetPreparation roleinplanningandpreparation 2013.SkoufiasandShapiro,2006. oftheschoolbudget?Arethese Sawada,2003.DiGropelloand responsibilitiessharedwiththe Marshall,2005.DiGropello,2006. schoolprincipal? 3B Participationin Doestheschoolcouncilhave Togetherwith3A financialoversight legalstandingasanorganization, orlegalauthoritytohaveavoice oroversightonbudgetissues? 3C Participationin Doestheschoolcouncilhave WB2007.Sawada2003.Di personnel legalrightorvoiceoverteacher GropelloandMarshall,2005.Di management appointments,transfers,and Gropello2006.Duflo,Dupas,and removals? Kremer2007. 3D Community Arethereformalinstructions, LugazandDeGrauwe2010. participationin manuals,ormandatesfor Mulindwa,etal,2013.Banerjee, schoolactivities organizingvolunteers?Dothese Banerji,Duflo,Glennerster,and guidelinesaddressimplementing, Khemani2008.Jesse,Tan, planning,orevaluatingactivities? Lassibille,Nguyen2010.Yuki,Igei, andDemas2015. 3E Community Doestheschoolcouncilhave OddenandWohlstetter1995. participationon legalauthoritytovoicean Caldwell2005.OECD2011.WB learninginputs opiniononlearninginputs?Does 2007.Barreraetal,2009. theschoolcouncilhavelegal oversightforalllearninginputs? 3F Transparencyin Arethereprovisionsforopen Lierl2014.Pradhan,etal,2011. community electionofschoolcouncil MansuriandRao,2013.Lugazand participation members?Aretheretermlimits DeGrauwe2010.Nielson,2007. orregularlyscheduledelections? Arethereguidelinesforcalling generalassemblies?  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 26  PolicyGoalFour:SchoolandStudentAssessment  In any education system, it is critical for stakeholders to know how well schools, teachers, and children are performing. Evidence shows that in schools where schoolͲbased management has worked well, a varietyofcommunicationmechanismswereusedtoshareinformationonassessmentresults.Thus,itis recommended to create a wellͲdeveloped system for sharing schoolͲrelated information with a broad rangeofconstituents(Wohlstetter1995).Thispolicygoallooksattheimportanceofschoolandstudent assessment—theirexistence,frequency,anduseofinformationgeneratedfromassessmentsformaking adjustments.  School assessments can have a big impact on school performance because they encourage parents and teachers to agree on indicators, scoring rules and ways to keep track of them. Student assessment is anotherimportantwaytodetermineifaschooliseffectiveinimprovinglearning.AkeyfunctionofSAA istheregularmeasurementofstudentlearning,withtheintentofusingtheresultstoinformparentsand society,andtomakeadjustmentstomanagerialandpedagogicalpractices.Withoutaregularassessment of learning outcomes, school accountability decreases and improving education quality becomes less certain.  Themainideaofthepolicyactionsforschoolandstudentassessmentisthattheschoolanditscommunity takeassessmentsasaroutinetasktobeperformed.Bymakingthistaskroutineitisboundtobecomea managerial tool and not a punitive tool. Results from PISA suggest that, when autonomy and accountability are intelligently combined, they tend to be associated with better student performance (OECD2011). Based on the review of the evidence on school autonomy and accountability policies, the SABERͲSAA framework has identified five policy actions for assessment and their links with autonomy and accountability.  PolicyGoal4 PolicyAction SchoolandStudent 4A.Existenceandfrequencyofschoolassessments Assessment 4B.Useofschoolassessmentsformakingschooladjustments 4C.Existenceandfrequencyofstandardizedstudentassessments 4D.Useofstandardizedstudentassessmentsforpedagogical, operational,andpersonneladjustments 4E.Publicationofstudentassessments  4A. Existence and frequency of school assessments. School assessment can be defined as any type of assessmentthatcollectsschoolinformationandmeasuresitsachievementeitherdirectlyorthroughthird parties.  There are many forms of school assessments such as school inspection, teacher appraisal by school directors, school selfͲevaluation or own quality assurance, school report cards, and testͲbased ranking.Schoolassessmentmaytakemanyforms,butitshoulddealwithtwomainissues:(i)theintegrity of the assessment process and (ii) the use of yearly assessment of teachers and students to evaluate school performance. Without the existence of and a suitable frequency in administration of school assessments, the education system, school, and community cannot understand how the different parts oftheschool(teachers,resources,curriculum,infrastructure,etc.)areeithersupportingornotsupporting studentlearning.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 27  EducationqualityassuranceintheUnitedKingdomeffectivelyintegratesstudentassessmentresultsinto both selfͲevaluations and school inspections. Schools participate in selfͲevaluations on an ongoing basis thattypicallyincorporatedataonattainment,predictedresults,andvalueͲaddedscores.Inaddition,the OfficeforStandardsinEducation,Children’sServicesandSkills(Ofsted2014)externallymonitorswhether schools have met standards in learning outcomes, student wellͲbeing, school facilities, and other areas (typicallyevery5years).SchoolsarealsoabletoanalyzetheirperformancewithRAISEonline,atoolthat providesinteractiveanalysisofschoolandstudentperformance,aswellascomparisonstopeers(Ofsted 2014).  4B.Useofschoolassessmentsformakingschooladjustments.Schoolassessmentsarethevehiclesused by schools to determine their needs for changes in pedagogical practices and to determine the training needs of their teachers (teacher quality). For these purposes, the regular management and analysis of school data plays an important part. The role of EMIS is linked to accountability as it makes it easier to reportonschoolassessmentresultsincludingindicatorsofinternalefficiencyandstandardizedtestscores (Bruns,Filmer,andPatrinos2011).ItalsoprovideslocaldecisionͲmakerswithkeyinformationneededfor improvingefficiencyandeffectiveness.  In the absence of centralized systems of school assessments, it may be possible to rely on school report cards.Severalreportcardprogramshavehadsuccessinincreasingparentalknowledgeaboutthequality of instruction in schools and have helped to raise parents’ voice in school matters at the school council and state levels. They can also increase awareness among school personnel about their schools’ instructionalqualityandacademicperformance–Brazil,Uganda,Nigeria,CentralAmerica,Namibia–which canleadtopedagogicalandoperationaladjustments(Bruns,Filmer,andPatrinos2011).  TheexperienceofParanáStateinBrazilinusingschoolreportcardsfrom1999to2002suggestspositive effects(Winkler2005).Ashort,simplesummaryofindicatorswasdisseminatedtoparentsandteachers through local workshops and through publication in the state’s education newsletter. The report cards servedasamanagementtoolattheschoollevelhelpingprincipals,teachersandschoolcouncilstomake managerial decisions or changes to practices. Teachers and parents engaged in discussions on how they mightimproveschoolperformanceandthroughtheschoolcouncilsincreasedtheirvoiceinpolicydebates abouteducation(Winkler2005).  In Liberia, a randomized controlled trial was conducted in 180 schools where a group of 60 schools received a full treatment that included the Early Grade Reading Assessment (EGRA) along with teacher training in the assessment of reading performance, frequent pedagogical support, and books and pedagogicalmaterials.Asecondgroupof60schoolsreceivedalighttreatment,inwhichonlytheschool reportcardsbasedonEGRAweredistributedamongparentsandthecommunity.Thecontrolgroupof60 schools did not get any of the above interventions (Piper and Korda 2011). The fullͲtreatment students showedasignificantlyhigherimprovementinreadingabilitythanthelightͲtreatmentandcontrolgroups, andthelightͲtreatmentgroupshowedbetterresultsthanthecontrolgroup.  4C. Existence and frequency of standardized student assessments. The experience of highͲperforming countries on PISA indicates that education systems with standardized student assessment tend to do betterthanthosewithoutsuchassessments(OECD2011).Systemiceffortstomeasurelearningoutcomes are important for all countries.  Some developing countries have dabbled in standardized student assessments but fail to make them a part of their education strategy and do not implement them with regularity or maintain them over time (Clarke 2012).  An important difference between one time assessmentsandasustainedassessmentsystemisthatonetimeassessmentsonlyprovideasnapshotof WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 28  studentachievementlevelswhileanannualassessmentorthoseprovidedwitharegularfrequencyallows for the possibility of monitoring trends in achievement and learning levels over time.  Additionally, research shows a weak but positive link between the uses of data from large, systemͲlevel assessments toholdschoolsandeducatorsaccountableandforbetterstudentlearningoutcomes(Clarke,2012).  