MyanmarTowardsARoadmaptoReform en Red

FiscalDecentralizationinMyanmar:TowardsaRoadmapforReform
HamishNixonandCindyJoelene
June2014
SubnationalGovernanceinMyanmarDiscussionSeries
Hamish Nixon is an independent governance specialist focusing on the areas of

decentralization, peace and conflict, and public service delivery. Cindy Joelene is a
researcheratMDRIsCentreforEconomicandSocialDevelopment.Theauthorswould
like to thank all the individuals who gave time to help them better understand
Myanmars decentralization context. The paper benefited significantly from
commentsbyRobertEbelonearlierwork,andfromsuggestionsfromJamieBoexon
internationalstrategiesforfiscaldecentralizationreforms.AndreaSmurra,Susanne
Kempel, Henri Schreurs, Paul Minoletti, and Matthew Arnold all provided helpful
commentsonearlierdrafts.ChristopherFreemancarefullyeditedthemanuscript,The
MyanmarTranslationCompanyablytranslatedthereport.Anyerrorsofsubstanceor
argumentrestwiththeauthors.
TheSubnationalGovernanceinMyanmarDiscussionPaperSeriesisacollaborative
research initiative between the Myanmar Development Resource Institutes Centre
forEconomicandSocialDevelopmentandTheAsiaFoundation.Thepaperseriesaims
to provide Myanmar policymakers at national and local levels, civil society
organizations,thebusinesscommunity,developmentpartners,andotherinterested
stakeholders with timely research on subnational governance issues that directly
informpolicyandreformprocesses.Theresearchbehindtheseriesincorporatesthe
perspectivesofarangeofgovernment,political,nongovernmental,civilsociety,and
communitystakeholdersinsubnationalgovernance,whilealsobringingtobearthe
most relevant policy analysis and international experience. The sponsoring
organizations welcome inputs and suggestions on published, ongoing, or future
research.
The Myanmar Development Resource Institute's Centre for Economic and Social
Development is an independent thinktank dedicated to the economic and social
transformation of Myanmar.The Centre undertakes participatory policy research
studiesrelatedtoeconomicreform,povertyreduction,sustainabledevelopment,and
goodgovernanceinMyanmar.Italsoprovidestrainingandeducationservicesforkey
institutionsandorganizationscontributingtotheongoingprocessofreform.
The Asia Foundation is a nonprofit international development organization
committedtoimprovinglivesacrossadynamicanddevelopingAsia.Informedbysix
decades of experience and deep local expertise, the Foundation works through a
networkof18countryofficesinAsiaonprogramsthataddresscriticalissuesaffecting
theregioninthe21stcenturygovernanceandlaw,economicdevelopment,women's
empowerment,environment,andregionalcooperation.
Preface
The Myanmar Development Resource Institute's Centre for Economic and Social
Development(MDRICESD)andTheAsiaFoundationarepleasedtopresentthisfifth
volumeintheSubnationalGovernanceinMyanmarDiscussionPaperSeries.
ThepaceofchangeinMyanmarisrapidandconstantlyevolving,anddecentralization
and local governance are issues of critical importance to the countrys longterm
development and prioritiesinthegovernmentsreformagenda.Assuch,thereisa
real need for timely research and analysis on key reform areas related to
decentralization and local governance. As an extended collaboration between the
Foundation and MDRICESD, this series of discussion papers aims to provide
Myanmarspolicymakersatnationalandlocallevels,civilsocietyorganizations,the
businesscommunity,developmentpartners,andotherinterestedstakeholderswith
research findings on subnational governance issues that directly inform policy and
reformprocesses.
FiscaldecentralizationformsthebackboneofMyanmarseffortstostrengthenpublic
services,encouragedevelopmentacrossthecountry,andsecurepeaceandstability.
Thecountryhasmadeimportantfirststepsinthedecentralizationprocess,butasyet
thesedonotformaconsistentframeworkfordistributingbudgetaryresources.This
discussion paper by Hamish Nixon and Cindy Joelene intends to inform the wider
discussionabouthowbesttoproceedwithfiscaldecentralizationinMyanmar.Itsgoal
is to inform policy makers, civil society, political parties, and international
development partners of principles and processes to guide fiscal decentralization
policy. The paper presents ideas for a fiscal decentralization roadmap that are
groundedinthecountryscontext,andthatbuildonexistingstructuresandreforms
whileleavingspaceforthelongertermevolutionofthedecentralizationprocess.
TheUnitedKingdomsDepartmentforInternationalDevelopment(DFID)generously
funded this report. The opinions expressed in this report are solely those of the
authors and do not necessarily reflect those of DFID, MDRICESD, or The Asia
Foundation.
Dr.ZawOoDr.KimNinh
ExecutiveDirectorCountryRepresentative
MDRICESDTheAsiaFoundation
TABLEOFCONTENTS
Acronyms
ExecutiveSummary
ii
ONE:FiscalDecentralizationinTheoryandPractice
1.1 THETEXTBOOKAPPROACHTOFISCALDECENTRALIZATION,ANDITSLIMITS
1.2 BETWEENGRANDPLANSANDMUDDLINGTHROUGH
3
5
TWO:MyanmarsFiscalDecentralizationEnvironment
2.1 POLITICALANDINSTITUTIONALCONTEXTFORDECENTRALIZATIONREFORMINMYANMAR
2.2 MYANMARSCURRENTSUBNATIONALFISCALSTRUCTURE
7
10
THREE:SketchingMyanmarsFiscalDecentralizationRoadmap
17
3.1 CHANNEL1:STRENGTHENINGSECTORALDECONCENTRATION
HORIZONTALALLOCATIONANDLOCALINPUTTOPLANNINGANDBUDGETING
3.2 CHANNEL2:RATIONALIZINGSTATEANDREGIONBUDGETS
3.3 CHANNEL3:ALLOCATINGSUBNATIONALDEVELOPMENTFUNDS
UNIONGOVERNMENTFUNDS:THEPOVERTYREDUCTIONFUND
THECONSTITUENCYDEVELOPMENTFUND
DONORPROVIDEDFUNDS:COMMUNITYDRIVENDEVELOPMENT
IMPLICATIONSANDCHALLENGES
3.4 TOWARDSASTATE/REGIONTRANSFERSYSTEM
18
24
27
29
31
32
34
35
37
FOUR:ManagingMyanmarsFiscalDecentralization
43
4.1 THETHREETRACKSTRATEGY
4.2 INSTITUTIONSANDPROCESSESFORSUCCESSFULFISCALDECENTRALIZATION
43
45
Annex1:Political,administrativeandfiscaldimensionsof
decentralization
49
Annex2:Designingtransferschemes
50
Annex3:SchedulesTwoandFiveoftheMyanmarConstitution
53
REGIONORSTATELEGISLATIVELIST(SCHEDULETWO)
TAXESANDFEESCOLLECTEDBYREGIONORSTATES(SCHEDULEFIVE)
53
54
Annex4:ElementsforMyanmarsDecentralizationStrategy
56
References
57
Acronyms
CDF
ConstituencyDevelopmentFunds
GAD
GeneralAdministrationDepartment
MBA
MinistryofBorderAffairs
MDRICESD
MyanmarDevelopmentResourceInstituteCentrefor
EconomicandSocialDevelopment
FESR
FrameworkforEconomicandSocialReform
MIMU
MyanmarInformationManagementUnit
MoF
MinistryofFinance
MNPED
MinistryofNationalPlanningandEconomicDevelopment
NCDD
NationalCommunityDrivenDevelopmentProject
PFM
PublicFinancialManagement
PRF
PovertyReductionFundsorPovertyEradicationFunds
SEE
StateEconomicEnterprises
SPDC
StatePeaceandDevelopmentCouncil
TDAC
TownshipDevelopmentAffairsCommittee
TDSC
TownshipDevelopmentSupportCommittee
TMC
TownshipManagementCommittee
ExecutiveSummary
Intheory,decentralizationcancreateimportantbenefitsconsistentwiththepolicy
prioritiesofthegovernmentandtheneedsofMyanmarspeople.Decentralization
issupposedtoalignpublicservicesmorecloselytothepreferencesandneedsoflocal
people, increasing responsiveness. By enabling stronger accountability and
transparency,andreducingadministrativebottlenecks,decentralizationmayimprove
thetechnicalefficiencyandimprovethequalityofpublicservices.Andasameansto
achieve greater equity and participation, decentralization may support state
legitimacy,andhelpaddresssome,thoughperhapsnotall,kindsofinternalconflict.
But,decentralizationisveryriskyanddifficultinpractice.Thereisevidencethaton
the whole more fiscally decentralized countries enjoy higher income and other
benefits,butthereislittleevidencethattheprocessofdecentralizationsystematically
produces these benefits. This gap between theory and reality is a result of lack of
attentiontocontextualfactorsandpoordesignorimplementationofdecentralization
reforms.Successfuldecentralizationreformsentailaligningdecentralizationstrategy
withthepoliticaldriversfordecentralizing,theinstitutionalstartingpoints,andthe
capacityoftheexistingsystem.
ThispaperintroducesaroadmapforMyanmarsfiscaldecentralizationthatis:(1)
basedinMyanmarscurrentpolitical,institutional,andfiscalrealitiesandaddresses
key local policyaims;(2)informedbyinternational andtheoretical lessons,and(3)
cautiousandflexibleenoughtolimitfiduciaryandcapacityriskswhileallowingforthe
possibilityofmoresignificantfuturereforms,includingdevelopmentsinMyanmars
peaceprocess.
The process of decentralization has accelerated since the Thein Sein government
came into power in 2011. The government has emphasized peoplecentered
development,creatingconsultativebodiesorelectedpositionsatdistrict,township
andvillagelevels.Ithasalsoorderedstateandregionlevelcivilservantstocoordinate
withstateandregiongovernmentoverUnionlevelresponsibilities,andincreasedthe
budgetsavailableforlocalpublicfunctionsanddevelopmentprojects.
However, important longerterm political aspects of the decentralization process
are still unclear. Several major policy processes that are still at an early stage will
shape the longterm direction of Myanmars decentralization. There are active
discussionsaboutconstitutionalchange,bothwithinthelegislativebranchandamong
political parties as the election draws nearer. It is also clear that should the peace
processproceedtothepoliticaldialoguestage,thepositionofmostethnicarmed
groupscallsforareorientationofcenterlocalrelationstowardsamorefederalmodel.
In large territories of contested governance, administration takes place via various
ii
hybrid or nonstate systems, and decentralizing central government services has

politicalimplicationsthatmaybearonthepeaceprocess.Finally,achievingincreased
transparency, equity, sustainability and justice in the sharing of natural resource
revenuesisintimatelyrelatedtothebroaderissueofeffectivefiscaldecentralization,
aswellasthepeaceprocess.
Myanmarsreformisatabalancepoint.Theendgoalsofthefiscaldecentralization
aredifficulttosee,buttherearestillopportunitiestoachieveresultswhilelayingthe
groundforalongertermdecentralizationprocessofwhichthefullscopeandnature
isasyet,unclear.Todothismeansbuildingupontheinstitutionsthatarecurrently
functioningandthereformsthathavebeenundertaken,whileleavingspaceformore
thanonepossiblefuture.
Myanmarscurrentsubnationalfiscalstructureisamixtureofhorizontalandvertical
decentralization.TherearethreemainchannelsforfiscalresourcesfromtheUnion
tothestate/regionlevel:
Channel 1: Line ministries or departments that remain under Union

jurisdiction,andwhichhaveassigned,toagreateroralesserdegree,some
activities and their accompanying budget to their state and region
departments,throughadeconcentratedchannel.
Channel2:TheUnionbudgetgivesaidintheformofgrantsorloanstothe
state/region fund to support a range of devolved departments through a
devolution channel; these state/region budgets are also supported by tax
andnontaxrevenuesandlocalSEEincome.
Channel3:Variousseparatelybudgetedfundsmainlyforcapitalexpenditure
on local infrastructure, also devolved to the state/region level, local
parliamentarians, or townships according to various distribution and
managementcriteriadetailedbelow.
The thrust of government fiscal decentralization policy has been to increase the
shareofnationalexpenditureincludedinstateandregionbudgets,whichtripledin
thelastyearto11.8percent.Thistypeofpolicyisquitecommonaroundtheworld,
butisinadequatebecauseitfocusesattentiononthestateandregion. Thus itdoesnot
addressthecentralsystemsfordeterminingwhythecentershoulddevolveacertain
shareofthebudget,orhowthatshareshouldbedividedamongdifferentstatesand
regions. It also focuses on one or two of the channels between the center and
state/regionlevel,andignoresthemajorsectordeconcentrationchannel.
AverticaldeconcentrationprocessforimportantsectorsinChannel1hasalready
begun in Myanmar. These steps represent important advances in decentralizing
service delivery. However, they do not as yet form a systematic and transparent
framework for allocating these resources across the country. An incremental fiscal
decentralizationstrategyshouldbuilduponthembyintroducingamoresystematic
iii
approachtothedeconcentrationoffunctions,providingtheresourcestofundthem
(verticalfiscalbalance),anddistributingtheseresourcesacrossthecountryinlinewith
policyobjectives(horizontalfiscalbalance).
Budgetingforstate/regiondepartmentsiscurrentlybasedonthedeficitbetween
departmentsrevenueandexpenditure,oradhocnegotiationswiththecentrallevel.
Thismethodofbudgetpreparationresultsininequitablebudgetsthatmirrorcurrent
ratesofspending,notwhatspendingshouldbetocarryoutassignedresponsibilities.
Whendeficitsarethemaindeterminantofthelevelsoftransfers,statesandregions
andotherlocalgovernmentsdonothaveanincentivetoimproverevenuecollection.
Myanmarhasalsointroducedanumberofoverlappingcommitteesandsubnational
development funds broadly aimed at poverty alleviation and local development
throughincreaseddiscretionaryresourcesandbottomupplanning.Differentfund
structures and the proliferating committees that have grown up to ensure local
participation carry overlapping objectives and mandates, which make coordination
difficult and risks poor targeting of resources. At the same time, the complex and
varied allocation procedures mean that the overall effect of these flows on the
horizontaldistributionofresourcesacrossthecountrymaybedifficulttodiscern.At
a time when Myanmar should be aiming for greater budget integration, there is
actuallyconsiderablebudgetfragmentationoccurring.
This report proposes an incremental fiscal decentralization reform strategy with
three reform tracks, corresponding to the major existing channels of subnational
resourceflowstostatesandregions.
The first track consists of a framework for sectoral deconcentration built on a
systematic analysis and adjustment of functional assignments, followed by the
introduction of sector budgeting formulae to improve subnational budget
allocationswithinmajorlineministries.Thisprocessmustruninparallelwithchanges
to budget preparation and presentation to allow ceilings and final allocations to
state/regiondepartmentstobetransparentwithintheunionbudget,andtoallowline
ministries to provide budget ceilings for deconcentrated functions and activities to
theirstateandregiondepartments.
Alongsidetheprocessofsectoraldeconcentration,Myanmarshouldmovetowards
asystemoftransfersthatworkstomeettheobjectivesofthegovernmentforits
devolved activities in the second and third channels. The immediate priority is to
move from the current system of budgeting for state/region transfers to one that
rationalizes (1) the vertical balance of resources between levels; (2) the horizontal
balanceacrossthecountry;(3)theincentivesforownsourcerevenuecollection,while
(4)notoverwhelmingfiscalorimplementationcapacityinweakorsmallstatesand
regions.
iv
The best starting point for designing astate/region transfersystem for Myanmar
would be for those responsibilities already under the control of state/region
government, and the various development aims of the poverty reduction funds.
Based on analysis of Myanmars situation, a tentative transfer system proposal for
stateandregiontransfersmightinclude:
an unconditionalblock grant forordinaryfunctionsassigned tothatlevel,

possibly with either a matching or mild performancebased component
aimed at improving planning and revenue performance relative to local
capacityandconflictsensitivities;and,
anadditional unconditionalgrantaggregatingthePovertyReductionFund
and, if politically feasible, the Constituency Development Fund, but with
guidelinesforcoordinationwithdeconcentratedsectorservices;
a possible general purpose grant for equalization, compensation or
derivationbasedonnaturalresourcerevenues.
Figure1:Threereformtracks
Track1
Track2
Track3
Coordinatingandeventuallyintegratingthesethreetracks,isamassivepolicytask,
but there is currently no clear policy home for this activity within the Myanmar
government. Decentralization can demonstrate benefits quickly for example
throughlocaldevelopmentprojectsandnewstaffinsectordepartmentsbutitisnot
aquickwinreform.Itrequiresaconsistentandclearefforttoestablishobjectives,
coordinate many stakeholders, and develop and implement overarching policies.

Thereisanurgentneedforanationalpolicycoordinationcapacity.
Thepolicycoordinationmechanismneedstobalancetherequirementsofpolitical
decisionmakingandsupportforsoundtechnicalanalysisandpolicydevelopment.
As decentralization is a crosscutting issue that influences public financial
managementofallkinds,itneedsparticipationofallkeybudgetaryinstitutions.
A high political body such as a national commission or secretariat is likely most
suitedtoMyanmarscontext wherepoliticaldecentralizationleadsadministrativeand
fiscal dimensions and its pending highlevel questions about the constitution, the
peaceprocess,andnaturalresources.
In addition to a higher level body to guide policy with wider participation, the
Financial Commission should be given through law in accordance with the
constitutionapolicyroletodesignthefiscalframeworksfordeconcentrationand
devolution called for in the roadmap introduced in this paper. The Commission
shouldmovefrombeingaforumforintergovernmentalbargainingtobeingapolicy
settingbodyforthetransfersystem.Todosowouldrequiresubstantialinvestmentin
atechnicalsecretariat,andlegislativerequirementsforittoconsultwidelywithinand
beyondgovernment.
vi
ONE:FiscalDecentralizationinTheoryandPractice
Along with other reforms, Myanmar is undertaking the most significant
decentralization in its history. The 2008 Constitution established state and region
governments with partially elected assemblies, or hluttaws, under the direction of
centrallyappointedChiefMinisters.1Morerecentlythegovernmenthasestablished
variousindirectlyordirectlyrepresentativebodiesattownshipandvillageorvillage
tractlevels,thoughthesefallshortofdevolvedlocalgovernmentsandaredependent
oncentralizedinstitutions,particularlytheGeneralAdministrationDepartment(GAD)
oftheMinistryofHomeAffairs.2ThegovernmentofMyanmarrecognizesthatthere
is still confusion over the administrative and expenditure roles of different
administrativelevels,anddifficultiescoordinatingUnionlevelfunctionswithstateand
region governments.3Various presidential orders have been issued to increase the
roleofstateandregionministers,andforUnionministriestodecentralizesomeof
theirfunctionstotheirlocaloffices.4
Themantraofpeoplecentereddevelopmenthasbeenwidelyproclaimedbycentral
andlocalofficials,andthegovernmentsFrameworkforEconomicandSocialReform
(FESR) attaches a high priority to developing a participatory process of local
budgeting,whichshouldreflectlocalprioritiesandneeds,whilecorrespondingwith
nationalpolicydirections.5Arangeofpotentialandactualrevenuesourcesnowlie
with the subnational levels of administration in line with Schedule Five of the
Constitution,includinglandandexcisetaxes,arangeoffees,andrevenuesfromafew
of the smaller State Economic Enterprises. 6 Along with these measures, there has
beenasignificantdecentralizationofthegovernmentbudgettothenewlyestablished
stateandregionbudgets.IntheyearfromFY201314toFY201415theshareofthe
Union budget estimate assigned as aid to state and region governments tripled,
fromabout3.6percentto11.8percent.7

1
HamishNixonetal.,StateandRegionGovernmentsinMyanmar(Yangon:TheAsia
Foundation/MDRICentreforEconomicandSocialDevelopment,September2013).
2
KyiPyarChitSawandMatthewArnold,TheGeneralAdministrationDepartment:AnOverviewof
MyanmarsAdministrativeBackbone(Yangon:TheAsiaFoundation/MDRICentreforEconomicand
SocialDevelopment,forthcoming2014).
3
GovernmentoftheUnionofMyanmar,FrameworkforEconomicandSocialReform,December
2012,Para.105.
4
SpeechofPresidentTheinSeindeliveredtoUnionMinisters,RegionandStateChiefMinisters,and
DeputyMinisters(9August2013).
5
GovernmentoftheUnionofMyanmar,FrameworkforEconomicandSocialReform,Para.106.
Peoplecentereddevelopmentfiguresprominently,amongelsewhere,inSpeechofPresidentThein
Sein(9August2013).
6
ConstitutionoftheRepublicoftheUnionofMyanmar,ScheduleFive(reproducedasAnnex4).
7
Unionbudgetlaws201314,aspublishedinUnionGovernmentGazette.

Intheory,decentralizationcanhaveimportantbenefitsconsistentwiththepriorities
ofthegovernmentandtheneedsofMyanmarspeople.8Decentralizationissupposed
to increase the responsiveness of public services to the preferences and needs of
communities. 9 By enabling more accountability and transparency, and reducing
administrativebottlenecks,decentralizationmayimprovetheefficiencyandqualityof
publicservices.Andifitsucceedsinachievinggreaterequityandparticipationacross
differentpartsofthecountry,decentralizationmaysupportstatelegitimacy,andhelp
addresssomekindsofinternalconflict.
Aparadoxofdecentralizationisthatwhilethereisevidencethatonthewholeand
inthelongtermmorefiscallydecentralizedcountriesenjoyhigherincomeandother
benefits, the process of decentralization does not produce these benefits easily,
particularly in developing countries. 10 Instead, there is broad agreement that
althoughtherehavebeenmeaningfulandproductivedecentralizationreformsand
developmentpartnersupportprogramsmanyperformwellbelowexpectations.11
Decentralization in general, and fiscal decentralization in particular, is risky and
difficultinpractice.
Thisgapbetweentheoryandrealityisnowunderstoodtocomefromlackofattention
tocontextualfactorsandpoordesignorimplementationofdecentralizationreforms.
Key sources of these weaknesses include a failure to recognize and work with the
politicaldriversofthedecentralizationprocess;inconsistenciesininstitutionaldesign
and poor integration of political, administrative and fiscal dimensions of
decentralization (see Annex 1); and sequencing mistakes that overwhelm existing
capacity.12There can be many political drivers for decentralization. The population
may feel that decentralization will improve government performance; groups of

8
JesperSteffensen,FiscalDecentralizationandSectorFundingPrinciplesandPractices(DANIDA,
November2010),3.Thereisanextensiveliteratureonthetheoreticalbenefitsoffiscal
decentralizationandfiscalfederalismdatingfromthe1950s.AusefuloverviewisPaulSmoke,Fiscal
FederalisminDevelopingCountries:AReviewofCurrentConceptsandPractices(NewYork:United
NationsResearchInstituteforSocialDevelopment,2001).
9
Ineconomicsthisformofresponsivenessiscalledallocativeefficiencyandisconsideredmost
importantwhenthereisheterogeneityofdemand,orwidelyvaryingneedsandpreferencesacross
differentplaces.
BalzsgertandHansjrgBlchliger,DecentralisationandEconomicGrowthPart2:TheImpacton
EconomicActivity,ProductivityandInvestment,OECDWorkingPapersonFiscalFederalism,(June3,
2013):http://www.oecdilibrary.org/taxation/decentralisationandeconomicgrowthpart2the
impactoneconomicactivityproductivityandinvestment_5k4559gp7pzwen[accessed26May
2014].
11
KentEaton,KaiKaiser,andPaulJ.Smoke,ThePoliticalEconomyofDecentralizationReforms
(Washington,D.C:WorldBank,2011),7.
12
Foracomprehensiveanduptodatereviewoftheliteratureonimplementationproblemsin
decentralization,seeLocalDevelopmentInternational,TheRoleofDecentralisation/Devolutionin
ImprovingDevelopmentOutcomesattheLocalLevel:ReviewoftheLiteratureandSelectedCases
(DFID/LocalDevelopmentInternational,November2013).
10

citizens who are disconnected due to ethnic or cultural difference may seek
decentralizationtorecognizetheirdifferences;andpoliticiansandbureaucratsmay
support it to generate electoral legitimacy or appease political demands for
autonomy. 13 Successful decentralization reform strategies must align with the
prevailingpoliticaldriversfordecentralizing,theinstitutionalstartingpoints,andthe
capacityoftheexistingsystem.
Withinsuchastrategy,fiscaldecentralizationtheadequateandeffectiveresourcing
oftheresponsibilitiesgiventosubnationaladministrationiscentraltorealizingthe
benefitsofdecentralization:
Whereasthedecisiontodecentralizeispolitical,theeconomicandfinancial
payoffflowsfromawelldesignedsystemoffiscaldecentralizationthatis,
theintergovernmentalsortingoutofresponsibilitiesforexpendituresand
financingamongthevarioustypes,tiers,orlevelsofgovernment,ina
mannerthatisinharmonywiththepoliticalframework.14
Ifdifferentcontextsanddesignscanleadtosuchdifferentresults,howcanacountry
designafiscaldecentralizationstrategythatismostlikelytosucceed?
1.1 Thetextbookapproachtofiscaldecentralization,anditslimits
Technical input to decentralization processes tends to suggest a common set of
buildingblocksandasequencedprocesstodefineafiscaldecentralizationstrategy,
shaped by normative principles and considerations of best or good international
practice.15Fundamentalamongthesebuildingblocksarewhatareoftencalledthe
fourpillarsoffiscaldecentralization,eachwiththeirownsetofbestpractices:

13
RoyW.BahlandJorgeMartinezVazquez,SequencingFiscalDecentralization(WorldBank
Publications,2006),7.
14
AbduMuwongeandRobertD.Ebel,IntergovernmentalFinancesinaDecentralizedWorld,in
MunicipalFinances:AHandbookforLocalGovernments,ed.CatherineFarvacqueVitkovicandMihaly
Kopanyi(Washington,DC:WorldBank,2014),2.
15
Eaton,Kaiser,andSmoke,ThePoliticalEconomyofDecentralizationReforms,7.

1. Assignmentofexpenditureresponsibilitiesaccordingto associatedprinciples
of public finance such as subsidiarity, economies of scale, externalities and
equity.
2. Assignmentofrevenuesourcesaccordingtothetypesofservicesthatareto
be funded, avoiding economic distortions, and promoting subnational fiscal
autonomy.
3. Design of intergovernmental transfer systems to meet any gap between
revenueandexpenditureforassignedfunctions,andtopromoteequityacross
locationsandotherpolicygoals.
4. Subnational borrowing and fiscal discipline to ensure longterm investment
andpreventfiscaldecentralizationfromcreatingexcessdebt.
Internationalexperiencesuggestsadditionalelementsthatareessentialforasound
fiscal decentralization process, including effective civil service rules, financial
managementandcoordinationprocedures,andclearpoliticalaccountability.16Since
eachofthesebuildingblocksneedscarefultechnicalconsideration,thereisanideal
or best practicesequence for thedevelopment of fiscal decentralization strategy
(Figure2).Broadlyspeaking,assumingaminimumsetofstatecapacitiesalreadyexists,
an inclusive and popular debate about the goals of decentralization should be
followed by the development of a policy and a program structure to carry it out,
passageofnecessarylegislation,andimplementationsupport.
Figure2:Anormativeapproachtosequencingfiscaldecentralization17
Enablingadministrativeand
institutionalstructures
Inclusive(preferablyformal)
Nationaldebate
Programandpolicy(white)
paper
Proceduresfor
intergovernmentaldisputes
Orderlybudgetprocessandfiscal
rules
Election
Peaceprocess
Nationalcommission
Componentsofprogram
Timetable
Guidanceforwritinglaws
Organiclawor
Constitutionalprovisionsor
amendment
Regulations
Implementation
PublicFinancialManagement
systems
Transferofcivilservants
Shortandmediumtermaction
plans
Programsupportfrompartners
Monitor,evaluateandrevise
Possiblereferendum
Whileeachofthesestepsareimportanttotheoverallsuccessofreforms,theyare
rarely implemented this way in practice a problem that is also acknowledged by
manyofthoseproposingthem.Thisidealizedapproachwhereeachstephappensin

16
BahlandMartinezVazquez,SequencingFiscalDecentralization,14;Steffensen,Fiscal
DecentralizationandSectorFundingPrinciplesandPractices,5.
17
AdaptedfromBahlandMartinezVazquez,SequencingFiscalDecentralization,25.

therightorderisusuallynotpossible,andintherealworldthereisnosinglebest
approachtosequencingfiscaldecentralization.18
Therearemanyexamplesoffailedordelayedgranddesignsfordecentralization.In
Afghanistan, a Subnational Governance Policy of 400 pages was painstakingly
developed with extensive and costly international assistance in 2008, but remains
largelyunimplemented.InNepal,manyelementsoftheLocalSelfGovernanceActof
1999 remain on paper 15 years later, with intentions to decentralize services
unfulfilled and only a small proportion of the budget given to local governments.
Cambodia,aftertenyearsofpostconflictstatebuilding,onlyembarkedonamodel
processwitha2005strategy,asubnationalOrganicLawin2008,andthebeginningof
atenyeardecentralizationprogramin2010.19
Inreality,thetextbookapproachtodecentralizationconfrontstheconstraintsset
by the underlying political motivations and conditions for reform, which are often
differentfromthetheoreticalbenefits.Asaresult,
Countriesembarkingondecentralizationoftenstrugglewithdecisionsabout
theessentialcomponentsofdecentralizationincludingtheorderofan
introductionofdecentralizationpolicies,thenumberofyearsnecessaryto
bringafullprogramonline,andthecomponentsofthetransitionstrategy.20
As well as difficulties with clarifying policy aims, the design and implementation of
decentralization policy often encounters political and bureaucratic resistance as it
proceeds:Intherealworldthereareimportantconstraintsthatsteercountries
awayfromsuchanoptimalsequencing.Andsometimes,thesedeparturesareinthe
bestinterestsofgettingthejobdone.21Forsuccessfuldecentralizationreformsaway
mustbefoundbetweentheidealsequenceandtheserealpoliticalconditions.
1.2 Betweengrandplansandmuddlingthrough
Recently, development policy research has focused on the practical challenges of
successfully implementing public sector reforms including decentralization
programs. 22 To be successful, reforms should build on contextual and institutional