4D.Useofstandardizedstudentassessmentsforpedagogical,operational,andpersonneladjustments. The existence, use and publicity of school and student assessment for making school adjustments including pedagogical, operational and personnel aspects are key enabling conditions for SBM to work better. More successful schools have teachers and administrators who have formed a professional learningcommunitythatfocusesonstudentwork(assessment)andbasedoninformationresultingfrom thoseassessments,theychangetheirinstructionalpractice(pedagogicalsupportforlearning)inorderto getbetterresults(NewmannandWehlage1995;FullanandWatson1998).Totakefulladvantageofthe benefits to learning, it requires that teachers develop better assessment literacy (Hargreaves and Fullan 1998). Therefore it is important to build capacity of principals and teachers to examine and accurately understand student work and performance data to be able to develop corresponding school and classroom plans and to alter conditions necessary to achieve better results (Hatch 2013, Fullan and Watson1998).  A study by DarlingͲHammond and Wentworth (2010) noted that among other positive effects, student assessment activities in high performing education systems regularly provide feedback to students, teachers and schools about what has been learned and identify information that can shape future learning.  Information in this case is used to enable better decisionͲmaking rather than to leverage accountability. Data resulting from student assessments may also assist in addressing inequities. In educationsystemsthatusestandardizedstudentassessments,PISAscoresamongschoolswithstudents from different social backgrounds differ less than in systems that do not use standardized student assessments(OECD2011).Testscoredisclosurealsohasthepotentialtochangethebehaviorofteachers andschoolmanagersasnotedinprivateschoolsbyaffectingthemarketincentivesfacedbysuchschools (Camargoetal2014).  Differences among schools on what types of information they receive about student performance (including comparative analysis) and on whether they use this information for making school improvements (pedagogical, operational, and personnel) matter for learning results (Yuki, Igei, Demas 2015).  A study in Senegal investigating SBM policies and their implementation found that in order to makeSBMmorefunctionalforbetterlearningresults,itisimportantnotonlythatschooldirectorsknow there are comparative analyses and use student assessment results for schoolͲlevel improvement, but also to increase the functional level of school councils, as well as to ensure a better implementation of the budgetary autonomy decentralized to communes. The study confirmed that one of the key policy messages of the SABERͲSAA framework, the 3 As working together in balance enable better learning results,isvalidforruralpublicschoolsinSenegal.  4E. Publication of student assessments. Making student assessment results available at the different levelsoftheeducationsystem(central,regional,municipal)aswellasattheschoollevelandtothepublic canhaveapositiveeffectonstudentperformance.SomefindingsontestͲbasedaccountabilityintheU.S. suggestthatsimplyreportinginformationaboutaverageschooltestscoresledtoincreasedperformance (Hoxby 2001, Hanushek and Raymond, 2003). Additionally, education systems in which schools publish test results and have more autonomy over resource allocation perform better than schools with less autonomy(OECD2011).  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 29  Table5.PolicyGoal4:Policyactions,indicatorsandevidence  PolicyActions DetailsofPolicyAction Evidence 4A Existenceand Howoftenisschoolperformance Ofsted2014. frequencyofschool assessed?AreMinistryofEducation assessments criteriausedtoassessschools?Are specialtopicsaddressedduringschool assessment,suchaspoverty,nutrition, ormigration? 4B Useofschool Doschoolsuseschoolassessmentsfor Bruns,Filmer,andPatrinos assessmentsfor makingpedagogical,personnel,or 2011.OECD2011.Winkler makingschool operationaladjustments?Areschool 2005.Brooke2005.Piper adjustments assessmentresultsanalyzed?Whois andKorda2011. mandatedtoreceiveschool assessmentresultsand recommendations? 4C Existenceand Howfrequentlydostudents OECD,2011.Clarke,2012. frequencyof participateinstandardized standardizedstudent assessments?Arestudent assessments assessmentscarriedoutonasample orcensusbasis? 4D Useofstandardized Arestudentassessmentresultsused NewmannandWehlage studentassessments tomakepedagogical,personnel,or 1995.FullanandWatson forpedagogical, operationaladjustments?Who 1998.HargreavesandFullan operational,and analyzesstudentassessmentresults, 1998.Hatch2013.DarlingͲ personnel andwhoismandatedtoreceive HammondandWentworth, adjustments studentassessmentresultsand 2010.OECD2011.Camargo, recommendations? Camelo,Firpo,andPonczek 2014.Yuki,Igei,Demas2015 4E Publicationofstudent Whoismandatedtoreceiveresultsof Hoxby,2001.Hanushekand assessments studentassessments? Raymond,2003.OECD, 2011.   PolicyGoalFive:SchoolAccountability  AccountabilityisattheheartofschoolͲbasedmanagement.Thesystemicconnectionbetweenbudgetary and personnel autonomy, parent participation in the financial and operational aspects of a school, and the measurement of learning outcomes are all aimed to reinforce accountability. Only by being accountabletolocalstakeholderscaneducationalqualitybesustainable.  AwayinwhichSAAcantheoreticallychangeeducationaloutcomesisbypromotingmoreinvolvementby the community and parents in the school and by holding accountable and monitoring those making decisions about school management (World Bank 2007a).  Ways to increase voice, transparency and accountability include some innovative tools such as public expenditure tracking, school report cards, communitymonitoringandsocialaudits(GaventaandMcGee2013).  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 30  In order to be consistent with the use of assessments as managerial tools, accountability should also be routineandobjective.Tothateffect,theproposedpolicyactionsrelyontheschool’ssocialandeconomic contexttomakecomparisonswithotherschoolsinsimilarconditions.Themainideaisthataccountability hasacontextandthattheschool’scontextisimportantforevaluatingprogress.Asaresult,teachersmay feelencouragedtoseeschoolassessmentstakingintoaccountinternalprogressthroughouttheyears,as wellasprogressrelativetoschoolswithsimilarsocialandeconomicconditions.  The following policy actions address aspects of accountability that can be implemented within the frameworkofschoolautonomyandaccountability.  PolicyGoal5 PolicyActions Accountability 5A.Guidelinesfortheuseofresultsofstudentassessments 5B.Analysisofschoolandstudentperformance 5C.Degreeoffinancialaccountabilityatthecentral,regional,and schoollevels 5D.Degreeofaccountabilityinschooloperations 5E.Degreeoflearningaccountability   5A. Guidelines for the use of results of student assessments.  An important defining feature of informationͲforͲaccountabilityinterventionsisthattheyfocusontheuseofinformationastheinstrument for change (Bruns, Filmer, Patrinos, 2011). There is reason to believe that if wellͲinformed, parents will trytoholdtheschoolsaccountablefortheirchildren’slearningoutcomes.However,Hanushek,Linkand Woessman(2013)arguethatthesuccessofautonomyreformsmaydependonthegenerallevelofhuman capital which affects the quality of parental monitoring.  Some studies show that low adult literacy communitiesmayhaveanegativeeffectornoeffectonstudentlearningoutcomescomparedtovillages with a high adult literacy rate that may post gains, as was the case in The Gambia (Blimpo and Evans, 2011). Regardless of literacy levels, however, evidence suggests that a mechanism to supply accurate information to communities about the relative performance of their children and schools could be desirable (Blimpo and Evans 2011). Establishing guidelines for packaging and use of school and student results in the appropriate form is important for reaching all parents so that they may be able to understand, react, and take action if need be. Communication of information alone is insufficient, but information that is perceived as actionable in an enabling environment can motivate collective action especiallyifcommunityvoicetriggersresponsivenessoftheauthorities(Fox,2014).  5B. Analysis of school and student performance.  Accountability with a context means that the publicationoftheresultsofstudentperformanceandthepositionoftheschoolrelativetoyearspast,as wellasinrelationtoschoolsinsimilarconditionscanbeempoweringtoteachersandparents.Therefore, it is important not just to present scores and other achievement data, but to conduct a comparative analysisthatoffsetsdisadvantagessuchassocioeconomicstatus,sizeandtypesofschools,acrossregions, andforpreviousyears.Otherwise,forexample,schoolsinpoorareasthatperformbadlyonachievement tests because they receive lowͲincome students may be discouraged from improving teaching quality, sinceschoolrankingdoesnotreflectalltheeffortmade(McEwen,Urquiola,andVegas2008).  