18
Ibid.,2.
OnCambodiaspainstakingdecentralizationprocessseePakKimchoeun,FiscalDecentralisationin
Cambodia:AReviewofProgressandChallenges,CDRIWorkingPaperSeries50(PhnomPenh:CDRI,
2011);TariqH.Niazi,DeconcentrationandDecentralizationReformsinCambodia(Philippines:Asian
DevelopmentBank,2011).
20
BahlandMartinezVazquez,SequencingFiscalDecentralization,1.
21
Ibid.,5.
22
SeeLeniWildandMartaForesti,WorkingwiththePolitics,2013,
http://www.odi.org.uk/sites/odi.org.uk/files/odiassets/publicationsopinionfiles/8587.pdf;Matthew
Andrews,TheLimitsofInstitutionalReforminDevelopmentChangingRulesforRealisticSolutions
19

realities, address specific and grounded problems, and proceed incrementally with
ongoinglearningandadjustment:[e]venthoughaclearvisionoflongertermreform
is needed, initial steps should be modest and logically phased. 23 For example,
policymakers can take advantage of momentum offered by other reforms, as in
Indonesia,PolandandEstonia.
While a lot of knowledge exists about what one should do to fiscally decentralize
successfully,developinganunderstandingofwhatonecandoandaworkableplan
todoitisamorecomplexanddifficultprocess.Thispaperisanefforttosupport
thatprocessinMyanmar.ItintroducesapotentialroadmapforMyanmarsfiscal
decentralization that is: (1) based in Myanmars current political, institutional, and
fiscalrealitiesandaddresseskeylocalpolicyaims;(2)informedbyinternationaland
theoretical lessons, and, (3) cautious and flexible enough to limit risks to public
financesandservices,whileallowingforthepossibilityofmuchmoresignificantfuture
reformsincludingdevelopmentsinMyanmarspeaceprocess.
It is important to begin from where we are now, so the next section reviews the
current political and institutional context for Myanmars fiscal decentralization
reforms,andintroducesthemainexistingchannelsofthesubnationalfiscalsystem.
Thethirdpartofthepaperreviewsthecurrentstateofreformacrosseachofthese
channels, and introduces a framework for strengthening each of them, while
preparing the ground for longerterm reforms. The last section emphasizes the
importanceofcoordinatinginstitutionsandpoliciestomanagesuchcomplexreforms,
andpresentssomerecommendationsforthem.

(Cambridge:CambridgeUniversityPress,2013).Thisapproachhasbeengiventhesomewhat
inaccessiblelabel:ProblemDrivenIterativeAdaptation,andechoesearlierworkonincremental
policychange:CharlesE.Lindblom,TheScienceofmuddlingthrough,PublicAdministrationReview
19,no.2(Spring1959):7988.
23
PaulSmoke,StrategicFiscalDecentralizationinDevelopingCountries:LearningfromRecent
Innovations,inLocalDynamicsinanEraofGlobalization,ed.ShahidYusuf,WeipingWu,andSimon
Evenett(WorldBank,2000),107.
TWO:MyanmarsFiscalDecentralizationEnvironment
Effective decentralization strategies need to build on existing context. Important
contextualfactorsincludethepoliticaldriversbehinddecentralization,therolesand
interestsofkeyactorsandstakeholders,andthecurrentstateofreform.24Keyactors
and stakeholders may be political or bureaucratic, and it is important to include
development partners among them.25Reforms also need to start with the existing
fiscalinstitutionsforplanning,budgetingandspendinginstitutions,evenwhileaiming
toimprovethem.
2.1 PoliticalandinstitutionalcontextfordecentralizationreforminMyanmar
Myanmarsformalprocessofdecentralizationbeganwiththeadoptionofthe2008
Constitution, itself the culmination of the State Peace and Development Council
(SPDC) regimes carefully managed sevenstep road map towards a modern,
developed, and democratic nation. As part of a carefully orchestrated transition,
decentralizationhasbeensubjecttoclearlimitsthatareapparentinconstitutional
provisions relating to the structure of the state. In particular, the constitution sets
somecrucialconstraintsonMyanmarsearlyfiscaldecentralizationprocess.
First,itestablishesamixedpoliticalsystematthestateandregionlevelthatcombines
a semielected local representative body with an, essentially, centrally appointed
Chief Minister. 26 This type of arrangement is quite common in decentralizing
countries:
Fullhorizontalaccountabilitytovotersislimitedinotherdevelopingand
transitioncountriesbylessthanfulldirectelectionsofexecutiveand
legislativebranchesofgovernment.Inacommonmodality,thelegislative
branchisdemocraticallyelectedbuttheexecutivebranchisappointedbythe
center.27
Thiscombinationallowsthecentrallyappointedexecutivetooverseetheactivitiesof
newlydemocraticlocalgovernments,andexercisecontroloverimplementation,and

24
Eaton,Kaiser,andSmoke,ThePoliticalEconomyofDecentralizationReforms,8.
Thereisevidencethatdevelopmentpartnerpracticesposeparticularriskstodecentralizationin
moreaiddependentcountries:LocalDevelopmentInternational,TheRoleof
Decentralisation/DevolutioninImprovingDevelopmentOutcomesattheLocalLevel:Reviewofthe
LiteratureandSelectedCases,4044.
26
ConstitutionoftheRepublicoftheUnionofMyanmar(MinistryofInformation,Governmentof
Myanmar,2008),Art.261.Asinthenationalparliament,25percentofstateandregion
representativesaredirectlyappointedbytheCommanderinChiefofthearmedservices.
27
BahlandMartinezVazquez,SequencingFiscalDecentralization,16.Oncomparisonbetween
CambodiaandMyanmar,seeLeeJones,ThePoliticalEconomyofMyanmarsTransition,Journalof
ContemporaryAsia44,no.1(January2,2014):2024.
25

is particularly common where there is an emphasis on centrally dominated and
managedreformprocesses,asinRussiasregionsorCambodiasprovinces.
Second, the constitution mandates important administrative roles in the state and
regiongovernments,intheselfadministeredzones,andintheNayPyiTawcouncil,
for the General Administration Department under the direction of the militaryled
Home Affairs ministry. 28 Finally, the constitution establishes a division of
responsibilities between central and subnational levels that retains many
developmentally and fiscally important responsibilities such as health, education
andmajoreconomicsectorsasUnionlevelresponsibilitiesandthereforesubjectto
theUnionbudgetprocess(SeeAnnex3).
In addition to the constitutional constraints, Myanmars public sector transition is
takingplaceviaacivilservicethathaslongbeencharacterizedbyahighdegreeof
centralization, a weak degree of administrative and managerial autonomy and an
almost nonexistent consultative process. 29 At subnational levels, the bureaucracy
were sidelined by the Tatmadaws regional commander system until the return of
civiliangovernment after 2011. In addition to this bureaucratic culture, Myanmars
politicalcultureechoesthatofsomeneighbours,withwide acceptanceofpoliticaland
bureaucratic elites, patronage, limited experience of accountability, and vertically
managed approaches to managing territory through governors or officers. 30 This
organizational culture means that reform is primarily being accomplished via top
downfiatratherthanmoreorganicorlocallydrivenprocesses,andthatambiguities
created by changing policies from above tend to result in stasis rather than
innovation.31
Despite these somewhat rigid foundations, the process of decentralization has
acceleratedandcreatednewspacesforreformsincetheTheinSeingovernmentcame
into power in 2011. This decentralization is aimed at rapid improvement of public
services and development with quick wins, bolstering efforts to conclude a
nationwide ceasefire and advance the peace process, and to generate legitimacy
beforethe2015election.32Concretedecentralizationreformssince2011havethus
had a primarily developmental focus. The government has emphasized people
centereddevelopmentbycreatingconsultativebodiesorelectedpositionsatdistrict,

28
ConstitutionoftheRepublicoftheUnionofMyanmar,Art.260,283,287.SeealsoKyiPyarChit
SawandArnold,TheGeneralAdministrationDepartment:AnOverviewofMyanmarsAdministrative
Backbone.
29
AlexM.Mutebi,MyanmarsCivilBureaucracyandtheNeedforReform,inMyanmar:Beyond
PoliticstoSocietalImperatives(Singapore:InstituteofSoutheastAsianStudies,2005).
30
TariqH.Niazi,DeconcentrationandDecentralizationReformsinCambodia(Philippines:Asian
DevelopmentBank,2011),ix.
31
Nixonetal.,StateandRegionGovernmentsinMyanmar,35.
32
Jones,ThePoliticalEconomyofMyanmarsTransition,12.

townshipandvillagetractlevels,althoughthesebodiesaremostlyappointedandnot
necessarilyrepresentative.Ithasalsoorderedstateandregionlevelcivilservantsto
coordinatewithstateandregiongovernmentoverUnionlevelresponsibilities,and
increasedthebudgetsavailableforlocalfunctionsanddevelopmentprojects.These
efforts are limited, but they have begun to address an imbalance between the
politicallysignificantestablishmentofstateandregiongovernmentsandtherelatively
weakadministrativeandfiscalsystemstosupportthem.33
However, the longerterm political direction of the decentralization process is still
unclear. Several majorpolicy processes that willshape Myanmars decentralization
arestillatanearlystage.Thereareactivediscussionsaboutconstitutionalchange,
bothwithinthelegislativebranchandamongpoliticalpartiesastheelectiondraws
near. 34 The constitutional review being managed through legislative committees
registered strong interest in amendment of the state structure provisions of the
constitution, but has not yet committed to supporting any particular reforms. The
reportofthejointparliamentarycommitteeonconstitutionalreformshowsthat161
of 162 submissions relating to Chapter Two of the Constitution (State structure)
suggest revisions or rewriting, while all 245 suggestions relating to Schedules One,
Two and Five setting out the powers of the states and regions, requested changes
(Annex3).35
Itisclearalsothatshouldthepeaceprocessproceedtothepoliticaldialoguestage,
the position of most ethnic armed groups calls for a reorientation of centrelocal
relations towards a more federal model, possibly through a redrafted rather than
amendedconstitution.36Itisavitalcaveattotherecommendationsinthispaperthat
in large territories of contested governance, administration takes place via various
hybrid or nonstate systems, and decentralizing central government services has
politicalimplicationsthatmaybearonthepeaceprocess.Thesituationintheseareas

33
Onthebalancebetweenpolitical,administrativeandfiscaldecentralizationintheseearlystagesof
stateandregiongovernmentseeNixonetal.,StateandRegionGovernmentsinMyanmar,6973.On
thepolitical,administrativeandfiscaldimensionsofdecentralization,seeAnnex1.
34
Seeforexample,NyeinNyein,USDPLeaderUrgesCommitteetoReviewKeyConstitutional
Reforms,TheIrrawaddy(18February2014).
35
Byfarthelargestnumberofrequests(5865separatesubmissions)forchangestotheconstitution
relatetoChapter12,whichsetsouttheprovisionsforconstitutionalamendments:Pyidaungsu
Hluttaw,ReportofConstitutionalReviewJointCommitteesFindingsandAssessment(NayPyiTaw:31
January2014).
36
BurmaNewsInternational,DecipheringMyanmarsPeaceProcess(ChiangMai:BNI,March2014),
3233.

isdescribedinEthnicConflictandSocialServicesinMyanmarsContestedRegions,a
newdiscussionpaperinthisseries.37
Finally,amongexistingstates/regionsandthecentre,politicalparties,andnonstate
politicalandarmedgroups,theissuesofincreasedtransparency,equity,sustainability
andjusticeinthesharingofnaturalresourcerevenuesareintimatelyrelatedtothe
broaderissueofeffectivefiscaldecentralization,atopictakenupbytheMDRICESD
andAsiaFoundationresearchpaperNaturalResourcesandSubnationalGovernments
inMyanmar.38
Myanmarsreformisthereforeatabalancingpoint.Unlikethetextbooksequence
that suggests a broad national consensus should precede reform, there is already
considerable political drive for some decentralization in action at both central and
locallevels.Butongoingdifferencesintheconstitutionalreformandpeaceprocesses
meanthatanationalconsensusisstilldeveloping.Inshort,thereisnotyetapolitical
settlement regarding decentralization or centrallocal relations in Myanmar.39The
endgoalsandfullscopeoffiscaldecentralizationarethereforedifficulttosee,but
therearestillopportunitiestoachieveresultswhilelayingthegroundforalonger
termdecentralizationprocess.Todothismeansbuildingupontheinstitutionsthat
arecurrentlyfunctioningandthereformsthathavebeenundertaken,whileleaving
thespaceformorethanonepossiblefuture.
2.2 Myanmarscurrentsubnationalfiscalstructure
These political and institutional factors mean that so far fiscal decentralization has
been largely driven from the top, focused on servicedelivery and development
spending, and hampered by limited administrative capacity. It has had to
accommodate the mixed political system for states and regions, and the division
between the relatively minor state and region responsibilities that have been
devolvedunderSchedulesTwoandFiveoftheConstitution,andthemoresignificant
Unionfunctions(Annex3).
Decentralization occurs through both horizontal (territoriallybased) and vertical
(sectorbased) processes. Horizontal decentralization refers to the assignment of

37
KimJoliffe,EthnicConflictandSocialServicesinMynamarsContestedRegions(Yangon:TheAsia
Foundation/MyanmarDevelopmentResourceInstituteCentreforEconomicandSocialDevelopment,
June2014).
38
ThetAungLynnandMariOye,NaturalResourcesandSubnationalGovernmentsinMyanmar
(Yangon:TheAsiaFoundation/MyanmarDevelopmentResourceInstituteCentreforEconomicand
SocialDevelopment/InternationalGrowthCentre,June2014).
39
Foranoverviewoftheconceptofpoliticalsettlement,seeThomasParksandWilliamCole,
PoliticalSettlements:ImplicationsforDevelopmentPolicyandPractice,OccasionalPaper(TheAsia
Foundation,July2010).
10

functions to a unified local government. In devolution, that local government has
autonomyoverdecisionsaboutthesefunctions,andisusuallymadeaccountablefor
those decisions to local people through elections. 40 In contrast, vertical
decentralization, sometimes called deconcentration, is the assignment of functions
and resources to lower administrative levels of a central organization (typically a
ministry),withaccountability forthesefunctionsstillbeingtothecentralauthority
(Box 1). A third form of decentralization, not as important for the purposes of this
paper,isdelegation.
Box1:Deconcentrationvs.devolution
Deconcentration is the redistribution of decision making, financial and management
responsibilityamongdifferentlevelsofthecentralgovernment.Itoccurswithinministriesor
toadministrativeunitsofthecentralgovernment,andisconsideredarelativelyweakformof
decentralizationasaccountabilityremainslargelyupwardtothecentre.
Devolution is the transfer of responsibilities and powers from the central government to
independentlocalgovernmentswiththeirowncorporateidentity,budget,legalstatusand
rights. Devolution still requires a national policy framework but involved downward
(democratic)accountabilityandusuallycoversseveralsectors.
There are important differences between devolution and deconcentration. In

devolutionthereistypicallymuchgreateraccountabilitytothepeoplebecausethe
local government has more decisionmaking power, and is elected. There is also
usually scope for local governments to make decisions about resource allocation
acrossdifferentsectors.Forthesereasons,devolutionisnormallyconsideredmore
politicallydecentralizedthandeconcentration.
Inreality,almostallcountrieshavesomemixtureoftheseformsofdecentralization:
Thetermfiscaldecentralizationencompassesthreedistinctarrangements
orvariants,eachofwhichhasaplaceinacountrysintergovernmental
financialsystem.Thethreearedeconcentration,devolution,anddelegation.
Animportantpolicyquestioniswhichofthethreevariantscanbesaidto
dominateanationspublicfinances.41
The defining characteristic of Myanmars current subnational fiscal structure is its
mixture of horizontal and vertical decentralization. In Myanmar, while there are
devolvedsubnationalgovernmentsinthestatesandregionswiththeirownsources
of revenue and a distinct budget for some functions, a large majority of fiscal
resourcesareretainedinthebudgetoftheUnionlineministries.Figure3illustrates
thishybridfiscalstructureatstateandregionlevel,usingafewsampledepartments
andministries.

40
41
Thisissometimesreferredtoasdemocraticdecentralization.
MuwongeandEbel,IntergovernmentalFinancesinaDecentralizedWorld,5.
11

Figure3:Myanmarshybridverticalandhorizontalfiscalstructuretostate/regionlevel
Channel1
Channel2
Channel3
Thereare,broadlyspeaking,threechannelsforfiscalresourcestoflowfromtheUnion
budgettosubnationallevels.
1. On the left are line ministries or departments that remain under Union
jurisdiction, and which may have assigned, to a greater or a lesser degree,
some activities and their accompanying budget to their state and region
departments(inred),throughthisdeconcentratedchannel(Channel1).
2. Ontheright,theUnionbudgetgivesaidintheformofgrantsorloanstothe
state/region fund to support a range of departments that operate under
state/regioncontrolthroughadevolvedchannel;thesestate/regionbudgets
arealsosupportedbytaxandnontaxrevenuescollectedatstate/regionor
townshiplevel,andlocalSEEincome(Channel2).
3. Finally, there are various separately budgeted funds mainly for capital
expenditure on local infrastructure, also devolved to the state/region level,
local parliamentarians, or townships according to various distribution and
managementcriteriadetailedbelow(Channel3).
This fiscal structure has important implications for both current and future fiscal
decentralization policies. First, each of these three channels has different
arrangements for determining both the total pool of available resources, and its
distributionacrossthedifferentpartsofthecountry.Thesearrangementsdetermine
12

ifadecentralizedfiscalsystemcancorrectverticalandhorizontalfiscalimbalances,
which refer to a mismatch between the expenditure needs and available revenue
betweenlevelsofadministration,andacrossdifferentpartsofthecountry(Box2).In
Myanmar,thearrangementsfor eachchannelarenotyetdesignedtocompensate
theseverticalandhorizontalimbalances.
Box2:Verticalandhorizontalfiscalbalance
Verticalfiscalbalanceistheratiobetweentheexpenditureneedsofeachlevelofgovernment
asawholeandthesourcesofrevenueavailabletoit.Averticalimbalanceexistsifrevenues
areinadequatetomeetspendingrequirements(unfundedmandates)or,lessoften,ifthey
exceedthem.
Horizontal fiscal balance is the degree of equity of available resources between different
subnational unitsat the same level, such as states or regions.Horizontal imbalanceexists
when some localities have more resources and others less for equivalent expenditure
requirements.
Second,thedivisionbetweenwhatresponsibilitiesfallundereachofthesechannels
isunclear.AsoutlinedinStateandRegionGovernmentsinMyanmar,someministries
areassignedtoeitherUnionorstate/regionlevel,buttheconstitutionisalsounclear
abouttheassignmentofotherfunctions,orevenassignsdifferentfunctionsof the
sameministrytodifferentlevels.42Asaresult,differentsubdepartmentsofthesame
Ministrymaybefundedthroughdifferentchannels,withsomeformingpartoftheir
parentministrybudget,whileothersareassignedtothestate/regionfund.Inpractice
instatesandregions,thismeansthatsomedepartmentshavebudgetsviaboththe
Unionandstate/regionbudgetchannels,whichcanleadtopooruseofresources,and
weakaccountability.
State/region budget staff confirm that some departments receive funds from both
channels,andsometimesspendavailabledeconcentratedunionfundswhileleaving
state and region funds unspent. The budget process is also not set up to allow
coordination of the funding sources: the Union ministries submit their budget
estimates before there is any opportunity for these to be discussed locally, while
state/region finance departments do not have time afterwards to discuss needs
departmentbydepartmentbeforesubmittingtheirstate/regionbudgetproposalto
thehluttawandthentheFinancialCommissionwhichfinalizesthestate/regionand
unionbudgetproposalsatnationallevel.43Thissituationweakensaccountabilityby
preventingaclearlinkbetweenthetwotypesoffinancialflowsandthefunctionsthey
aremeanttosupport.

42
Nixonetal.,StateandRegionGovernmentsinMyanmar,2526.Forexample,somedepartments
andactivitiesoftheMinistryofAgricultureandIrrigation,orForestryandMines,aredevolvedto
statesandregions,whileotherpartsofthesameministryarenot.
43
Interview,State/RegionBudgetDirector(March2014).
13

Finally,andrelatedtothesestructuralfeatures,eachchannelofsubnationalfinancial
resourcesispresenteddifferentlyinMyanmarsbudget.Thedevolvedaidtostates
andregions,andeachstateandregionsownsourcerevenues,arepresentedinunion
andstate/regionbudgetdocuments(Channels2and3).However,thedeconcentrated
funds for states and region departments of Union ministries in Channel 1 are not
specified in national budget documents; the parent Union ministry is the primary
budget unit, and there is currently no secondary budget classification that clearly
indicates Union ministries state/region spending in the national budget, making it
difficulttodistinguishthepoliciesforhowtheseministriesallocateacrossthecountry
interms.Ministriesdohavebudgetsfortheirsubnationalunitsbutthesearemainly
formanagementandbudgetcontrolreasons.
Together,thesefactorsmeanthatitiscurrentlyverydifficulttoanswerthemostbasic
question about fiscal decentralization: how much of Myanmars public spending is
decentralized,towhatlevels,andforwhatpurposes?Ifweunderstanddecentralized
public spending to be just the state and region budgets, then indeed we can see
increasedfiscaldecentralization.Thethrustofgovernmentpolicyhasbeentoincrease
theshareofnationalexpenditureincludedinstate/regionbudgets.Asnotedinthe
introduction,theshareoftheUnionbudgetassignedasaidtostatesandregions
indeedtripledinthelastyear,to11.8percent.Duringthe201415nationalbudget
preparation process the Ministries of Finance and National Planning and Economic
Development were under instructions from the Presidents Office to increase this
sharesubstantiallyinordertosupportdecentralizeddevelopment,potentiallyupto
15percent.44
This type of policy a topdown directive to give an increased share of revenue,
expenditure,orGDPtobedistributedtosubnationalunitsisquitecommonaround
theworld:
Politically,atthecentrallevel,itisrathereasyforaPresidenttoproposeand
aparliamenttoapproveapropositionforagivenpercentageofcentral
governmentrevenuestobesharedwithlocalgovernments.45
However,ithassomeveryimportantproblemsandperniciouseffects.46First,while
it does indicate the budget over which the state/region has some decision making
power,thisfiguredoesnotrepresentthetrueextentofsubnationalspendingsince
most of the expenditure on important social and economicservices is budgeted at
nationallevel.

44
Interviews,MinistriesofFinanceandNationalPlanning(November/December2013).
46
Ibid.,2.
45
14

Mostimportant,itremovesthelinkbetweenthefunctionsassignedtodifferentlevels
ofgovernment,andtheavailabilityofresourcestocarrythemout.Asexpressedby
governmentfinanceofficials:
Wefacedifficultiesinknowinghowtofinancethisexpenditureinstatesand
regions,becausetheguidelinesforwhatisstateandregionspendingarestill
unclear,andmuchspendingisforunionministries.47
Inpractice,thiskindofpolicymeansthatstateandregionbudgetsaremanipulated
andpiecedtogethertomeetapolicyquota,insteadofbeinganexpressionofpolicy
objectivesandpriorities,plans,andexplicitservicestandards.Oneresponse,adopted
already inMyanmar, is toshiftactivities from one budget to anotherarbitrarily,in
ordertomakeupthepoliticallydeterminedlevelofexpenditure:
ThePresidentsOfficeministerswerepushingforincreasedbudgetsforthe
statesandregions,soduringthebudgetnegotiationssomeworkswere
transferredontothestateandregionbudgets.48
Some of these works may not even be matched to state/region responsibilities
accordingtotheconstitution.Similarly,somestatesandregionsaretryingtoincrease
[Union]grantsbyshiftingSEEexpensestoministerialdepartments; byincreasingthe
deficitbetweenstate/regionrevenuesandexpenditurestheUnionwouldbeforced
toincreaseitsaid,makingthedistinctionbetweenuniongrantsandloansunclear
andinconsistentacrossdifferentterritories.49
Oftenalackofconnectionbetweenfundingsourcesandthefunctionstheysupport
results in unfunded mandates, when a subnational government is assigned
insufficient resources for its responsibilities. An opposite risk may be emerging in
Myanmar:giventhelimitedexpenditureresponsibilitiescurrentlydevolvedtostate
andregioncontrol,furtherincreasesinthestateandregionbudgetsmayinfacttake
resources away from essential social and economic services that are in the Union
budgetforexamplebyoverfundinglocalservicesatthecostofeducationorhealth.
Withoutcostingtheactualfunctionsunderstateandregioncontrol,itispossiblethat
the budgetshare approach to fiscal decentralization currently used will risk
overfundinganarrowrangeoflocalservices,whilecreatingshortagesandpotentially
largerdeficitsatnationallevel.Suchasituationinwhichlocalgovernmentswereable
tooverspendwhilecentralgovernmentbecamecashstrappedoccurredinChinain
the mid1990s, during a similar revenueled fiscal decentralization. 50 In order to
prevent these scenarios the four pillars of fiscal decentralization described in the

47
Interview,UnionGovernmentOfficial(November2013).
Interview,UnionGovernmentOfficial(February2014).
49
Interview,UnionGovernmentOfficial(November2013).Ingeneral,Unionaidforministry
departmenthasbeenconsideredgrants,butforSEEshasbeenaccountedasloans.
50
48
15

introductionrequirealignment,andaclearerdivisionoffunctionsbetweentheUnion
andsubnationallevelswillbeneeded.
16
THREE:SketchingMyanmarsFiscalDecentralizationRoadmap
Decentralization reform strategies need to consider the full range of channels by
which resources reach subnational levels: A comprehensive analysis of the local
publicsectorshouldtakeintoaccountthattherearedifferentapproachesbywhich
governments interact withand deliver services tothe people. 51 The most
importantfeatureofthecontextandinstitutionsdescribedintheprevioussectionis
the mixture of deconcentration and devolution that is used to plan and deliver
resourcestothestate/regionandlocallevels.52Eachofthesechannelsisnotonly
about the organizations and capacities at the local level. Instead they describe the
relationshipsbetweenthecentralandsubnationallevels,aswellaslocalgovernance
andaccountabilitysystemsatthecenter.Anystrategytoaddressthesechannelsalso
includesbuildingthecapacityofthecentertobecomeintergovernmental.53
This multilevel approach is increasingly seen as essential to understanding and
reformingdecentralizingsystems:
Ratherthanisolatedactors,subnationalauthoritiesandcentralgovernments
aremutuallydependent.Inthiscontextthekeyunderlyingquestionis
notwhethertodecentraliseornotorevenoptforaspecific
decentralisationmodel,buttolookatwaystoimprovecapacityandco
ordinationamongpublicstakeholdersatdifferentlevelsofgovernmentto
increaseefficiency,equityandsustainabilityofpublicspending.54
AsuccessfulstrategyforfiscaldecentralizationinMyanmarshouldthereforestartby
workingonallthesechannels, andfocusonimprovingfiscal relationshipsbetween
thecenterandlocallevelsforeachofthem.Againstthisbackground,themainthrust
offiscaldecentralizationreforminMyanmaruptothispointtopdowneffortsto
increase the share of the budget appearing on state and region budgets is
inadequateintwoways.
First,itfocusesattentionmostlyononelevelofgovernance,thestateandregion,but
does not yet address the central systems for determining why the center should
devolveacertainshare ofthe budget,orhowthatshareshouldbedivided among
differentstatesandregions.Second,itonlyfocusesononeortwoofthechannels

51
LocalPublicSectorInitiative,MeasuringtheLocalPublicSector:AConceptualandMethodological
Framework(TheUrbanInstitute,December2012),1.
52
Forthepurposesofthispaper,andconsistentwithgeneralusageinthedecentralizationliterature,
state/regionlevelisusedtorefertoitself,localreferstotownshipandvillagelevelwhereservices
aremostdirectlydelivered,andsubnationalreferstoalltheselayersbelowtheUniongovernment.
53
54
ClaireCharbit,GovernanceofPublicPoliciesinDecentralisedContexts,OECDRegionalDevelopment
WorkingPapers(Paris:OrganisationforEconomicCooperationandDevelopment,July19,2011),5.
17

betweenthecenterandstate/regionlevelthechannelofdevolvedfundsinthestate
andregionbudgetfortheirlimitedfunctions,andtheprovisionofcapitalinvestment
throughthevariouslocaldevelopmentfunds(ontherightofFigure2).Giventhelarge
fiscalimportanceofthesocialandeconomicfunctionsstillunderUnionguidanceand
delivered via deconcentration (Channel 1), this reform misses an important
componentoftheLocalPublicSector:itdoesnotincludethefullrangeofchannels
bywhichresourcesandfunctionsreachlocallevels.55
Thissectiondevelopsinitialstepsforafiscaldecentralizationroadmapthatbuildson
the existing channels already described, and focuses attention on the relationships
betweenthecenterandthestate/regionlevelandtherequiredreformsandcapacity
ateachlevel.Theremainderofthesectionfollowsthestructureofthethreechannels
introducedinthepreviouspartofthepaper:
Channel1:thedeconcentrationofsectorspecificresources;
Channel2:thedevolvedstate/regionbudgetforassignedfunctions;
Channel3:provisionoflocalcapitaldevelopmentfunds.
Itdescribesthecurrentsituationineachchannel,andsuggestswaystoimprovetheir
functioning. In this sense, it is an incremental approach as suggested in the
introduction.Atthesametime,becauseoftheopenendednatureofthebigpicture
aspectsofMyanmarsreformenvironmentdescribedinthesecondpartofthispaper,
these approaches must not close off too many options for more substantial or
dramaticreformsinthefuture.
3.1 Channel1:Strengtheningsectoraldeconcentration
A vertical deconcentration process for important sectors in Channel 1 has already
beguninMyanmar.Whileitisbeyondthescopeofthispapertoexaminethesituation
in all sectors, from 2012 onwards ministries such as education and health have
increased the number and the responsibilities of staff at state/region, district, and
township level. Increased responsibility for implementing and managing both
recurrent and capital parts of the budget have also been deconcentrated in these
sectors. From the progress in sectors still centralized under Schedule One of the
constitution,suchashealthandeducation,conclusionsmaybedrawnabouthowto
strengthenthedeconcentrationframework.