InPakistan,anexperimentwithreportcardsin823schoolscovering12,000childrenwasshowntohave helped increase test scores by 0.10 standard deviations and reduced private school fees by 23 percent (Andrabi, Das, and Khwaja 2012). The report card was designed in collaboration with parents and the schools.Parentswantedtoknowtheirchild’sscoreandhis/herrankrelativetootherstudentsintheclass, WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 31  theaveragescoreforeachschoolintheirvillage,andthescoresbycategory(wordrecognition,sentence building,etc.)sotheywouldknowinwhichcategorytheirchildneededmorehelp. AnotherexamplecomesfromNigeria.InKanoState,simplifiedreportsshowingschoolͲspecificindicators comparedwithlocalareaandstateͲwidedatawereusedbylocalstakeholdersandincreasedtransparency in the management of the system. Furthermore, the use of the comparative data in the reports helped to strengthen accountability links between the communities and schools as well as between the central andlocalgovernments(WinklerandHerstein2005).  5C.Degreeoffinancialaccountabilityatthecentral,regional,andschoollevels.Thedegreeoffinancial accountability is important because it can be used as a source of support for teachers and the principal. Parents tend to look at teachers and principals more favorably if they are aware that the production of good results is tied to adequate funding and management of those funds. As academic accountability becomes more routine, financial support can be forthcoming. Participation of the school council in the development of the school improvement plan encourages ownership and deepens local stakeholders’ knowledgeofschoolgoals.Thisinturnenablesthecouncilandwidercommunitytomonitortheinputs, outputs,andoutcomes.Keydecisionsandchangesaboutspendingintermsofmagnitudeofspending,on what items, and who made the investment decisions are critical for maintaining financial accountability andkeepingontrackwitheducationoutcomegoals(WorldBank2007a).  The education budget represents the plan for implementing education policy.  Reported education expenditureatalllevelsofgovernmentalongbudgetprioritiesallowsschoolfinancesystemstomonitor the success of budget execution and make adjustments as necessary (World Bank 2011c). Internal and external audits can provide regular feedback to education authorities on management of funds. Establishing rewards and sanctions for financial and operational compliance provides a clear framework and incentives for transparently executing the budget and implementing the corresponding education goals.Consequences,suchasimprovedsupervision,dismissal,orsalarycutsshouldexistforfailedinternal audits(WorldBank2011c).Inadditiontoformalauditmechanisms,disseminationoffinancialinformation totheschoolandthepubliccanbeaneffectivemonitoringtool.AnexamplefromUgandademonstrates thatreportingonschoolgrantsthroughanewspapercampaignboostedtheabilityofschoolsandparents tomonitorlocalofficials’handlingofthegrants.Thegainsinfinancialknowledgeabouttheschoolgrants andincreaseinmonitoringreducedthediversionoffundsfrom80percenttolessthan20percentovera sixͲyearperiod(ReinikkaandSvensson2004).  5D. Degree of accountability in school operations. The introduction of consequential accountability systemshasaclearlybeneficialimpactonoverallperformance(HanushekandRaymond2005).Avariety of approaches to using information for accountability can be used  including report card programs discussed earlier, testͲbased score cards and participatory public expenditure or inputͲtracking (Bruns, Filmer, Patrinos 2011). For example, preliminary research has found that the improvements introduced to Chile’s voucher program combined with the testͲbased accountability policies provided schools with incentivestooperatemoreefficiently(ElacquaandAlves2014).AnotherexampleisthePhilippineswhere publicresourcetrackingwasusedtoincreaseaccountabilityoftextbookdistribution toschoolswiththe goal of increasing learning outcomes. ParentͲteacher associations and other local groups were used to verify and report on the delivery of textbooks. Anecdotal evidence suggests the program successfully reduced corruption in school operations and by 2005, all textbooks produced were delivered in comparison to a loss of 40 percent in 2001 (Bruns, Filmer, and Patrinos 2011). Efforts like this could be expandedtoschoolbuildingsandfurnitureaswell.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 32  5E. Degree of learning accountability. Despite design flaws in most existing accountability systems, Hanushek and Raymond, 2003 find that they have a positive impact on achievement in US schools. However, the impact holds only for states attaching consequences to performance.  Those that only provide information through report cards without attaching consequences to performance get similar resultstothosewithnoaccountability.Workstillneedstobedoneondeterminingthebestsetofrewards and sanctions.  Similarly, in Chile, a study analyzing the effects of accountability pressures on teacher policiesandpracticesinlowͲperformingschoolsfindsthattheimpactofthePreferentialSchoolSubsidies Law, 2008, which offers an additional subsidy to schools that serve the most vulnerable students, has beeneffectiveingeneratingincentivesforschoolstoseekstrategiestoimprovetheirresultsasquicklyas possible.Participatingschoolsmustmeetaseriesofminimumacademicperformancestandardsandface sanctionsiftheydonot(ElacquaandTreviño2008).  Table6:PolicyGoal5:Policyactions,indicators,andevidence  PolicyActions DetailsofPolicyAction Evidence 5A Guidelinesfortheuse Doguidelinesexistfortheuseof Bruns,Filmer,andPatrinos, ofresultsofstudent studentassessmentresults? 2011.Hanushek,Link,and assessments Whichlevels(schooland Woessman,2013.Blimpoand government)dotheseguidelines Evans,2011.Fox,2014. concern? 5B Analysisofschool Arethereprovisionsforthe McEwen,Urquiola,andVegas andstudent comparativeanalysisofstudent 2008.Andrabi,Das,andKhwaja performance assessmentresults?Are 2009.Bruns,Filmer,and comparativeanalysesofstudent Patrinos2011. assessmentresultscarriedoutat differentlevels?Arecomparative analysesmandatedtobe distributedtoparentsoravailable online? 5Ci, Degreeoffinancial Areregulationsinplacefor WorldBank2007a.Reinikka Cii, accountabilityatthe complyingwithrulesoffinancial andSvensson2004.World Ciii centrallevel; managementandtransparency, Bank2011c. regional/municipal reportingtothosewithoversight level;schoollevel authority,andlinkingrewardsand sanctionstocompliance? 5D Degreeof Areregulationsinplacefor HanushekandRaymond2005. accountabilityin complyingwithrulesoffinancial Bruns,FilmerandPatrinos, schooloperations managementandtransparency, 2011.ElacquaandAlves2014. reportingtothosewithoversight authority,andlinkingrewardsand sanctionstocompliance? 5E Degreeoflearning Isthereamandatetosimplifyand HanushekandRaymond2003. accountability explainresultsofstudent ElacquaandTreviño2008. assessmenttothepublic?Howis feedbacksolicitedfromparents andcommunitymembers?  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 33  V. ImplementingtheSABERFramework  SABERInstrumentandMethodology SABERͲSAAisadiagnostictoolthatbenchmarkseducationpoliciesaccordingtoevidenceͲbasedglobal standards and best practice. The School Autonomy and Accountability (SAA) tool is part of SABER (Systems Approach for Better Education Results). SABER helps countries systematically examine and strengthen the performance of their education systems to achieve learning for all (Rogers and Demas 2013).TheWorldBankhasbeenworkingwithpartnersaroundtheworldtodevelopdiagnostictoolsthat benchmark education policies according to evidenceͲbased global standards and best practice. By leveragingglobalknowledge,SABERfillsagapintheavailabilityofpolicydata,informationandknowledge onwhatmattersmosttoimprovethequalityofeducation.  Thepolicyactions/indicatorslistedinthepreviousSAApolicygoalsectionhelpustoidentifykeyfeatures of effective service delivery systems. Indicators are identified based on a combination of criteria, including: x empiricalresearchonthecharacteristicsofeffectiveschoolautonomyandaccountability x experienceofhighͲperformingsystems x theoryorgeneralconsensusamongexpertsregardingeffectiveservicedelivery x rationalconnectionsbetweenpolicyevidenceandindicatorapplication  SABER SAA benchmarks a country’s policy intent based on a country’s policies, laws, and guidelines using a standardized protocol.  The data collection process includes a review of laws, policies, and manuals currently in place as an enabling environment for school autonomy and accountability. The SABERͲSAAteamhasdevelopedaninstrumenttocollectdataoneachcountry’spolicies.Answerstothe data collection instrument (DCI) include references from education legislation, policy documents, and explanationsagainstwhichresponsesareverified.ApointofemphasishereisthattheaimoftheSABER SAApolicyintenttoolistoassessacountry’sofficialandestablishedpolicies.Anadaptationofthistool– the SABER SAA policy implementation tool Ͳ has also been developed to determine onͲtheͲground implementation of these policies at the local level. Education policies regarding school autonomy and accountability can be found in a range of documents such as national and local constitutions, education acts and laws, memoranda of understanding, official education policy documents, school regulations manuals,andeducationreformdocuments,amongothers.Theofficialdocumentsvaryforeachcountry. Oncethepoliciesandlawshavebeencollected,thedataareanalyzedandscoredusingarubricforeach correspondingpolicygoal.Theserubricsallowforcountrypoliciestobescoredbyeachindividualpolicy action (see Appendix 1 for a list of all policy goals and policy actions and Appendix 2 for the detailed rubrics).  