55
Thetermlocalpublicsectorhasrecentlybeenadoptedbyoneinternationallytransferablewayto
takeaccountofthesedifferingsystemsaroundtheworldelaboratedbytheLocalPublicSector
InitiativeattheUrbanInstitute:LocalPublicSectorInitiative,MeasuringtheLocalPublicSector:A
ConceptualandMethodologicalFramework.
18

In education, district level offices headed by a deputy director and ten staff were
established, and the grades and numbers of sanctioned state/region and township
officers were increased. Township education officers wereelevated from thesame
rankasheadmastertoassistantdirector.However,progressinfillingthesepostshas
beenslower thantheircreation.In2013oneregionseducationofficeestimatedit
wouldtakeuntil2015tofillalloftheirsanctionedposts.56Alongwiththesestaffing
changes, some human resource management functions were deconcentrated
townshipofficerswouldassignortransferprimaryschoolstaff,andthestate/region
candothesameforsecondaryteachers.Allhiringandfiringremainedwiththecentral
level or the three Departments of Basic Education until 2013, when the hiring of
primary school teachers was deconcentrated to state/region education offices by
Presidentialorder.57
There has also been deconcentration of implementation of the capital budget for
education.Theprocurementandmanagementofschoolconstructionisnowwiththe
state/region education office, and they work through the state/region government
procurement board to select construction contractors, and involve the township
education department in supervision. 58 However, these budget execution
responsibilities do not correspond to decisionmaking power over the numbers or
types of school construction projects, or their locations. Bottomup input is sought
from township offices, and education officers reported that it has some impact on
decisions.Buttheyalsoreportedthatpoorsitingofprojectsandtheselectionofnew
construction over their recommendations for refurbishment of unsafe or
overcrowdedschoolswerestillweaknessesintranslatingbottomupinputintoreal
planningresults.59
Anequallyimportantissueisthatthisinputisnotreallyplanningorbudgetinginput
atallitisgatheredduringthefiscalyearandprojectsarethenapprovedforthat
sameyear.Forexample,atownshipsubmitsitsprioritiesforcapitalprojectsinJune
or July, receives funds for procurement around October, and must carry out the
constructionbytheyearendinMarch.Thisinputisnotshapingthebudgetingprocess,
only its execution, and some of the gains of having local input are surely lost by
reducingtheavailableimplementationperiod.
There are also other sources of capital spending in education, notably projects
supportedbythevariouslocaldevelopmentfundmechanismsdescribedlaterinthis
paper. These projects are difficult to coordinate with the Union ministrys sector

56
Interviews,StateandRegionEducationOfficers(January2013March2014).
58
Interviews,RegionEducationOfficer(January2014)andTownshipEducationOfficer(March2014).
59
Interview,TownshipEducationOfficer(March2014).
57
19

spending,sinceprojectselectiondoesnottakeplaceuntilwellintotheyear,sothere
islittlechancetocreateaconsolidatedlistofprojectspriortosubmittingeitherthe
ministryorthestate/regionbudgets.Thesefundsarewelcomed.Oneremotestates
education director and chief minister observed that the new availability of
constructionfundsmeantthatparentswhotheyestimatedhadbeenpayingforup
to 40 percent of the school construction no longer were asked to contribute. 60
However, they also noted that a local preference for refurbishment tended to be
overriddenbyUnionofficialsinfavourofnewschoolconstruction.Finally,thereisa
channelforschoolimprovementsthatisdecentralizedtoschoollevelviatheschool
grantsprogramsupportedbytheWorldBankandotherdonors.
Changesinthehealthsectorinstaffinglevelsandhumanresourcemanagementhave
been less dramatic, in part due to the different technical nature of health service
delivery.Forexample,intheHealthMinistrythereisnoroleyetfortownshipoffices
in staff assignment, transfer or promotion, other than reporting on vacant posts.61
Healthofficersatstate/regionandtownshiplevelpointtounfilledpostsasseeninthe
educationsector.62Themanagementofdoctorsandothertechnicalstaffisstillthe
Unionministrysresponsibility,andsomehealthdirectorshavepointedtotheneed
formoreconsultationontheseassignments,aswellasregulationsandincentivesfor
stafftoworkinremoteareas.63However,state/regionhealthboardsarenowgivena
role in deploying doctors, nurses and pharmacists, though decisions about the
numberstobehiredandthehiringitselfaredonebytheUnion.64
Significantcapitalandoperatingbudgetmanagementnowalsotakeplaceinhealthat
subnational levels. At the end of 2012 the procurement of medicine and materials
weredeconcentratedtothestate/regionlevelforclinics,withhospitalspurchasing
theirownmedicinesandmaterials.65However,aswitheducation,decisionsoverthe
budgetlevelsfortheseitemsarestillcentralized.Somespendingfromthedevolved
capitaldevelopmentfunds(Channel3)isalsodirectedtowardshealth,andthesame
issuesasineducationwerenoted:spendingisnotalwayscoordinatedwithministry
plansduetoalackoftimeinthebudgetpreparationcycle.Thischannelalsotendsto
favoursuggestionsfromtheUnionlevelfornewconstructionoverlocalpreferences
for renovations, and has not always used the most up to date building standards
determinedbytheministry.66

60
Interviews,StateEducationOfficerandChiefMinister(March2013).
Interview,TownshipHealthOfficer(March2014).
62
Interview,RegionHealthOfficer(January2013).
63
Accordingtooneofficial,rulesrequiringdoctorstoserveatimeinremoteplaceswereinplace
throughthe1980s:Interview,StateHealthDirector(March2013).
64
Interview,TownshipHealthOffice(March2014).
65
Interview,RegionSocialAffairsMinister(January2013).
66
Interview,StateHealthDirector(March2013).
61
20

Thesestepsrepresentimportantadvancesindecentralizingservicedelivery.However,
they do not yet form a systematic and transparent framework for allocating these
resourcesacrossthecountry.Thefirstphaseofanincrementalfiscaldecentralization
strategyshouldbuilduponthembyintroducingamoresystematicapproachtothe
deconcentration of functions, providing the resources to fund them (vertical fiscal
balance), and distributing these resources across the country in line with policy
objectives(horizontalfiscalbalance).
Introduceasystematicframeworkfordeconcentratingfunctionalassignments
The steps to deconcentrate service delivery so far have been taken by individual
ministerstorespondtotheobjectiveofpeoplecentereddevelopment,orinresponse
to specific instructions from the Presidents Office. In August 2013 the President
directededucationauthoritiestobeginappointingprimarylevelstaffatstate/region
levelwiththeparticipationofthosegovernments,andtoldtheMinistryofHealthit
must assign nurses and midwives in coordination with state and region
governments. 67 Other ministers in the Union government show varying interest in
makingthesekindsofchanges.Thisechoesthesituationdescribed,forexample,in
Cambodia:
inimportantsectorsforlocaldevelopment,suchashealth,educationand
naturalresources,mostdecisionsarestillmadeatthecentralleveldespite
ministrieshavingofficesandstaffatprovincialanddistrictlevelsandthereis
oftenalackofdisclosureofthefulldetailofthecostofimplementing
differentfunctions.68
Thiskindofpiecemeal,topdowndeconcentrationmaybeusefultogettheprocess
started,butshouldtransitionintoarationalandsystematicframeworkfordeciding
whatfunctionsareassignedtowhatadministrativelevels.Asthisisdone,appropriate
costingmethodscanbeusedtodecidetheappropriatelevelofoverallspendingfor
that level (vertical fiscal balance). There are two general steps in systematically
assigningfunctionalassignmentsforanysector.
Thefirstistounbundletheservice(usuallyasubsector)intoitscomponentfunctions
andactivities.Thismeansoutliningboththegeneralfunctionsrequiredtodeliverthe
service, and the activities that are needed to fulfill those functions. The broadest
functions that can be separated are typically policy (setting rules and standards),
provision(ensuringthroughfinanceandaccountabilitythattheserviceisdelivered),

67
68
FactorsAffectingFiscalDecentralizationinCambodia(ParliamentaryInstituteofCambodia,
November2012),2.
21

and production (actual delivery). These need to be further broken down and the
constituentactivitieslistedindetail,asintheexamplesforeducationgiveninTable1.
Table1:Unbundlingfunctionsinbasiceducation69
Function
Policy,Standards
Planning
AssetCreation
Operation
Monitoring&
Evaluation
ExamplesofActivities
syllabus
language
teacherstudentratio
catchmentsize
schoollocationandnumbers
roomstandards
teacherrequirements
constructionofschoolsandotherfacilities
teaching
maintenanceofschools
selecting/purchasingteachingmaterialsandbooks
hiring,training,anddeployingteachers
examinationofstudentsandteachers
internalaudit
Thenextstepistoconsidertheassignmentofthesefunctionsandactivitiesinterms
ofappropriateprinciplesoffiscaldecentralizationandaccountability,asdescribedin
Box 3, and propose adjustments. Some of these principles will tend to suggest
functionalassignmentsthataremorecentralized,whileothersmightencouragemore
decentralization.Itisimportanttonotethattheseprincipleslargelyapplyinthesame
ways whether decentralization is through deconcentration to lower administrative
offices,ordevolutiontosubnationalgovernments.Theseprinciplescanthereforealso
beusedtoconsideradditionalfunctionsfordevolutiontosubnationalgovernments
laterinareformprocess.
Box3:Principlesoffiscaldecentralizationandaccountabilityforfunctionalassignments70
Subsidiarityanddifferencesinpreferencesandconditions:afunctionshouldbecarriedout
atthelowestlevelitcanbeperformedeffectively.Thismaydependoncapacity,beneficiary

69
AdaptedfromLantPritchettandVaradPande,MakingPrimaryEducationWorkforIndiasRural
Poor:AProposalforEffectiveDecentralization,SocialDevelopmentPapers(Washington,D.C.:World
Bank,June2006),2429.
70
AdaptedfromPritchettandPande,MakingPrimaryEducationWorkforIndiasRuralPoor:A
ProposalforEffectiveDecentralization,3040.
22

population,andlandarea.Italsodependsonhowvariedlocalpreferencesareforthefunction
(heterogeneityofdemand).Thisprincipletendstosuggestmoredecentralizedfunctions.
Economyofscale:afunctionshouldbecarriedoutwhereitsunitcostislower.Productionof
some functions may be cheaper at higher levels. This principle, unlike subsidiarity, may
suggestlessdecentralization,andifso,abalanceneedstobefoundbetweenthem.
Externality/spillover:afunctionshouldbeatthelevelthatminimizesbenefitsthatspillover
tootherjurisdictionswithoutcompensation.Thisdependsonthetypeofcatchmentarea.
Equity/equalization: a function should be carried out at a level that ensures equitable
treatment.Thisisveryimportantformanysocialservicessuchashealthandbasiceducation
wherethegoalmaytendtobeuniversalityandequalityofaccess.
Localdiscretion:Whendecisionscanbemadelocallywithouthigherinvolvement,thenlocal
accountabilityisimportant.
Intensity of transactions: Activities with many repeated transactions/actions should have
localaccountabilitybecausetheyareinefficienttomonitorfromadistance.
Requires technical or local monitoring: Some activities require more technical verification
and checks due to the complexity of the service, and may need more centralized
accountability.
Intheend,theprecisemixoffunctionalassignmentsisamatteroflocalcontextand
history, and not a one-size-fits-all process. However, it requires a framework and
guidelines set at the national level to guide the process, and technical support in
implementing it.71 It is also important that the framework to govern this process
should have clear time-bound steps and should initially target a few of the most
importantfunctionsotherwiseitcanbecomecaughtupinanextendedandoverly
detailed process that can take years to implement. For example, in Cambodia
functionalmappingbeganin2010andreassignmentisstillongoing.72Mostimportant
isthattheassignmentsforthemainfunctionsandactivitiesthathavelargebudget
implicationsareveryclear.
Aneffectivesystemoffiscaldecentralization,whetherdeconcentratedordevolved,
ensuresthatbudgetsaresufficienttomeettheseclearlyassignedfunctions.Initially
budget levels may be determined by highlevel policy goals or existing overall
spendingonagivenfunctionoractivity.Later,morecomplexactivitybasedcosting
modelsmightbeused.However,thiscostingprocessiscomplexandsectorspecific
and would require more detailed analysis anddiscussion. The eventualgoal is that
transparentbudgetsforclearlyassignedfunctionscanallowlocaldiscretioninhow

71
Forexample,inCambodiathenationaldecentralizationprogramrecognizesthatnational
coordinatinginstitutions,theMinistryofEconomyandFinance,andMinistryofPlanninghavean
importantroleinfinalizingpoliciesonfunctionalassignments,whichshouldnotjustbelefttosector
ministries:Niazi,DeconcentrationandDecentralizationReformsinCambodia,59.
72
FactorsAffectingFiscalDecentralizationinCambodia,10.
23

thingsgetdone,whilesupportingaccountability.Deconcentrationcanbeafirststep
intheprocessofreachingthatgoal.
Horizontalallocationandlocalinputtoplanningandbudgeting
Onceministriesdevelopaclear,sequencedplanfordeconcentrationoffunctions,and
normsforcosting,planningandbudgetingthosefunctions,theycanrationallyallocate
resourcesamongtheiradministrativeunits.Whilesomepartsoflineministrybudgets
areincreasinglybeingmanagedorexecutedbysubnationalofficesasdescribedabove,
the planning and budgeting process is not yet decentralized. Subnational ministry
departmentsdonotplanwithinaforecastbudgetceilinginadvanceofthestartofthe
year.Instead,budgetdecisionsarestillmadecentrallyandthesubnationalunitsare
onlyawareoftheirallocationsafterthebudgethasbeenpassed.
At the same time, in line with thenotion of peoplecentered development many
ministries have tried to gather more planninginformation and priorities from their
subnational units. This information is important, but it does not in itself constitute
decentralizationofplanningorbudgeting:
Underdecentralizedfinancingconditions,localplanningisnolongerthe
preparationofawishlistofprojectsthataremeanttobefedbottomupto
regionalnationalagenciesfortheirselection,financingandimplementation
aspartofsectorprogramstheyadminister.Instead,localresourceavailability
anddiscretionovertheiruseallowsamoremeaningfullocalplanning
processoneinwhichlocalchoicescanandmustbemadeunderclear
budgetconstraints.73
Whatisneededisabudgetprocessthatallowsthelineministrytoprovidebudget
estimates for deconcentrated functions and activities to, at least, its state/region
departments.Thisneedstohappenearlyenoughinthebudgetpreparationcyclefor
thelocaldepartmenttoplanitsactivitieswithinthoseconstraints,andtoconsultand
coordinatewiththestate/regiongovernmentandpotentiallythepublicasitprepares
its own budget (see next section). An example of how this could work for basic
educationcapitalschoolconstructionisgiveninBox4.
Box4:Deconcentratingcapitalplanningandbudgetinginbasiceducation
1. The Union ministry, based on policy goals and using analysis of data on schoolaged
children,currentenrolmentandschools,assignsatransparentlyformulabasedbudget
ceilingforschoolconstructiontoeachstateandregioninadvanceoftheunionbudget
submission.
2. State/regioneducationofficescompleteaconsultativeprocesswiththeirlocalunits,the
state/regiongovernmentandthepublic,andselectthecombinationofrepair,renovation

73
LeonardoRomeo,TakingRisks:DecentralizedDevelopmentPlanning,IssuesandEarlyLessonsfrom
UNCDFSupportedLocalDevelopmentFundProgrammes,(NewYork:UNCDF1999).
24

andnewconstructiontheywishtoundertake,usingnationalstandardsandguidelinesfor
constructionandcosts.
3. Asthestate/regioneducationdepartmentalreadyimplementsthisbudgetthroughthe
localbranchoftheMyanmarEconomicBank,andprocurementtakesplacelocally,the
worksshouldbeabletostartvirtuallyassoonasthebudgetispassedinApril,ratherthan
sometimeinOctober.
4. Thestate/regiongovernment,byknowinginadvancefromtheirownbudgetsubmission
whatworkwillbeundertakenbytheDepartmentofEducation,canplantosupplement
theseorpursueotherprioritieswiththedevolvedcapitalresourcesavailabletothem.
This approach has two very important advantages. First, it allows for a more
meaningful role for state/region governments in planning and
implementation/monitoringofsectorserviceseventhoughthesearenotdevolvedto
state/regioncontrol. State and regiongovernment ministers andhluttaw members
continuetocomplainthattheirroleininfluencingthesesectorservicesisweak.This
isinpartbecausetheirinputintoplanningthingslikeschoolconstructiondoesnot
feeddirectlyintoalocalbudgetpreparationprocess,butmustbefiltereduptoUnion
levellater.74
Second,itprovidesamuchmoreeffectivewaytoconnectthebottomupinformation
beinggatheredbyministriesthemselveswiththetopdownbudgetingprocess.When
townshipofficerscreateprioritizedlistsofactivitiesfortheirparentministries,these
are done without knowledge of the available resources and only approved after
frequent and inefficient processes of adjustment and negotiation at higher levels
during the fiscal year. For example, with support from UNICEF, the Ministry of
Education is developing Township Education Improvement Plans in targeted
townshipsinlinewithlocalneedsandconditions,butthesearenotpreparedinthe
contextofabudgetceilingorconstraint,andthereforewillinevitablyrequirereview
andadjustment.75Aclearsubnationalbudgetingsystemwouldhelpthisbottomup
planninginformationtobetterfinditswayintothefinalbudget.
Thereareimportantprerequisitesinthebudgetformulationprocessforthisapproach
tobesuccessful.TheUnionbudgetpreparationcalendarandproceduresshouldbe
adjustedtoallowceilingstobeprovidedtostate/regionsectordepartmentsfortheir
deconcentrated functions near thebeginning of the budget cycle. Stateand region
officials would be provided a supplementary budget circular with guidelines on
consultationwithstate/regiongovernmentontheuseofthefunds,andthat outlines
limitations and restrictions on the use of these resources. Crucially, the budget
classificationsystem must be adjusted toallow forstate/region departments to be
designatedasbudgetunitswithintheUnionbudgetfortheselineministries,toallow

74
75
Interviews,RegionHluttawMembers,RegionEducationDirector(January2014).
Presentation,UNICEFstaff(February2014).
25

forthetransparentallocationoffundsbystate/regionaswellasbyfunction.76Since
theseimportantPublicFinancialManagementchangeswouldrequirejointactionby,
at a minimum, the Ministries of Finance, National Planning and Economic
Development,andthelineministries,thereisapressingneedforpolicycoordination
onfiscaldecentralization,asdiscussedinthefinalpartofthispaper.
A final important consideration about deconcentration reforms for the Myanmar
context is that successful deconcentration can help prepare the ground for more
democratically decentralized public services in the future, whether through
broadeneddevolution or even more federal arrangements. The new local planning
andbudgetingresponsibilitiesachievedthroughdeconcentrationcanbeanimportant
sourceofcapacity:
Inacountrywithalongtraditionofcentralization,thatlocalgovernments
lackthecapacitytobeselfgoverningisanobservationthatislikelytobe
bothtrueandtautology.Aslocalgovernmentsinseveralnewlydeveloping
countrieshavedemonstrated,developingcapacitytogovernisalearningby
doingprocess.77
At the Union level, introducing the planning tools to effectively cost and equitably
distributeresourcesarealsoimportantprecursorstothekindsoftransfersandblock
grantsthatwouldbeneededtosupportasystemwheremoreservicesaredevolved
to state/region and local governments. This kind of preparation for planning and
budgetingthroughdeconcentrateddeliveryisseenascontributingtothesuccessof
Indonesiaswelldocumenteddecentralizationreforms:
Asuccessfuldeconcentrationofcentralgovernmentservicedeliverymay
provideanimportantplatformforimplementingdecentralizationsuccessfully
Whenthedecentralizationpolicyiseventuallycrafted,thelocalcivilservice
willbemorereadytotakeontheresponsibilitiesassignedtothesubnational
level,andtherisksofextremedisruptionofservicedeliverywillbeless.This
appearstobeamajorreasonwhyIndonesiasbigbangdecentralization
wasaccomplishedwithoutabreakdowninpublicservicedelivery.78
Given the uncertainty over Myanmars longerterm decentralization process, this
deconcentrationfirststrategyhastheadvantageofgeneratingrapidimprovements
in the targeting of budgets and creation of local capacity within the existing
subnationalfiscalframework,whilelayingthegroundforpossiblefuturechangesto
theintergovernmentalfiscalsystem.

76
ElementsofthisprocessdrawonrecommendationsinHamishNixonandJamieBoex,
ImplementingGIROAsSubnationalPlanningandFinancePolicy:OptionsandInternationalPractice
(WorldBank:PolicyNotePreparedforGovernmentofAfghanistan,2010).
77
78
26

3.2 Channel2:Rationalizingstateandregionbudgets
SubnationalbudgetsinMyanmarhavethreebroadsourcesofrevenue,asoutlinedin
thepreviouspartofthispaper:aidfromtheUniongovernmenttothevariousstate
and region departments; several mechanisms for delivering discretionary funds for
local development and infrastructure initiatives; and own source revenues from a
rangeoftaxesandfeescollectedthroughseveraldepartmentsassetoutinSchedule
Five(Annex3).
Thesizeoftheaidflowstostateandregiongovernmentsarecurrentlybasedonpast
patterns, the need to finance the deficit between departments revenue and
expenditure,and/oradhocnegotiationswiththecentrallevel.Broadlyspeakingthe
Unionbudgetpreparationprocessbeginswithboththeministriesandstates/regions
submitting their proposals, including estimates of revenue and expenditure for all
their constituent departments and SEEs. The office of the second VicePresident
assessesthese,primarilyagainsttheaggregatefivepercentdeficittargetwhichisthe
primaryfiscalguidanceusedinbudgetpreparation,aswellasthemainprioritiesof
thegovernment.79ThesearesubmittedtotheFinancialCommissionmandatedinthe
ConstitutiontobeaggregatedandsubmittedtothePyidaungsuHluttawforapproval
as the national budget. State and region budgets are therefore largely a control
mechanismonrevenueandexpenditure,notanexpressionofstate/regionlevelpolicy
orplanninggoalsinfinancialterms.
Thismethodofbudgetpreparationhastwoimportantweaknesses.First,ittendsto
resultinbudgetsthatmirrorwhatcurrentratesofspendingareamongdifferentstates
andregions,notwhattheyshouldbetocarryouttheirassignedresponsibilities.Ifa
givenstateorregionhashistoricallyspentmorethanitsshare,itwilltendtocontinue
to receive higher allocations. Places that have had limited capacity to spend may
receiveless,eveniftheirneedsareasgreatorgreater.Ifbudgetsgrowincrementally
(i.e.byacertainpercentageeachyear),thegapbetweenthosereceivingthemostand
theleastwillgrowovertime.Thissituationalsocreatesaperverseincentiveamong
statesandregionstobiduptheirbudgetproposals:byaskingformoreandshowing
larger deficits they hope to end up with higher allocations from the Union. As one
regionalbudgetdirectorobserved:Somestatesandregionssubmittheirneedsonly,
whileotherssubmitabroaderlistofwantsinordertosecuremoreresources.80This
incentive may also contribute to the generally large supplementary budgets in
Myanmar,bothatcentralorstate/regionlevel.81Inturn,thiscanhaveimplications
forthebudgetdeficitasawhole.

79
Interview,BudgetDepartment,MinistryofFinanceandRevenue(November2013).
Interview,RegionBudgetDirector(March2014).
81
WorldBank,RepublicoftheUnionofMyanmarPublicFinancialManagementPerformanceReport,
PublicExpenditureandFinancialAccountabilityReport(Yangon:WorldBank,May2013),9,17,46.
80
27

Secondly,whenthedifferencebetweenstate/regionrevenuesandexpendituresare
the main determinant of the levels of transfers, states and regions and other local
governments do not have an incentive to improve their revenue collection. If local
departmentsshowmoreownsourcerevenueintheirbudgetsubmission,itislikely
thatthiswillresultinalowerleveloftransfer,oratleastalowerrateofincrease.
Under such circumstances, what are the benefits to state/region and township
departments to undertaking the politically unpopular business of taxation? The
resultingincentivestructureisforstatesandregionstoboostexpenditureswithout
considering revenue. In the words of one budget director: Departments want to
spend,andthinklessaboutrevenueorreducingexpenditure.82
Figures4and5illustratetheevolutionofthestate/regionbudgetsbetweenFY2012
13andFY201415.BudgetestimatesfortotalstateandregionexpendituresforFY
201415havenotbeenpublishedatthetimeofwriting,butintheprevioustwoyears
budgets havegenerally grown, with theexception of Yangon (which experienceda
hugesupplementalbudgetattheendof201213),andRakhine.However,figuresfor
Unionaidtostatesandregions,whichareavailableforthecurrentFY201415,show
adramaticincreasetoallstatesandregionsexceptYangonandMandalayoverthe
previoustwoyears.Thischangemeansthattheoverallstateandregionbudgetsgrew
quickly(asnotedintheintroduction),andthattherelativeshareofthebudgetfunded
byUnionaidwillalsogrow.Inotherwords,statesandregionsarereceivingbigger
budgets,butarealsomoredependentonUniontransfers.83

82
FiguresforUnionaidhaveincludedbothgrantsandloansfortworeasons:thereisanunclear
expectationofrepayment,andtherehavebeenchangesinpolicyoverwhichdepartmentsandSEEs
receivegrantsorloans.Itshouldbenotedthatthisiscontrarytotypicalpublicaccountingpractice,in
whichloanswouldnotbeconsideredatransfer.
83
28

Figure4:State/regionexpenditureFY2012/13and2013/1484
300000
250000
200000
150000
100000
50000
0
TotalExpenditureFY201213
TotalExpenditure201314
Source:FY201213andFY201314BudgetEstimatesandRevisedEstimatesaspublishedinGazette
Figure5:Unionaidtostates/regionsFY201213toFY201314
250000
200000
150000
100000
50000
0
UnionAidFY201213
UnionAidFY201314
UnionAid,FY201415
Source:FY201213toFY201415BudgetEstimatesaspublishedinGazette
3.3 Channel3:Allocatingsubnationaldevelopmentfunds
Since 2011 Myanmar has introduced a number of overlapping committees and
subnational development funds broadly aimed at poverty alleviation and local
developmentthroughincreaseddiscretionaryresourcesandbottomupplanning.

84
In201213Yangonhasaparticularlylargesupplementary(RevisedEstimate)whichdoubledits
expenditures,contributingtotheapparentanomalyofadecliningbudgetfrom201214.
29

This process began in 2011 with a list of eight priorities set by the President in an
addressmadetothePovertyAlleviationCentralCommittee.85Tosupportthesepolicy
priorities, committees were formed starting from the Union level through working
committeesdowntovillagetractlevel.However,thesecommitteeshavehadlimited
roles in programming subsequent funds that have been directed to subnational
development.
Sincethen,therehasbeenaremarkableproliferationofcommitteesatthetownship
level to manage development planning and implementation. 86 The basic
administrativeorganistheTownshipManagementCommittee(TMC),headedbythe
GAD Township Administrator, with the other members being the officers of
government departments present locally. The Township Development Affairs
Committee (TDAC) consists of a respected community member as chair, the GAD
deputy township administrator, the rural development, livestock and fisheries
departmentdeputy,thesecretaryofthedevelopmentaffairsoffice(formerlyknown
asMunicipalOrganization),andrepresentativesfromsocial,business,andacademic
sectors.87Finally,TownshipDevelopmentSupportCommittees(TDSC)wereformedto
contributetodevelopmentplanning,althoughwithoutdirectcontroloverresources.
ThesecomprisetheAssistantTownshipAdministratorfromGAD,thechairfromthe
Township Development Affairs Committee and six selected sector representatives
fromsocialsectors,businessassociations,laborers,andfarmers.88
ThesethreecommitteesandtheTownshipFarmlandManagementCommitteemeet
onceortwiceamonthtodiscussneedsacrosstheeightdevelopmentprioritiesset
outbythePresident.TheTMCisthedecisionmakerwhiletheTDSCtakesanadvisory
role between the TMC and the public. 89 There are also Ward/Village Tract
ManagementCommitteesandWard/VillageTractDevelopmentSupportCommittees,
thelatterwithfivetosevenelectedmembersdependingonthesizeofthewardor

85
Theeightprioritieswere(1)Developmentofagriculturalproductionsector;(2)Developmentof
LivestockBreedingandFishandMeatProduction;(3)Developmentofruralproductivityandcottage
industries;(4)Developmentofmicrosavingandcreditenterprises;(5)Developmentofrural
cooperativetasks;(6)Developmentofruralsocioeconomy;(7)Developmentofruralenergy;(8)
EnvironmentalConservation:ChairmanofRuralDevelopmentandPovertyAlleviationCentral
CommitteePresidentUTheinSeinaddressesthecentralcommitteemeeting(1/2011),20June2011,
[http://www.presidentoffice.gov.mm/en/?q=issues/ruralurban/id1845]
86
Forathoroughpresentationoftownshipanddistrictlevelcommitteesmembershipsandroles,see
KyiPyarChitSawandMatthewArnold,TheGeneralAdministrationDepartment:AnOverviewof
MyanmarsAdministrativeBackbone,Annex2.
87
FormorediscussionoftheTownshipDevelopmentAffairsOfficeseeJaredBissinger,Subnational
GovernmentsandBusinessinMyanmar(Yangon:TheAsiaFoundation/CentreforEconomicandSocial
Development,February2014),1315.
88
UnionofMyanmar,PresidentOfficeOrderNo.(27/2013),Art.4.
89
Interview,RegionTDSCChairman(March2014).
30

villagetract.90Theprocessforfillingthesepostshasbeenunclear,andinmanycases
existingofficialsmayeffectivelynominatethem.91
UnionGovernmentFunds:ThePovertyReductionFund
The principle subnational development fund provided by the Union Government is
known as the Poverty Reduction Fund (PRF). 92 The Poverty Reduction Fund was a
presidential initiative to address rural development and poverty, and began in FY
201213withonebillionkyatsallocatedtoeachstateandregion.ChinState,dueto
its extreme poverty rates and remote condition, was subsequently allocated an
additionaltwobillionkyatsthesameyear.Asthisfirstroundofthepovertyreduction
fund was distributed as an equal share for each state/region it resulted in widely
varyingentitlementsonapercapitaorpovertybasis.93Withinstatesandregions,the
one billion kyats were generally divided equally among the townships, resulting in
differing entitlements per township as well. This equal shares approach raised
concernsamongsomelocalrepresentatives:
Thedevelopmentofregionsandstatesareunequalwithdifferentneedsand
differentpovertyratesandsoallocatingpovertyreductionfundsequallyto
allstatesandregionswouldnotbethemostappropriatesolutiontotackle
thepoverty.94
The funds were both budgeted and executed under the General Administration
Department,becausestateandregiongovernmentinstitutionswerenewlyformed
and considered unable to manage these funds. 95 The second round of poverty
reductionfundsforFY201314alsoonebillionkyatperstateandregionwere
budgeted and transferred directly (via drawing rights) to the state/region
administrativeorganizationbudgetcategory,andnottheHomeMinistrysbudget
line.Thisisthenewlycreatedbudgetcategoryforthestateandregiongovernments
executiveinstitutions,andintheorycouldprovideformorefiscalautonomyoverthe
useofthefunds.However,executionofthestate/regionadministrativeorganizations
budgetisstilltheGADExecutiveSecretarysresponsibility,andthereforeunderthe
GADsinfluence.Thetownshipcommittees,particularlytheTownshipDevelopment
SupportCommittee,aremeanttoprovideinputtoprojectselectionandmonitoring.
Forthe201415budget,thefundnolongerprovidesanequalshareforall14regions
andstates.Theoverallbudgethasbeenincreaseddramaticallyto50billionKyat,and