Asdiscussedpreviously,eachpolicygoalconsistsofasetofkeypolicyactions.Foreachpolicyaction,the country will be benchmarked into one of four levels of development: 1 (Latent), 2 (Emerging), 3 (Established),or4(Advanced):  • Latentisthelowestperformancelevelandreflectsthatpolicyisnotinplaceorthereislimited engagement; • Emergingrepresentssomegoodpracticewithpolicydevelopmentstillinprogress; • Establishedrepresentsgoodpracticewithsomelimitations • Advancedrepresentstheinternationalbestpractice WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 34  The country team or consultant first collects information in order to answer the questions in the DCI. OncetheDCIiscompleted,thescoringrulesthatcorrespondtotherubricsareapplied.Usingtherubrics, alevelofdevelopmentandscoreisassignedtothepolicyactionswithineachpolicygoal: Latent   =1point Emerging  =2points Established  =3points Advanced  =4points Thescoreforeachpolicygoalisthencomputedbyaggregatingthescoresforeachofitspolicyactionsto the policy goal level. The Data Collection Instrument was designed to provide equal weight to each indicator. For example, a hypothetical country receives the following scores for one of its policy goals: Policy Action IA = 2 points; Policy Action IB = 3 points; Policy Action IC = 3 points; Policy Action ID = 3 points;PolicyAction IE=2points.The hypotheticalcountry’soverallscorefor thispolicy goalwouldbe: (2+3+3+3+2)/5=2.6.Thefollowingscaleisusedtodetermineadevelopmentlevelforthepolicygoal: Latent: 1.00–1.75 Emerging: 1.76–2.75 Established: 2.76–3.75 Advanced: 3.76–4.00  Forthisexample,thepolicygoalscoreis2.60placingitintheEmergingrange.  ApplyingtheSABERinstrumentandcomparingcountriescanhelpgovernmentsmakeevidenceͲbased policy decisions. Through the application of SABERͲSAA, countries will gain a better understanding of how well developed the set of policies are to foster managerial autonomy, assess school and student performance, and use information from assessments to promote accountability. The scoring allows for easy identification of strengths and weaknesses of the school autonomy and accountability policies, in reference to eachpolicy goal.Oncethedataare collectedandanalyzed,and scoresare produced,then theresultscanbeusedtoinformlargerstudiesinadditiontothetypicalproductwhichisacountryreport. SABERͲSAA produces a country report, summarizing the results of this benchmarking process and the policyimplicationsinconnectionwiththecountry’sparticulareducationalcontext.TheSABERreportalso includes policy recommendations to guide countries in further developing an enabling environment for SAAortoactasabasisforinͲcountrydialoguetakingintoaccountthelocalcontext.Thus,theSABERtool is to be used not as prescriptive policy assessment but rather as an informed policy assessment of the country policies in reference to the current knowledge of effective approaches. This information is compiled in a comparative database where interested stakeholders can access framework papers, detailed country reports, and other resources describing how different education systems have approached school autonomy and accountability  with a view  to  improving  learning outcomes http://saber.worldbank.org/index.cfm.    WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 35  Conclusions LearningforAllisacommongoalforalldevelopedanddevelopingcountriesaswellastheWorldBank’s EducationSectorStrategy2020.ItiscriticallyimportanthowSABERͲSchoolAutonomyandAccountability can help to identify the enabling environment that can lead to better learning. There are two important conclusionswhichshouldbeemphasizedfromthispaper.OneconclusionisthatbalancingtheThreeAs, namelyautonomy,assessment,andaccountability,matters.SAAreformsjustbecomeamenuofactivities if the Three As are not interconnected to form a closedͲloop system. This is what makes the difference. Implementingautonomyoraccountabilityalonehasnoadditionaleffectonlearning.IfSAAisconsidered asaschoollevelsystemthatincludestheperiodicassessmentofteachersandstudentsandtheincentives for improving teacher quality, then the impact of SAA activities related to increased autonomy and accountability may yield improved learning more consistently than if implemented in isolation. School autonomy must be complemented with school accountability to promote academic excellence. This has been well documented through various impact evaluations. We also know that the highest PISA scores comefromcountrieswhereautonomyandaccountabilityareimplementedtogether(OECD2011).  The second conclusion is that context matters. Knowing the context for SAA can help to avoid the risks inherent in its implementation in developing countries. In fact, context makes all the difference. SAA reformscantakemanyshapesandformsandthisisusuallyinresponsetothecountry’spoliticaleconomy, educationgoals,capacity,performanceissues,andhistory,justtonameafewcontextualfactorsthatmay influencethepoliciesanddesignoftheintendedSAAmeasures.  In countries where educational institutions are strong, especially in the area of training and in the selection of good teachers, autonomy and accountability in SBM may not be a necessary condition for success. In these conditions, which are found in many high performing countries in Europe, trust is the mainelementofaccountability.Parentstrustandsupportthesystembecausetheempiricalevidence— shown by the results in international testing exercises such as PISA—indicates that it is producing good results. Nevertheless, even in highͲperforming countries, trust and professionalism flourish in a context ofschoolautonomyandaccountability.  What should be noted is that when undertaking school autonomy and accountability reforms, there is a general assumption that actors and stakeholders at the school level will come together in an automatic and collegial way to put schoolͲbased authority and accountability into practice. However, due to weak data, there is only a small body of robust evidence to show that this happens. At present, we have few randomized experiments to draw upon; however, more and more projects are building in impact evaluation so that we can understand better what combination of SAA policies and activities are meaningfulforbetterlearningoutcomesinavarietyofcontextualsituationsinthefuture. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 36  References  Andrabi, Tahir, Jishnu Das and Asim Ijaz Khwaja. 2012. "What Did You Do All Day?: Maternal Education and Child Outcomes," Journal of Human Resources, University of Wisconsin Press, vol. 47(4), pages873Ͳ912.  Arcia, Gustavo, Raja Bentaouet Kattan, Harry Anthony Patrinos and Angie RiveraͲOlvera. 2013. “From GoodtoGreat.TheImpactofProgramadeEscuelasdeCalidadonLearningOutcomesinMexico.” Washington,DC:TheWorldBank.  Arcia,Gustavo,KevinMacdonald,HarryA.Patrinos,andEmilioPorta.2011.“SABERSchoolAutonomyand AccountabilityBackgroundPaper”WorkingPaper,No.94450.Washington,DC:TheWorldBank.  Arcia, Gustavo, Harry Patrinos, Emilio Porta, and Kevin Macdonald, 2011. “School Autonomy and Accountability in Context: Application of Benchmarking Indicators in Selected European Countries.”WorkingPaper,No.94444.Washington,DC:TheWorldBank.  Arcia,G.,E.PortaandJ.R.Laguna.2004.“OtrovistazoalaautonomíaescolardeNicaragua:Aceptacióny percepciónen2004.”(“Anotherlookatschoolautonomy:Acceptanceandperceptionsin2004”). Dirección General de Prospección y Políticas. Ministerio de Educación, Cultura y Deportes. Managua,Nicaragua.  Banerjee,Abhijit,RukminiBanerji,EstherDuflo,RachelGlennerster,andStutiKhemani.2008.“Pitfallsof Participatory Programs: Evidence from a Randomized Evaluation in Education in India” NBER WorkingPaper,Cambridge,MA:http://nber.org/papers/w14311  Bardhan, Pranab. 2002. “Decentralization of Governance and Development.” Journal of Economic Perspectives.Vol.16,No.4,185Ͳ205.  Barrera,Felipe,TazeenFasih,andHarryPatrinoswithLucreciaSantibáñez,2009.DecentralizedDecisionͲ MakinginSchools.TheTheoryandEvidenceonSchoolͲBasedManagement.Washington,DC:The WorldBank.  Benveniste,LuisandJ.Marshall.2004."SchoolGrantsandStudentPerformance:EvidencefromtheEQIP ProjectinCambodia."Unpublishedmanuscript.Washington,DC.:TheWorldBank.  Blimpo, Moussa P. and David K. Evans. 2011. “SchoolͲBased Management and Educational Outcomes: LessonsfromaRandomizedFieldExperiment.”Washington,DC:TheWorldBank  Bloom,Nicholas;RenataLemos,RaffaellaSadunandJohnVanReenen.2014.“DoesManagementMatter inSchools?”NBERWorkingPaperSeries.Cambridge,MA.  Bold,Tessa,MwangiKimenyi,GermanoMwabuandJustinSandefur.2013.“WhyDidAbolishingFeesNot Increase Public School Enrollment in Kenya?” Africa Growth Initiative Working Paper No. 4. WashingtonDC:TheBrookingsInstitution.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 37  Borman, Geoffrey D., G. Hewes, L. Overman, and S. Brown.2003. “Comprehensive School Reform and Achievement:AMetaͲAnalysis.”ReviewofEducationalResearch73(2):125Ͳ230  Bruns,Barbara,DeonFilmer,andHarryAnthonyPatrinos,2011.MakingSchoolsWork:NewEvidenceon AccountabilityReforms.Washington,DC:TheWorldBank.  Burde, Dana, Leigh L. Linden. 2012. “The Effect of VillageͲBased Schools: Evidence from a Randomized ControlledTrialinAfghanistan,”NBERWorkingPaper18039.  Caldwell,BrianJ.2005.SchoolͲBasedManagement.UNESCOIIEP.Paris.  Camargo,Braz,RafaelCamelo,SergioFirpo,andVladimirPonczek.2014.Information,MarketIncentives, and Student Performance. IZA Discussion Paper No. 7941, Institute for the Study of Labor: Germany.  Chaudhury, Nazmul and D. Parajuli. 2010. "Giving It Back: Evaluating the impact of Devolution of School ManagementtoCommunitiesinNepal."Unpublishedmanuscript.Washington,DC.:WorldBank.  Chen, Dandan. 2011. “SchoolͲBased Management, School DecisionͲMaking and Education Outcomes in Indonesian Primary Schools.” World Bank Policy Research Working Paper. Washington, DC.: The WorldBank.  Clark, D. 2009. “The Performance and Competitive Effects of School Autonomy.” Journal of Political Economy117(4),745Ͳ783.  Clarke,Marguerite,2010.“RoadmapforBuildinganEffectiveAssessmentSystem.”HumanDevelopment Network.Washington,DC.:TheWorldBank.  Clarke, Marguerite, 2012. What Matters Most for Student Assessment Systems: A Framework Paper. SABERWorkingPaperSeries,No.1.Washington,DC.:TheWorldBank.  Contreras, Nicolas. 2015. “School Autonomy Education Quality and Development: An Instrumental Variable Approach.”  CES Working Paper, Centre d'Economie de la Sorbonne Ͳ Paris School of Economics.  Corrales,Javier,2006.“DoesParentalParticipationinSchools EmpowerorStrainCivilSociety? TheCase of CommunityͲmanaged Schools in Central America.” Social Policy & Administration 40 (4): 450Ͳ 470.  Cristina Ling, Nidhi Khattri, Shreyasi Jha. 2010. “The Effects of SchoolͲBased Management in the Philippines: An Initial Assessment Using Administrative Data.” The World Bank Policy Research WorkingPaper.  DarlingͲHammond, Linda and Laura Wentworth. 2010. “Benchmarking Learning Systems: Student Performance Assessment in International Context.” Stanford, CA: Stanford University, Stanford CenterforOpportunityPolicyinEducation.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 38  deVijlder,Frans,2001.“ChoiceandFinancingofSchoolsinTheNetherlands.”MaxGooteExpertCenter, UniversityofAmsterdam.  Di Gropello, Emanuela, 2004. “Education Decentralization and Accountability Relationships in Latin America.”WorldBankPolicyResearchWorkingPaper3453.Washington,DC.  Di Gropello, Emanuela and Jeffrey H. Marshall. 2005. “Teacher Effort and Schooling Outcomes in Honduras.”InIncentivestoImproveTeaching,ed.E.Vegas.Washington,DC:WorldBank.  DiGropello,Emanuela,2006.“AComparativeAnalysisofSchoolͲBasedManagementinCentralAmerica.” WorldBankWorkingPaperNo.72.Washington,DC.  Di Gropello, Emanuela, and Jeffery H Marshall. 2009. “Decentralization and Educational Performance: Evidence from the PROHECO Community School Program in Rural Honduras” Education Economics,1Ͳ20.  Duflo, Esther, Pascaline Dupas, and Michael Kremer. 2007. “Peer Effects, PupilͲTeacher Ratios, and Teacher Incentives: Evidence from a Randomized Evaluation in Kenya.” MIT Poverty Action Lab WorkingPaper.  Elacqua,GregoryandFatimaAlves.2014.“RisingExpectationsinBrazilandChile.”EducationNext14(1): 54.  Elacqua,GregoryandErnestoTrevino.2008.“TheEffectsofAccountabilityPressuresonTeacherPolicies andPracticesinLowͲperformingschools:TheCaseofChile.”PREAL.Santiago,Chile.  Elmore, Richard, Charles Ablemann, and Susan Fuhrman. 1996. “The New Accountability in State Education Reform: From Process to Performance.” In Ladd, Helen (Ed.). Holding Schools Accountable. PerformanceͲBased Reform in Education. Washington, DC: The Brookings Institution.  Eskeland,GunnarS.andDeonFilmer.2002.“Autonomy,Participation,andLearninginArgentineSchools: Findings,andTheirImplicationsforDecentralization.”Workingpaper,Washington,DC:TheWorld Bank.  Eurydice.2007.SchoolAutonomyinEurope.PoliciesandMeasures.Brussels:Eurydice.  Fox, Jonathan. 2014. “Social Accountability: What does the evidence really say?” Presentation delivered March3,2014attheWorldBank.AmericanUniversity,SchoolofInternationalService.  Fullan, Michael and Nancy Watson 1998. “SchoolͲbased Management: Reconceptualizing to Improve Learning Outcomes.” Working Paper. Ontario Institute for Studies in Education, University of Toronto.  Gaventa,JohnandRosemaryMcGee.2013.“TheImpactofTransparencyandAccountabilityInitiatives.” DevelopmentPolicyReview,Vol.31,pp.s3Ͳs28,2013.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 39  Gertler, Paul, Harry Anthony Patrinos, and M. RubioͲCodina. 2006. "Empowering Parents to Improve Education. Evidence from Rural Mexico." Policy Research Working Paper 3935. Washington, DC: WorldBank.  Gillies, John, Luis Crouch, and Anna Florez. 2010. Strategic Review of the EDUCO Program, Washington, DC:USAID.  Gunnarsson,Victoria,PeterF.Orazem,MarioA.Sánchez,andAimeeVerdisco.2009.“DoesLocalSchool Control Raise Student Outcomes? Evidence on the Roles of School Autonomy and Parental Participation.” Economic Development and Cultural Change, Vol. 58, No. 1, 25Ͳ52. University of ChicagoPress  Hahn, Youjin; Liang Choon Wang, and HeeSeung Yang. 2014. Do Greater School Autonomy and Accountability Make a Difference? Evidence from the Random Assignment of Students into PrivateandPublicHighSchoolsinSeoul.AvailableatSSRN:http://ssrn.com/abstract=2352765or http://dx.doi.org/10.2139/ssrn.2352765  Hallinger, Philip and Ronald H. Heck. 1996. “Reassessing the Principal’s Role in School Effectiveness: A ReviewofEmpiricalResearch,1980Ͳ1995.”EducationalAdministrationQuarterly32(1):5Ͳ44.  Hanushek, Eric A., Susanne Link and Ludger Woessmann. 2013. “Does School Autonomy Make Sense Everywhere?PanelEstimatesfromPISA.”JournalofDevelopmentEconomics104(C):212Ͳ232.  Hanushek,ErikA.andMargaretE.Raymond.2005.“DoesSchoolAccountabilityLeadtoImprovedStudent Performance?”JournalofPolicyAnalysisandManagement,Vol.24,No.2,297Ͳ327.  Hanushek, Eric A., and Margaret Raymond. 2003. “Lessons about the Design of State Accountability Systems.”InNoChildLeftBehind?InThePoliticsandPracticeofAccountability,ed.PaulPeterson andMartinWest,127Ͳ51.Washington,DC:BrookingsInstitutionPress.  Hanushek, Eric A. and Ludger Woessmann. 2007. “The Role of Education Quality for Economic Growth.” WorldBankPolicyResearchWorkingPaperNumber4122.Washington,DC:TheWorldBank.  Hargreaves, A. and Michael Fullan. 1998. What’s Worth Fighting For Out There? New York, Teachers CollegePress.  Hatch,Thomas.2013.“Beneaththesurfaceofaccountability:Answerability,ResponsibilityandCapacityͲ Building in Recent Education Reforms in Norway.” Journal of Education Change (2013) 14:113– 138(Publishedonline:17January2013).  Heim, Michael. 1996. “Accountability in Education: A Primer for School Leaders.” Pacific Resources for EducationandLearning,HawaiiDepartmentofEducation.Honolulu,Hawaii.  Hood, C. 2001. “New Public Management,” In N.J. Smelser, P.B. Baltes (eds), International Encyclopedia oftheSocialandBehavioralsciences.Amsterdam:Elsevier.  Hoxby,Caroline.2001.“TestingisaboutOpennessandOpennessWorks.”HooverDailyReport,July30.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 40  Jesse, Cornelia, Gerard Lassibille, JeeͲPeng Tan and Trang Van Nguyen. 2010. “Managing for Results in Primary Education in Madagascar: Evaluating the Impact of Selected Workflow Interventions.” TheWorldBankEconomicReview24:303Ͳ329.  Jimenez,EmmanuelandYasuyukiSawada.1999.“DoCommunityͲManagedSchoolsWork?AnEvaluation ofElSalvador’sEDUCOProgram.”TheWorldBankEconomicReview13:415Ͳ441.  Jimenez, Emmanuel and Yasuyuki Sawada. 2003. “Does Community Management Help Keep Kids in Schools?EvidenceUsingPanelDatafromElSalvador’sEDUCOProgram.”DiscussionPaper.Center forInternationalResearchontheJapaneseEconomy,UniversityofTokyo.  Jimenez, Emmanuel and Yasuyuki Sawada. 2014. “Does Community Management Help Keep Children in Schools?EvidenceUsingPanelDatafromElSalvador’sEDUCOProgram.”EconomicDevelopment andCulturalChange,Vol.62,No.2,pp.307Ͳ338.UniversityofChicagoPress.  Kaplan, R. S. and Norton, D. P. (2008). "Mastering the Management System." Harvard Business Review.January2008.  King, Elizabeth and Susana Cordeiro Guerra. 2005. “Education Reforms in East Asia: Policy, Process and Impact. In East Asia Decentralizes: Making Local Government Work. (Ed) Roland White and Paul Smoke.Washington,DC:WorldBank.  King,E.M.,andB.Ozler.1998."What’sDecentralizationGottodowithLearning?TheCaseofNicaragua's SchoolAutonomyReform."Unpublishedmanuscript.DevelopmentResearchGroup.Washington, DC:TheWorldBank.  Lierl, Malte. 2014. Preferences or Incentives? Experimental Evidence on the Accountability of Village LeadersinTanzania.Unpublished.YaleUniversity,DepartmentofPoliticalScience.Presentedon August26,2014attheWorldBank.  Litvak, Jennie and Jessica Seddon. 