90
UnionofMyanmar,PresidentOfficeOrderNo.(27/2013),Art.6.
91
Fieldresearchontownshipadministration,personalcommunication.
ThishasbeentranslatedinsomecasesasPovertyEradicationFund.
93
Nixonetal.,StateandRegionGovernmentsinMyanmar,45.
94
Interview,StateMinister(December2013).
92
95
Interview,UnionLineMinistryStaff(February2014).
31

someregionsandstatesarereceivingmuchhigheramountswhileothersremainat
onebillionkyats.Forexample,ShanStatereceivedfourbillionkyatsandRakhineand
Kachinreceivedfifteenbillionkyatseach,whiletheotherstatesandregions(seeTable
2below).Theselargeamountsareaimedatareaswithparticularlyhighreconstruction
needsduetoconflict,orremoteness.96
Theprocessforallocatingfundstoeachtownshipisalsonowdecidedbythestateand
regiongovernment.Somestatesandregionshavekepttoanequalsharesprinciple,
whileoneregiondescribedusingaclassificationoftownshipsintotwogroupsbased
onpovertyrateandpopulation.Inonestate,thequotaforeachtownshipisdecided
based on thesize oftheprojects,andsomeofthefundshavebeengiven assmall
loans.97The repeated changes to the procedures for distribution and use of these
fundshaveledtosomeconcernsoverthemonitoringandauditing:
Therehavebeenmanypolicychangesandtheauditorcouldnotkeepupwith
itbecausetherulesandregulationsusedbyauditorsareoutdated.Itisnot
updatedtokeepupwiththechangesinthebudgetsuchaspovertyreduction
funds.98
ThereareseveralotherdiscretionaryfundsfromtheUnionGovernmentdeliveredvia
differentchannels.BoththeDepartmentsofRuralDevelopmentandtheMinistryof
BorderAffairshavesomefundsavailableforuseatstate/regionlevelandbelow.In
addition,attownshipleveltheGADhasdiscretionaryuseofacommissionorrebate
on collected motor vehicle license fees in busier townships such as Myeik in
Tanintharyi,thiscanbeasubstantialamount.99
TheConstituencyDevelopmentFund
The Constituency Development Fund (CDF) aims to provide budget for local
development by allocating a fixed amount in each township for projects selected
under the guidance of the national and state/region hluttaw representatives.
PyidaungsuHluttawSpeakerThuraUShweMannintroducedtheconceptofaCDFin
April2013.ApparentlythisideawasinspiredduringavisittoIndiainDecember2011.
TheCDFwasrealizedbypromulgationoftheProceduresofcontrollingandusingof
Development Fund, Order No. (82/2013): by the Pyidaungsu Hluttaw on 15

96
Interview,StateFinanceDirector(May2014).Theexactdistributionwashardtodiscernatthetime
ofwritingbecausethisassistanceisaggregatedwithotheraidtothestatesandregionsintheUnion
budgetdocument,andtheindividualstateandregionbudgetestimateshadnotyetbeencollated.
97
Interview,StateMinister(May2014).
98
99
Interview,StateFinanceMinister(December2014).
Interview,TownshipAdministrator(March2014).
32

November2013,wherebyrepresentativeswerealsogiventherighttoopenjointbank
accountsforeachtownship.100
Theplancallsforthedistributionof33billionkyatsresultingfromcutstotheUnion
budget equally to all 330 townships, resulting in an allocation of 100 million per
township.101RepresentativesfromthetwohousesofthePyidaungsuandregion/state
hluttawsareallowedtoselecttownshipdevelopmentactivitiesintheirconstituencies
toamaximumoffivemillionkyatsperproject.102Prioritiesfortheseprojectsareto
bewatersupply,renovationsofruralroadsandbridges,renovationofschoolbuildings,
renovation of buildings related to health and other township needs. 103 The
implementation of CDF projects is to be done by Township Development
Implementation Bodies consisting of the four hluttaw representatives from the
township.
Thefund has beencontroversial.Initiallythe President refusedtosignthelawand
sent it back to the Pyidaungsu Hluttaw with comments that the law was
unconstitutional due to its granting the hluttaw executive power to implement
development activities. However, the hluttaw proceeded to promulgate Order No.
83/2013andNo.86/2013withminoramendments.Internationalexperiencesuggests
thatconstituencyfundsincludinginPakistan,India,andthePhilippinescanbreach
separationofpowersandweakenlegislativeoversight,weakengovernmentcapacity
to plan and monitor, undermine equity when the program allocates funds equally
across constituencies, and encourage patronage. 104 MPs themselves were divided,
andacademicsandinternationalorganizationsraisedconcerns.105Ontheotherhand,
localhluttawmembersindicatedthatthesefundswouldallowdevelopmentactivities
tobebetteralignedwithlocalneeds.106
InFebruary2014,MinisterforFinanceUWinSheinandUnionAuditorGeneralUThein
Htike explained the governments concerns over the program, and the Pyidaungsu

100
ProceduresofcontrollingandusingofDevelopmentFund,OrderNo.(82/2013),Pyidaungsu
Hluttaw,15thNovember2013,Art.7.
101
HluttawtoreleaseprocedurestospendK.33billiondevelopmentfund,MyanmarFreedom,(7
October2013).
102
Hluttaw,15thNovember2013,Art.8a.
103
Hluttaw,15thNovember2013,Art.7k.
104
AlbertvanZyl,WhatIsWrongwiththeConstituencyDevelopmentFunds?,BudgetBrief
(InternationalBudgetPartnership,2010).
105
ThomasKean,VotebuyingfearsoverMPfundsplan,MyanmarTimes,(15December2013),
[http://www.mmtimes.com/index.php/nationalnews/9057votebuyingfearsovermpfunds
plan.html].
106
Interview,RegionHluttawMembers(January2014).
33

Hluttawagreedtochangestobetteralignitwithgovernmentfinancialprocedures.107
TheCDFisnowbudgetedascurrentexpenditureandrevenueunderthestate/region
hluttawbudget.108
DonorProvidedFunds:CommunityDrivenDevelopment
In2013theWorldBankandtheMyanmargovernmentlaunchedthesixyearMyanmar
NationalCommunityDrivenDevelopmentproject(NCDD)withan80millionUSDgrant,
supplementedby6.3millionUSDofgovernmentsupport,overthelifeoftheproject.
ThedevelopmentobjectiveofNCDDistoenablepoorruralcommunitiestobenefit
fromimprovedaccesstoanduseofbasicinfrastructureandservicesthroughapeople
centeredapproach,andtoenhancethegovernmentscapacitytorespondpromptly
andeffectivelytocrisisoremergency.109Theprojectiscurrentlyplannedtoexpandto
15 poor and remote townships (i.e. one township per state/region plus the union
territory).Townshipselectiontakesplaceusingparticipatorymethodsstartingatthe
villagetract,townshipandstate/regionlevel,andthefirstphaseiscurrentlyunder
implementationinthreetownships.Eachvillagetractreceivesanaverageblockgrant
of27,000USDperyearforthreeyears,andthesegrantsareinthreebandsaccording
tovillagetractsize.
The institutional arrangements for NCDD have three levels: (1) the Union level, (2)
townshipleveland(3)villagetractandvillagelevel.110Atthevillagetractlevel,newly
formed Village Tract Project Support committees consist of the Village Tract
DevelopmentSupportCommitteeandVillageProjectSupportCommitteeleaders,and
are responsible for subproject selection. A finance subcommittee is formed at the
villagetractleveltomanagetheblockgrantsforvillagetractandvillagelevelalong
with other subcommittees for monitoring and procurement. The township level is
responsibleforcoordinatingwiththeRuralDevelopmentDepartment,andcertifying
the subprojects that are proposed by village and village tract meetings do not
duplicate existing government efforts. As in most community driven development
programs, this process is facilitated by partner NGOs, currently the International
RescueCommitteeandMercyCorps.

107
EiEiToeLwin,MPsagreetoamendconstituencyfundingprogram,MyanmarTimes,8February
2014:[http://www.mmtimes.com/index.php/nationalnews/9490mpsagreetoamend
constituencyfundingprogram.html].
108
109
MyanmarNationalCommunityDrivenDevelopmentProject,WorldBank
[http://www.worldbank.org/projects/P132500/myanmarnationalcommunitydrivendevelopment
project?lang=en]
110
NationalCommunityDrivenDevelopment,OperationalManuelPartIV,Governmentofthe
RepublicoftheUnionofMyanmar,2013,GovernmentoftheRepublicoftheUnionofMyanmar,3.
34

Thereareanumberofotherdonorprovidedlocalinfrastructureprogramsincludinga
proposed 22 million USD community block grant implemented by the Asian
Development Bank for several states and regions, and a large national local
infrastructure investment program of about 220 million USD support by Japan.111
However,thelatterprimarilydrawsonUnionministriestoselectandimplementits
79subprojects.TheUNDPUNCDFhasalsobeenexploringthepossibilityofatownship
leveldevelopmentfund.Asnotedabove,theDepartmentofRuralDevelopmentitself
also has a local development budget outside the NCDD framework, and is seeking
waystoincreasecommunityinvolvementintheirprogramming.
Implicationsandchallenges
The introduction of these important sources of funding for development initiatives
clearly provide new opportunities for local governments and administrations to
implement development activities that would have been impossible otherwise.
However, the different fund structures and modalities, and the proliferating
committee structures that have grown up to ensure local participation in their
targeting,presentchallengesoftheirown.Boththefundsandthecommitteeshave
overlapping objectives and mandates, which makes coordination difficult and risks
poortargetingofresources.
Figure6:Selectedlocalcapitaldevelopmentfundsandtheirbudgetingarrangements

111
Donorrepresentatives,personalcommunicationwithauthors.
35

Given the different funding and management procedures, duplication of effort in
somesectorsorareascouldoccur,whileothersgounderfunded.Atthesametime,
thecomplexandvariedallocationproceduresmeanthattheoveralleffectofthese
flowsonthehorizontaldistributionofresourcesacrossthecountrymaybedifficultto
discern.Somefundsaredistributednationwideandequallyshared,whileothersare
aimedatparticularareasorallocatedonamoreadhocortargetedbasis.Indeed,one
recentattempttodocumentfiscalspaceatthelevelofseveralsampletownships
found it virtually impossible to quantify systematically the various resources
available. 112 As a result, while individual funds may succeed relative to their own
objectives,theimpactonthesubnationalfiscalsystemasawholeandparticularly
horizontalfiscalbalanceisveryunclear.Inshort,atatimewhenMyanmarshould
be aiming for greater budget integration, there is actually considerable budget
fragmentationoccurring.Thereisaneedtoconsiderwaystobetterintegrateboththe
ordinarystateandregionbudgetsandthevarioussourcesofdevelopmentfunds.The
nextsubsectionintroducespotentialelementsofsuchatransfersystem.
Table2:Devolvedgovernmentlocaldevelopmentfundsbystate/region,FY201415
State/Region
#of
Township
s
Estimated
Population
*
Ayeyarwadd
y
Bago
Chin
Kachin
26
7,179,000
28
9
18
5,428,103
514,994
1,510,618
Kayah
Kayin
Magway
Mandalay
Mon
Rakhine
7
7
25
31
10
17
316,714
1,623,989
4,858,203
7,090,579
2,626,104
3,264,231
Sagaing
Shan
Tanintharyi
Yangon
37
55
10
45
5,836,529
5,076,654
1,539,734
6,492,478
Total
Per
Poverty Constituenc
yFund
(Millions Capit
Reductio
(Millions)
)
a
nFund
(Kyat)
(Millions)
1,000.00
2,600.00
3,600
501
1,000.00
5,000.00
15,000.00
1,000.00
2,000.00
1,000.00
1,000.00
1,000.00
15,000.00
1,000.00
4,000.00
1,000.00
1,000.00
2,800.00
900.00
1,800.00
3,800
700
5,900 11456
16,800 11121
700.00
700.00
2,500.00
3,100.00
1,000.00
1,700.00
1,700
2,700
3,500
4,100
2,000
16,700
5368
1663
720
578
762
5116
3,700.00
5,500.00
1,000.00
4,500.00
4,700
9,500
2,000
5,500
805
1871
1299
847
Sources:MyanmarInformationManagementUnitTownshipLists,InterviewswithMoFstaff,and
ConstituencyFundlegislation
*Populationestimatesareaveragesof2011MIMUpopulationestimatesand2011Myanmar
StatisticalYearbookandareintendedtobeindicativeonly.