1999. Decentralization Briefing Notes. World Bank Institute Working PapersincollaborationwithPREMnetwork.  LopezͲCalva, L.F., and L.D. Espinosa. 2006. "Efectos diferenciales de los programas compensatorios del CONAFE en el aprovechamiento escolar." In Efectos del Impulso a la Participacion de los Padres deFamiliaenlaEscuela,ed.,CONAFE.Mexico:CONAFE.  Lugaz,CandyandAntonDeGrauwe.2010.SchoolingandDecentralization.Paris.UNESCOIIEP.  Mansuri, Ghazala and Vijayendra Rao. 2013. Localizing Development–Does Participation Work? Washington,DC:TheWorldBank.  McEwen, Patrick J., Miguel Urquiola and Emiliana Vegas. 2008. School Choice, Stratification, and InformationonSchoolPerformance:LessonsfromChile.Economia:JournaloftheLatinAmerican andCaribbeanEconomicAssociation8:1.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 41  McLean, Keith and Elizabeth King. 1999 “Decentralization of the Education Sector,” in Decentralization BriefingNotes.EditedbyJennieLitvackandJessicaSeddon.WorldBankInstituteWorkingPapers inCollaborationwithPREMnetwork.Washington,DC:TheWorldBank.  Meza, Darlyn, José Guzmán, and Lorena De Varela. 2004. “EDUCO: A CommunityͲManaged Education PrograminRuralElSalvador(1991Ͳ2003).”Enbreve,No.51.Washington,DC:TheWorldBank.  Miljevic, Gordana, ed., 2009. School Governance and Social Inclusion. Involvement of Parents. Open SocietyInstituteandUniversityofLjubljiana,Slovenia.  Moynihan, Donald. P. 2007. “Citizen Participation in Budgeting: Prospects for Developing Countries” In Participatory Budgeting. Public Sector Governance and Accountability Series, No. 39498. Edited byAnwarShah.Washington,DC:TheWorldBank.  Mulindwa,InnocentNajjumba,JamesHabyarimana,andCharlesLwangaBunjo.2013.ImprovingLearning in Uganda, Vol.III SchoolͲBased Management: Policy and Functionality. Washington, DC: The WorldBank  Murnane,R.J.,J.B.Willett,andS.Cardenas.2006."DidtheParticipationofSchoolsinProgramEscuelasde Calidad(PEC)InfluenceStudentOutcomes?"WorkingPaper,HarvardUniversityGraduateSchool ofEducation.Cambridge,MA:HarvardUniversity.  Nielsen,H.Dean.2007.“EmpoweringCommunitiesforImprovedEducationalOutcomes:SomeEvaluation FindingsfromtheWorldBank.”Prospects,Vol.XXXVII,No.1.  Newmann, F. and G. Wehlage. 1995. “Successful School Restructuring.” Madison, WI, Center on OrganizationandRestructuringofSchools.  Odden, Eleanor R. and Priscilla Wohlstetter. 1995. “Making SchoolͲBased Management Work.” EducationalLeadership15(5),32Ͳ26.  Ofsted.(2014).“Theframeworkforschoolinspection.”Availableonline:http://www.ofsted.gov.uk  OECD (Organisation for Economic CoͲoperation and Development). 2011. “School Autonomy and Accountability:AreTheyRelatedtoStudentPerformance?”PISAinFocus,No.9October.  Parker E. Caroline. 2005. “Teacher Incentives and Student Achievement in Nicaraguan Autonomous Schools,”p.359Ͳ388.InIncentivestoImproveTeaching:LessonsfromLatinAmerica,ed.Emiliana Vegas.Washington,D.C.:TheWorldBank.  Patrinos, Harry Anthony. 2011. “SchoolͲBased Management.” In Bruns, B., D. Filmer, and H.A. Patrinos (2011), Making Schools Work: New Evidence on Accountability Reforms. Washington, DC: World Bank.  Patrinos, Harry Anthony, and Robin Horn, 2010. “Benchmarking Education Systems for Results.” PowerPointPresentation,HumanDevelopmentNetwork.Washington,DC:TheWorldBank.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 42  Piper, Benjamin and Medina Korda. 2011. EGRA Plus: LiberiaͲ Understanding the Causal Mechanisms: EGRAPlus’sEffectiveness.ResearchTrianglePark,NC:RTIInternational.  Piper,BenjaminandMedinaKorda.2010.EGRAPlus:LiberiaͲProgramReport.ResearchTrianglePark,NC: RTIInternational.  Pradhan, M., D. Suryadarma, A. Beatty, and R. P. Artha. 2011.” Improving Educational Quality through EnhancingCommunityParticipation:ResultsfromaRandomizedFieldExperimentinIndonesia.” WorldBankPolicyResearchWorkingPaper.Washington,DC:TheWorldBank.  Rechebei, Elizabeth. 2010. Accountability and Reality. Who Should Do What? And Who Should Be Accountable?ResearchintoPracticeSeries,PacificResourcesforEducationandLearning,Hawaii DepartmentofEducation.Honolulu,Hawaii.  Reinikka, Ritva and Jakob Svensson 2004. The Power of Information: Evidence from a Newspaper Campaign to Reduce Capture of Public Funds. The Quarterly Journal of Economics, 119 (2) 670Ͳ 705.  Ritzen,JosefM.M.,JanvanDomelen,andFransJ.deVijlder.1997.“SchoolFinanceandSchoolChoicein TheNetherlands.”EconomicsofEducationReview16(3):329Ͳ335.  Rogers,F.HalseyandAngelaDemas.2013.TheWhat,Why,andHowoftheSystemsApproachtoBetter EducationResults(SABER).Report#79901.Washington,DC:WorldBank  Shibuya,Kazuro.2014.CaseStudyofSchoolAutonomyandAccountability in BurkinaFaso.Washington, DC:TheWorldBank.  Skoufias, Emmanuel, and Joseph Shapiro. 2006. “Evaluating the Impact of Mexico’s Quality Schools Program:ThePitfallsofUsingNonͲexperimentalData.”PolicyResearchWorkingPaper4036.The WorldBankImpactEvaluationSeries8.Washington,DC:TheWorldBank  Takeda, Nazumi, Angela Demas and Kazuro Shibuya. 2014. World Bank Portfolio Review of School Autonomy and Accountability: Operations and Analytical Work, FY2003Ͳ2013. Report #88956, Washington,DC:TheWorldBank  Vegas,Emiliana.2001.“SchoolChoice,StudentPerformance,andTeacherandSchoolCharacteristics:The ChileanCase.”DevelopmentResearchGroup.Washington,DC:TheWorldBank.  Vegas, Emiliana, and Ilana Umansky. 2005. “Improving Teaching and Learning through Effective Incentives:LessonsfromEducationReformsinLatinAmerica?”InIncentivestoImproveTeaching: LessonsfromLatinAmericaeditedbyEmilianaVegas.Washington,DC:TheWorldBank.  Wampler, Brian. 2007 “A Guide to Participatory Budgeting.” In Participatory Budgeting. Public Sector Governance and Accountability Series, No. 39498. Edited by Anwar Shah. Washington, DC: The WorldBank.  Winkler,Donald.2005.IncreasingAccountabilityinEducationinParanáState,Brazil.EQUIP2PolicyBrief, USAID,Washington,DC WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 43   Winkler, Donald and Jon Herstein. 2005. “Information Use and Decentralized Education.” EQUIP2 Policy Brief,U.S.AgencyforInternationalDevelopment,Washington,DC  Wohlstetter, Proscilla. 1995. Getting SchoolͲBased Management Right, What Works and What Doesn’t. PhiDeltaKappan.  World Bank. 2003. World Development Report 2004. Making Services Work for Poor People. New York: OxfordUniversityPress.  _______. 2007a. “What Do We Know About SchoolͲBased Management?” Working Paper, No. 44921 Washington,DC:TheWorldBank.  _______.2007b.“WhatisSchoolͲBasedManagement?”WorkingPaper,No.44922.Washington,DC:The WorldBank.  _______.2010.AReviewoftheBulgariaSchoolAutonomyReforms.WashingtonDC:TheWorldBank.  _______. 2011a. “Enhancing SchoolͲBased Management in Indonesia.” HD Policy Brief 60457, Washington,DC:TheWorldBank.  _______. 2011b. Learning for All, Investing in People’s Knowledge and Skills to Promote Development: WorldBankGroupEducationStrategy2020.Washington,DC:TheWorldBank.  _______. 2011c. What Matters for School Finance: A Framework Paper. SABER Working Paper Series 2. Washington,DC:TheWorldBank.  _______. 2013a. Republic of the Philippines Basic Education Public Expenditure Review Phase II School Based Management in the Philippines: An Empirical Investigation. Washington, DC: The World Bank.  _______. 2013b. SABER School Autonomy and Accountability Mexico Country Report. Available at: http://saber.worldbank.org/index.cfm?indx=14.  _______. 2013c. SchoolͲBased Management in Lao PDR: Current Conditions and Recommendations for theFuture.Washington,DC:TheWorldBank.  Yuki,Takako;KengoIgei,andAngelaDemas.2015.MeasuringQualityofPoliciesandtheirImplementation ForBetterLearning:AdaptingWorldBank’sSABERToolsOnSchoolAutonomyandAccountability toSenegal.  Zafeirakou,Aglaia.2004.“SchoolManagementandtheChangingRolesofSchoolDirectors,CaseStudies, Bulgaria and Romania.” Analytical Sector Work, the World Bank, Europe Central Asia, Human DevelopmentSector,(ECSHD).   WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 44  Appendix1:SchoolAutonomyandAccountabilityPolicyGoalsand PolicyActions  PolicyGoal PolicyAction 1A Legalauthorityoverthemanagementoftheoperationalbudget Levelofautonomyin LegalauthorityoverthemanagementofthenonͲteachingstaff 1B theplanningand salaries managementof 1C Legalauthorityoverthemanagementofteachersalaries schoolbudget 1D Legalauthoritytoraiseadditionalfundsfortheschool 1E Collaborativebudgetplanning 2A Autonomyinteacherappointmentanddeploymentdecisions Levelofautonomyin AutonomyinnonͲteachingstaffappointmentanddeployment personnel 2B decisions management 2C Autonomyinschoolprincipalappointmentanddeploymentdecisions 3A Participationoftheschoolcouncilinbudgetpreparation 3B Participationoftheschoolcouncilinfinancialoversight Roleoftheschool 3C Participationoftheschoolcouncilinpersonneloversight councilinschool 3D Participationoftheschoolcouncilinschoolactivities governance 3E Participationoftheschoolcouncilinlearninginputs 3F Transparencyincommunityparticipation 4A Existenceandfrequencyofschoolassessments 4B Useofschoolassessmentsformakingschooladjustments Schoolandstudent 4C Existenceandfrequencyofstandardizedstudentassessments assessment Useofstandardizedstudentassessmentsforpedagogical,operational, 4D andpersonneladjustments 4E Publicationofstudentassessments 5A Guidelinesfortheuseofresultsofstudentassessments 5B Analysisofschoolandstudentperformance Degreeoffinancialaccountabilityatthecentrallevel,regional, SchoolAccountability 5C municipal,localandschoollevel 5D Degreeofaccountabilityinschooloperations 5E Degreeoflearningaccountability   WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 45  Appendix2:RubricforSABERSchoolAutonomyandAccountability  PolicyGoal1:Thelevelofautonomyintheplanningandmanagementoftheschoolbudget.  PolicyAction Latent Emerging Established Advanced 1A Legal Legal Legal NonͲsalaryexpenditure NonͲsalary authorityover management management canbemanagedby expenditurecanbe management authorityover authorityover schoollevelwithout managedbyschool ofthe theoperational theoperational consultationwith levelinconsultation operational budgetis budgetisatthe parents/community with budget centralized. regionalor membersunder parents/community municipallevels. governmentguidelines. members. 1B Legal Legal Legal NonͲteachingstaff NonͲteachingsalaries authorityover management management salariescanbe canbemanagedby the authorityover authorityover managedattheschool schoollevelin management nonͲteaching nonͲteaching levelwithout consultationof ofnonͲ staffsalariesis staffsalariesis consultationwith parents/community teachingstaff centralized. attheregional parent/community members.An salaries ormunicipal members.Acentralized establishedpayscale levels;a orregional/municipal maybeusedasa centralizedpay payscalemaybeused guide. scalemaybe asaguide. usedasaguide. 1C Legal Legal Legal Teachersalariescanbe Teachersalariescan authorityover management management managedbyschool bemanagedbyschool the authorityover authorityover levelwithout levelinconsultationof management teachersalaries teachersalaries consultationwith parents/community ofteacher iscentralized isattheregional parent/community members.An salaries ormunicipal members.Acentralized establishedpayscale levels;a orregional/municipal maybeusedasa centralizedpay payscalemaybeused guide. scalemaybe asaguide. usedasaguide. 1D Legal Budgetisfixed Schools can Schoolscanraise Schoolscanraise authorityto bytheMinistry requestmore additionalfundsfrom additionalfundsfrom raise ofEducation fundsfromsubͲ parents/community anysource additional andno national members,private fundsforthe additional governments. businesses,andfrom school fundingis nonͲgovernmental permitted institutions. 1E Collaborative Budgetary Provisionsallow Nationaland/orsubͲ Nationaland/orsubͲ budget decisionsare fortheschool nationalauthoritiesare nationalauthorities planning madeatthe leveltopropose tousetheproposed aretousethe nationaland aschoolbudget budgetbytheschool proposedbudgetby subͲnational tothesubͲ levelasareferencefor theschoollevelasthe levelsandthere nationallevelas thetransferof mainguideforthe isnosystemto arequestfor resourcestothe finaltransferof acceptabudget funding. school. resourcestothe proposalfrom school. theschoollevel.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 46   PolicyGoal2:Thelevelofautonomyinpersonnelmanagement  PolicyAction Latent Emerging Established Advanced 2A Autonomyin Teachersmust Regionalor Regionalor Schools(school teacher beappointed municipal municipal principals,school appointment anddeployedby governmentshave governmentshave council,parent and thecentral legalauthorityto legalauthorityto associationetc.)have deployment government appointteachers appointanddeploy legalauthorityto decisions levelundera underunionorcivil teachersunder appointteachers. unionorcivil service unionorcivilservice Unionandcivilservice service agreements. agreementswithout agreementmayor agreement. Appointmentsare reviewbycentral maynotregulatethe subjecttofinal authorities. appointments. reviewbycentral authorities. 2B Autonomyin NonͲteaching Regionalor Regionalor Schoolshavelegal nonͲteaching staffmustbe municipal municipal authoritytoappoint staff appointedand governmentshave governmentshave nonͲteachingstaff. appointment deployedby legalauthorityto legalauthorityto Civilservice and central appointnonͲ appointanddeploy agreementmayor deployment government teachingstaff nonͲteachingstaff. maynotregulatethe decisions levelundercivil undercivilservice Civilservice appointments. service agreements. agreementmayor agreement. maynotregulate theappointments. 2C Autonomyin Principalsareto Principalsaretobe Principalsaretobe Principalsaretobe school beappointed appointedand appointedand appointedand principal anddeployedby deployedbythe deployedby deployedby appointment thecentrallevel. centrallevel.Their regionalor municipal/local and Their performanceis municipal/local authoritiesin deployment performanceis evaluated authorities,whoare consultationwiththe decisions evaluated regionallyorby alsoresponsiblefor schoolcouncil/ centrallyand municipal theirevaluationand stakeholdersatschool theycanbe inspectors,which havetheauthority level,orbytheschool transferredor determinestheir fordetermining councilalone. firedbycentral tenure,transfer,or tenure,transfer,or Municipal/local authorities. removalbycentral removal. authoritiesare authorities. responsibleforthe principal'sevaluation todeterminetenure, transfer,orremoval.     WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 47  PolicyGoal3:Roleoftheschoolcouncilonschoolgovernance.  PolicyAction Latent Emerging Established Advanced 3A Participation Noroleforthe Schoolcouncilisto Schoolcouncilisto Schoolcouncilisto oftheschool schoolcouncil; haveavoiceinthe haveavoiceinthe haveavoiceinthe councilin budgetsare planningand planningand planningand budget prepared preparationofthe preparationofall preparationofall preparation centrallybythe nonͲsalarybudget expensesatthe expensesatthe Ministryof itemsattheschool schoollevel,but schoolleveland, Education. level,butfinal finalresponsibility dependingonthe responsibilityfalls fallsontheschool law,mayshare ontheschool principalorother responsibilitywith principalorother government theschoolprincipal. government authority. authority. 3B Participation Nolegalstanding Legalstandingasan Legalstandingasan Legalstandingasan infinancial asan organization,butno organization,and organization,legal oversight organization,no legalauthorityto legalauthorityto authoritytohavea legalauthorityto haveavoice,andno haveavoice,butno voice,andlegal haveavoice,and legaloversight legaloversight oversightauthority nolegaloversight authorityonbudget authorityonbudget onbudgetissues. authorityon issues. issues. budgetissues. 3C Participation Nolegalrightor Nolegalrightin Legalrighttohavea Legalrightto inpersonnel voiceinteacher teacher voiceinteacher oversee management appointments, appointmentsand appointments, appointments, transfers,and removals,buthave removals,and removals,or removals. avoiceinteacher transfers. transferofteachers. transfers. 3D Community Noformal Thereareformal Thereareformal Thereareformal participation instructions, instructions, instructions, instructions, inschool manuals,or manuals,and manuals,and manuals,and activities mandatesfor mandatesfor mandatesfor mandatesfor organizing organizing organizing organizing volunteersto volunteersto volunteerstoplan volunteerstoplan, performactivities. implement andimplement implement,and activities. activities. evaluateactivities. 3E Community Nolegalauthority Legalauthorityto Legalauthorityto Legalauthorityto participation tovoicean voiceanopinion, voiceanopinion voiceanopinion onlearning opinionandno butnolegal andlegaloversight andlegaloversight inputs legaloversighton oversighton onsomelearning onalllearning learninginputsto learninginputsto inputstothe inputstothe theclassroom. theclassroom. classroom. classroom. 3F Transparency Noprovisionsfor Noprovisionsfor Thereareprovisions Thereareprovisos incommunity theopenelection theopenelectionof foropenelectionof forregularly participation ofschoolcouncil schoolcouncil schoolcouncil scheduledelections membersandfor members,but membersbutno ofschoolcouncil general guidelinesfor termlimitsor membersand assemblies. callinggeneral regularschedulefor definedtermlimits. assemblies. elections.Thereare Thereareguidelines guidelinesforcalling forcallinggeneral generalassemblies. assemblies.  WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 48  PolicyGoal4:Schoolandstudentassessment.  PolicyAction Latent Emerging Established Advanced 4A Existenceand Schoolsdo Schools aretobe Schoolsaretobe Schoolsaretobe frequencyof notassess assessedeveryfew assessedeveryyear assessedeveryyear school school yearsusingMinistry usingMinistryof usingMinistryof assessments performance ofEducationcriteria. Educationcriteria Educationcriteria.In onaregular addition,there basis. shouldbesporadic evaluationsofspecific aspectsofschoollife, suchasstudent poverty,equity,and teacherquality.The resultsofall evaluationsshouldbe madepublicand easilyaccessible. 4B Useofschool Schoolsdo CentralMinistryof Centralorregional/ MinistryofEducation assessments notuseschool Educationmust municipalbranchof ormunicipal formaking assessments analyzeschool theMinistryof governmentsmust school tomake assessmentresults Educationmust analyzeschool adjustments pedagogical andsendthemtothe analyzeschool assessments,and adjustments, regions/municipalities assessmentresults makeresultseasily ortochange andmakebroad andsendthem accessibletoschools school recommendationson directlytothe andthepublic. materials. pedagogicaland schools.Schoolsmay Schoolsmustusethe operational usetheinformation informationtomake adjustments. tomakepedagogical pedagogical, andoperational personnel,and adjustments. operational adjustments. 4C Existenceand Studentsdo Assessmentsof Assessmentsof Assessmentsof Frequencyof nottake studentlearningare studentlearningare studentlearningare standardized standardized doneeveryfewyears doneeveryfewyears doneeveryyearin student tests. inselectedgrades inselectedgradesfor selectedgradesforall assessments usingrepresentative allstudentsinthe studentsinthe samplesofstudents. country. country. 4D Useof Schoolsdo CentralMOE must Centralorregional/ MOEormunicipal standardized notuse analyzeresultsof municipalbranchof governmentsmust student standardized standardizedstudent MoEmustanalyze analyzestandardized assessments student assessmentsandsend standardizedstudent studenttestscores; for assessments themtotheregions/ testscores;send makeresultseasily pedagogical, tomake municipalitiesand resultsand accessibletoschools operational, pedagogical makebroad recommendationsto andthepublic. and adjustments recommendationson regional,localoffices Schoolsmustusethe personnel ortochange pedagogical, anddirectlytothe informationtomake adjustments school operationaland/or schools.Schoolsmay pedagogical, materials. personnel usetheinformation operational,or adjustments. tomakepedagogical personnel andoperational adjustments. adjustments. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 49  4E Publicationof Resultsofthe Resultsofthestudent Resultsofthe Resultsofthestudent student student assessmentsare studentassessments assessmentsare assessments assessments madeavailableto aremadeavailableto madepublicand arenot centraland central, availableonline. reported. regional/municipal regional/municipal levelsoftheMOES. levelsoftheMOES, andtoschools.  PolicyGoal5:SchoolAccountability  PolicyAction Latent Emerging Established Advanced 5A Guidelinesfor Thereareno Thereare Thereare Thereareguidelines theuseof guidelinesforthe guidelinesfortheguidelinesforthe fortheuseofresults resultsof useofresultsof useofresultsof useofresultsof ofstudent student student student student assessmentsatall assessments assessments. assessmentsattheassessmentsatthe levels.Theguidelines nationaland national, areavailableonline municipallevels municipal,and andcanbeusedto only. schoollevels. foster/demand Schoolcouncils accountability. canusethe guidelinestovoice accountability. 5B Analysisof Thereareno Thereare Thereare Thereareprovisions schooland provisionsforthe provisionsfor provisionsfor forcomparative student comparative comparative comparative analysisofstudent performance analysisof analysisofstudent analysisofstudent assessmentresultsfor student assessmentresults assessmentresults differenttypesof assessment fordifferenttypes fordifferenttypes schools,across resultsfor ofschools,across ofschools,across regions,andfor differenttypesof regions,andfor regions,andfor previousyearsatthe schools,across previousyearsat previousyearsat national,regional, regions,andfor thenationaland thenational, municipal,andschool previousyears. regionallevels. regional,and levels.Detailedschool municipallevels. performanceresults Schoolsare attheschoollevel requiredto mustbepublished distribute online. summaryresults toparents. 5Ci Degreeof Thereareno Thereare Thereare Thereareregulations financial regulationsin regulationsin regulationsin inplaceforcomplying accountability placefor(i) placefor placefor withtherulesof atthecentral complyingwith complyingwith complyingwith financialmanagement level therulesof therulesof therulesof andtransparency;for financial financial financial reportingtothose managementand managementand managementand withoversight transparency;(ii) transparency,but transparency,and authority;andfor reportingto notforreporting forreportingto linkingrewardsand thosewith tothosewith thosewith sanctionsto oversight oversight oversight compliance. authority;and authority;andnot authority;butnot (iii)linking forlinkingrewards forlinkingrewards WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 50  rewardsand andsanctionsto andsanctionsto sanctionsto compliance. compliance. compliance. 5Cii Degreeof Thereareno Thereare Thereare Thereareregulations financial regulationsin regulationsin regulationsin inplaceforcomplying accountability placefor(i) placefor placefor withtherulesof atthe complyingwith complyingwith complyingwith financialmanagement regional/ therulesof therulesof therulesof andtransparency;for municipal financial financial financial reportingtothose level managementand management,but management,and withoversight transparency;(ii) notforreporting forreportingto authority;andfor reportingto tothosewith thosewith linkingrewardsand thosewith oversight oversight sanctionsto oversight authority;andnot authority;butnot compliance. authority;and forlinkingrewards forlinkingrewards (iii)linking andsanctionsto andsanctionsto rewardsand compliance. compliance. sanctionsto compliance. 5Ciii Degreeof Thereareno Thereare Thereare Thereareregulations financial regulationsin regulationsin regulationsin inplaceforcomplying accountability placefor(i) placefor placefor withtherulesof attheschool complyingwith complyingwith complyingwith financialmanagement level therulesof therulesof therulesof andtransparency;for financial financial financial reportingtothose managementand management,but managementand withoversight transparency;(ii) notforreporting transparency,and authority;andfor reportingto tothosewith forreportingto linkingrewardsand thosewith oversight thosewith sanctionsto oversight authority;andnot oversight compliance. authority;and forlinkingrewards authority;butnot (iii)linking andsanctionsto forlinkingrewards rewardsand compliance. andsanctionsto sanctionsto compliance. compliance. 5D Degreeof Thereareno Thereare Thereare Thereareregulations accountability regulationsin regulationsin regulationsin inplaceforcomplying inschool placefor:(i) placefor placefor withtherulesof operations complyingwith complyingwith complyingwith schooloperationsand therulesof therulesofschool therulesofschool forreportingtothose school operations,but operationsandfor withoversight operations;(ii) notforreporting reportingtothose authority,andfor reportingto tothosewith withoversight linkingrewardsand thosewith oversight authority,butnot sanctionsto oversight authority,andnot forlinkingrewards WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 51  authority;and forlinkingrewards andsanctionsto operating (iii)linking andsanctionsto operating performance. rewardsand operating performance. sanctionsto performance. operating performance. 5E Degreeof Nomandatefor Thereisa Theresultsof Theresultsofstudent learning simplifyingand mandatefor student assessmentsare accountability explainingresults simplifyingand assessmentsare simplifiedand ofstudent explainingresults simplifiedand explainedtothe assessmentsto ofstudent explainedtothe publicandthelocal thepublic. assessmenttothe publicandthe level/schoolsare public. locallevel/schools obligatedtohavea areobligatedto meetingwiththe solicitfeedback schoolcommunityto fromtheschool solicitfeedbackand communityon toinformthemofa thoseresults. planofactionto addresstheissues.         WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 52   worldbank.org/education/saber TheSystemsApproachforBetterEducationResults(SABER)initiativecollectsdata on the policies and institutions of education systems around the world and benchmarks them against practices associated with student learning. SABER aims to give all parties with a stake in educational results—from students, administrators, teachers, and parents to policymakers, business people and political leaders—an accessible, detailed, objective snapshot of how well the policiesoftheircountry’seducationsystemareorientedtowarddeliveringlearning forallchildrenandyouth. ThisframeworkpaperfocusesspecificallyintheareaofSchoolAutonomyand AccountabilitySector. Thefindings,interpretations,andconclusionsexpressedhereinarethoseoftheauthor(s)anddonotnecessarilyreflectthe viewsoftheInternationalBankforReconstructionandDevelopment/TheWorldBankGroupanditsaffiliatedorganizations, orthoseoftheExecutiveDirectoroftheWorldBankGrouporthegovernmentstheyrepresent. The World Bank Group does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations,andotherinformationshownonanymapinthisworkdonotimplyandjudgmentonthepartoftheWorld BankGroupconcerningthelegalstatusofanyterritoryortheendorsementoracceptanceofsuchboundaries. WhatMattersMostForSchoolAutonomyandAccountability:AFrameworkPaper 53