112
UNCDF,LocalFiscalSpaceinMyanmar,unpublisheddraft(April2014).
36

3.4 Towardsastate/regiontransfersystem
Both the aid from the Union government for state/region departments and the
varioussubnationaldevelopmentfundsdescribedaboveareeffectivelyfiscaltransfers,
sincesubstantiallocaldiscretionovertheiruseexistsatstate/regionlevelsorbelow.
Alsoasdescribed,boththeUnionaidtothestate/regionbudgetandthedevelopment
funds have weaknesses in their allocation systems: the former through its deficit
based budgeting process, and the latter through their collectively overlapping
objectivesandmandates.
Alongside the process of sectoral deconcentration already described, Myanmar
shouldalsoconsidermovingtowardsasystemoftransfersthatworkstomeetthe
objectivesofthegovernmentforthosedevolvedactivitiesdefinedinScheduleTwo.
TheimmediateprioritiesinMyanmararetostarttomovefromthecurrentsystemof
budgetingforstate/regiontransferstoonethatrationalizes(1)theverticalbalanceof
resourcesbetweencentralandsubnationallevels;(2)thehorizontalbalanceacross
thecountry;and(3)theincentivesforownsourcerevenuecollection,while(4)not
overwhelming fiscal or implementation capacity in weak or small states.
Strengthening revenue capacity is essential for subnational governments to avoid
being a spending agent for the center, and to develop more permanent balance
between revenues with expenditure assignments, discretion, autonomy and
accountability. 113 An additional objective of the system may be to enhance the
coordinationoftheordinaryanddevelopment/capitalsidesofthebudget.
The most appropriate way to accomplish these objectives is to carefully develop a
formulabasedsystemfortransfersforsome,ifnotall,stateandregionlevelresources
anddevelopmentfunds.InCambodia,comparisonofthehorizontalimbalanceamong
communes,whichreceiveaformulabasedallocation,andprovinces,whichundertake
a budget process similar to that in Myanmar, confirms that basic formuladriven
approachesthattakesomeaccountofcapacitiesandneedswillleadtomoreefficient
and fairer outcomes than will bargaining between subnational administrations and
officialsofthecentralgovernment.114
Thetransparencycreatedbythisprocesscanbepoliticallyusefulaswell.InTanzania,
onereasonfortheintroductionofaformulabasedtransfersystemwastoclarifyto
members of parliament the criteria for distribution of revenues among local
governments.115Aformulabasedbudgetforsubnationalallocationsetsahardbudget
constraint at the beginning of the fiscal year, and can be used to limit local
governments from turning to large requests for endofyear assistance from the

113
Niazi,DeconcentrationandDecentralizationReformsinCambodia,51.
115
114
37

central government which undermine the predictability of the national budget.
Withoutsuchaconstraint,localgovernmentsmayinducebailoutsordefaultssuchas
BrazilandArgentinaexperiencedinthemid1990s.116Transferscanthereforesupport
fiscalcontrolssuchasthegovernmentsgoalofmaintainingGDPbaseddeficittargets.
Itmayalsohelpsupportalignmentwithfuturepublicfinancialmanagementreforms
suchastheintroductionofasingletreasuryaccountormoreresultsbasedbudgeting
models.
Thedevelopmentofsuchatransfersysteminvolvesseveralsteps.Thefirstandmost
importantistoconsidercarefullytheobjectivesofthesystemasawhole,andofthe
individualtransfersthataretomakeitup.Someofthemostcommonobjectivesof
transferspotentiallyrelevanttoMyanmarareasfollows:
Improving vertical fiscal balance (ensuring sufficient resources for

subnationalresponsibilities);
Improvinghorizontalfiscalbalance(equalization);
Compensatinglocalitiesforspilloversofbenefitstoneighboringareas;
Pursuing national priorities and services with some locally provided
component,suchashealthandeducation;
Sharingrevenuesfromspecificsourcessuchasnaturalresourcesaccording
topoliticalagreements;
Encouraging certain kinds of local government performance, for example
revenuemobilizationorgoodfinancialmanagement.
Eachoftheseobjectivesmaybebestservedbydifferenttransferdesigns.Onekey
element of transfer design is whether the transfer is conditional or unconditional.
Conditional transfers are earmarked for specific purposes according to sector or
expenditureclass.Unconditionaltransferscanbeusedatthediscretionofthelocal
government and better serve purposes such as compensating for horizontal
imbalancesorinequalities,promotinglocalautonomyorrevenuesharingfromnatural
resources. In addition, transfers are increasingly being designed that incorporate a
performancebasedcomponent,bywhichtheamounttransferredvariesaccordingto
localgovernmentsperformanceinagivenareaforexamplefinancialmanagement,
revenuemobilizationordevelopmentoutcomes.117
The best starting point for designing a state/region transfer system for Myanmar
wouldbethoseresponsibilitiesalreadyunderthecontrolofstate/regiongovernment,
orthedevelopmentaimsofthefundsdescribedintheprevioussubsection.Certainly
thereisaneedformoreclarityonthescopeofsubnationalresponsibilities,andthis
canbedevelopedthroughasimilarprocessoffunctionalanalysistothatdescribedin
the section on sectoral deconcentration above. The current system does not

116
Ibid.,24.
ForathoroughexplanationofperformancebasedgrantdesignprinciplesseeUNCDF,
PerformanceBasedGrantSystems:ConceptsandInternationalExperience,(NewYork:UNCDF,n.d.).
117
38

effectively address some key objectives, particularly horizontal equity or wealth
sharing,butMyanmaralreadyhasasolidsetofobjectivesuponwhichtodevelopits
system of intergovernmental finance in a more rational direction without starting
afresh. To guide consideration of transfer designs, some common subnational
expendituresandappropriatefundingsourcesarepresentedinTable3.
Table3:Matchingexpenditureresponsibilitieswithappropriatefinancing118
Expenditure
Locallydeterminedservicesor
infrastructurethathasidentifiable
usergroup(excludable)andno
spilloversorequityimplications
Locallydeterminedservicesand
infrastructurethathavesomelocal
spillovers
Localservicesthathavespilloversto
otherjurisdictions
Localprovisionofnationally
determinedpriorityprograms
Localspendingonnationally
determinedprogramsforpromoting
equity(nationalspillover)
Localcapital/infrastructure
developmentwithlongterm
benefits
Examples
Localutilities,
recreationalfacilities
AppropriateFinancing
Userfees
Localparks,minor
roads,refuse
collection
Districtroads,
diseasecontrol
Localtaxesandtransfers
(unconditional)
Localtaxesandtransfers(conditional
orunconditional)
Publichealthcare
andeducation
Central transfers(conditional)
Socialwelfare,public
education,resource
sharing
Centraltransfers(conditional,
unconditional,orrevenuesharing)
Highway,transport,
telecommunications
Localborrowing,transfers
(conditionalorunconditional)and
revenuefordebtservicing
Ingeneral,transfersystemdesignshouldconsiderobjectivesinthefollowingorder.
The first is to achieve vertical fiscal balance by bridging the difference between
expenditure responsibilities and revenue raising powers. The second is to ensure
provision of national or regional functions where underprovision might take place
without assistance. Finally, equalization must address horizontal imbalances that
remainafterthefirsttwopartsofthetransfersystemaredesigned.119
Transfersystemsfallintofourbroadtypesbasedonwhethertransfersareconditional
or unconditional and their focus on recurrent or development spending. These are
presentedinTable4inorderofincreasingcomplexityanddecreasingdecentralization
from left to right. The more conditionality, the more complex the system and the
lower the degree of autonomy or decentralization granted to lower administrative
levels.Simplerand more unconditionalsystemssupportmorelocalautonomy, and

118
119
AdaptedfromSteffensen,FiscalDecentralizationandSectorFundingPrinciplesandPractices.
BahlandMartinezVazquez,SequencingFiscalDecentralization.
39

might be more conducive to further decentralization in the future, an important
considerationinlightoftheuncertaintyoverMyanmarslongtermdecentralization
path.
Table4:Basicgrantoptionsforsubnationalfiscaltransfersystems
BudgetElement
RecurrentBudget
Single
Unconditional
BlockGrant
Development
Budget
TransferSystemElements
Conditional
Unconditional
SectorGrants
BlockGrant
Unconditional
Unconditional
BlockGrant
BlockGrant
Lesscomplex /Moredecentralized
Conditional
SectorGrants
Conditional
SectorGrants
Morecomplex/Lessdecentralized
With the overall system in mind, the design of individual transfers can take place.
Therearefourstepstothisprocess:
1. Clearlyandsimplydefinetheobjectivesofthetransfer;
2. Select the appropriate transfer mechanisms (conditional, unconditional,
matching,performancebased);
3. Determine the basis for determining the transfer pool (the total amount
availableaccordingtoadefinedshareofrevenues,reimbursementofcosts,
estimationofneeds).
4. Determinehowthepoolisdividedamongthesubnationalunits(theformula).
EachofthesestepsisexplainedinmoredetailinAnnex2,butitshouldbekeptinmind
thattransferdesignisacomplexactivityandrequiresadequatetechnicalsupport.It
alsocausestransitionissues:
often,aparticularlydifficultissueariseswhenthesystemofdistributing
intergovernmentaltransfersamonglocalgovernmentsisdramatically
changed,asforexamplefromanegotiatedsystemtoaformulabased
system.Thischangewillalmostcertainlyproducelocalgovernmentswho
makebiggainsfromthenewformulatransfers,andotherswhofacebig
revenuelosses.120
Fortunately,oneadvantageofintroducingformulabasedtransfersisthattheirimpact
onbudgetoutcomescanbesimulatedbeforeanyimplementationneedstotakeplace.

120
Ibid.,31.
40

Appropriatemeasurescanthenbedesignedtodealwithtransitionalissuessuchas
thereallocationofexistingresourcesbetweenterritories.
In the case of Myanmars transfer system to states and regions, some significant
decisions are (1) whether to treat the ordinary state and region budget on a
conditional basis or unconditional basis; and (2) how to integrate the various
development funds with each other and with the recurrent budget; (3) how to
encourageownsourcerevenuegeneration;and(4)howtoencouragecoordination
withtheseparatechannelofdeconcentratedsectoralservices.
Atentativeproposalforstateandregiontransfersmightthereforeinclude:
an unconditional block grant for those ordinary functions assigned to

state/region level, possibly with a matching or small performancebased
componentaimedatimprovingplanningandrevenueperformancerelative
tolocalcapacityandconflictsensitivities;and
anadditionalunconditionalgrantaggregatingthePovertyReductionFund
and,ifpoliticallyfeasible,theConstituencyDevelopmentFund,withrules
forcoordinationwithdeconcentratedsectorservices;
a possible general purpose grant for equalization, compensation or
derivationbasedonnaturalresourcerevenues(seeBox5).
Box5:Naturalresourcewealth:thefourthchannel?121
Myanmars government has announced a commitment to both greater transparency in
naturalresourcesectorsandtofurtherfiscaldecentralization.Therehasalsobeenincreasing
discussion of sharing natural resource revenues in the contexts of economic restructuring,
constitutionalreform,andthepeaceprocess.However,thereisnotyetconsensusonwhatis
tobeshared,betweenwhichentities,how,orwhy.
Thegovernmenthasestablishedaworking grouptoimprovepublic financial management
andMyanmarhassubmittedacandidacyrequesttotheExtractiveIndustriesTransparency
Initiative (EITI)Significant sources of uncertainty include: the composition, extent and
distributionoftheresourceendowment;informalextractionandpayments;contestedareas
andparalleladministrationbynonstategroups;theroleofmilitaryownedcompanies;anda
lackofroutine,disaggregatedreportingforformalrevenuesfromtheextractivesector.
InMyanmar,importantpartsofthearchitectureneededforsuchafiscalsystemare
alreadymovingintoplace.Whilebudgetproposalsarecurrentlycompiledbasedon
individualdepartmentsrevenueandexpenditure,thereisnoserioustechnicalbarrier
ofwhichweareawaretopreventthestateandregiontakingfulldiscretionoverthe
allocationoffundswithinitsassignedresponsibilities.Themostcomplexchangeswill,
liketheproposeddeconcentrationreforms,involvealterationstothebudgetcycleto

121
ExcerptfromThetAungLynnandMariOye,NaturalResourcesandSubnationalGovernmentsin
Myanmar,ii.
41

allowmoreforbudgetceilingstobeissuedtostatesandregions,andforstate/region
governments to review and modify the budget.122 These reforms are complex and
requirecoordinatedeffortsatbothcentralandsubnationallevels,andacrossdifferent
partsofgovernment.Thefinalpartofthepaperemphasizesthatthecountrywillneed
strongerandmoreformalizedcoordinatinginstitutionsandreformprocessesifthese
smaller pieces are to later converge and support a coherent longterm vision for
effectivecentralsubnationalrelationsinMyanmar.

122
Interviews,State/Regionhluttawmembers(JanuaryMarch2014);BartRobertson,Overviewof
theMyanmarBudget(Yangon:ActionAid,April2014),15.
42
FOUR:ManagingMyanmarsFiscalDecentralization
Thisreportproposesstartinganincrementalfiscal decentralizationreformstrategy
withthreereformtracks,correspondingtothemajorexistingchannelsofsubnational
resourceflowstostatesandregionsandsmallerunitsbelow.123Theoutlinesofthe
initialreformsineachtrackwereintroducedintheprevioussectionofthisreportand
are reviewed here. The implementation of such a strategy requires the Union
governmenttoprovideaframeworktoguidereforms,andstrongcoordinatingpolicy
institutionsandprocesses,consideredinthefinalsectionsofthisreport.
4.1 Thethreetrackstrategy
Thefirsttrackconsistsofaframeworkforsectoraldeconcentration.Thisframework
beginswithsystematicanalysisandadjustmentoffunctionalassignmentsfollowedby
the introduction of sectoral budget norms where needed to improve subnational
budgetallocationswithinmajorUnionlineministries.Thisprocessruninparallelwith
changestobudgetpreparationandpresentationtoallowceilingsandfinalallocations
tostate/regiondepartmentstobetransparentwithintheUnionbudget.
The second and third tracks involve rationalizing the preparation of state/region
budgets through the development of a more integrated transfer system. Separate
transferscanbedesigned,simulated,andpotentiallypilotedforcurrentstate/region
responsibilities in the ordinary budget, and to integrate subnational development
fundsthatarecurrentlydistributedinwaysthatarenotbestsuitedtotheirobjectives.
Theremaybeneedforaparalleldiscussionoverageneralpurposegranttomeetgoals
relatedtothedistributionofnaturalresourcewealth.
Thereisanargumentthatthecontingentriskofadecentralizationreformisatleast
equal to the share of public expenditure being decentralized. 124 This three track
approachlimitssomeofthefiduciaryandcapacityrisksofrapiddecentralizationby
supportingamoregradualprocessofdeconcencentrationinmajorsectors,butalso
provides some of the rapid and democratic benefits of decentralization through
devolution by dramatically boosting the autonomy of state/region government
throughthesecondandthirdtracks.

123
TheseproposalsbenefittedfrompersonalcommunicationwithJamieBoexandcommentsfrom
RobertEbelonearlierwork,particularlyregardingreformofEgyptssimilarlyhybridsystemofvertical
andhorizontaldecentralization.SeeJamieBoex,DemocratizationinEgypt:ThePotentialRoleof
Decentralization,PolicyBrief(Washington,DC:UrbanInstitute,February2011);JorgeMartinez
VazquezandAndreyTimofeev,DecentralizingEgypt:NotJustAnotherEconomicReform,International
StudiesProgramWorkingPaper(Atlanta,GA:AndrewYoungSchoolofPolicyStudies,December
2008);KhaledZ.AminandRobertD.Ebel,EgyptianIntergovernmentalRelationsandFiscal
Decentralization,EgyptPublicExpenditureReview(Washington,DC:WorldBank,January2006).
124
43

Many of the principles for determining the appropriate vertical and horizontal
distributionofresourcesbetweenlevelsandacrossterritorialunitsaresimilarforboth
deconcentrationanddevolutionreforms.However, thelatterintroducesmorelocal
autonomyanddemocraticaccountability.Forexample,deconcentrationreformscan
beseenassimilartoakindofconditionaltransferwithinaministry.Certainlyboth
setsofreformsrequiresimilaradjustmentstonationalfiscalproceduresforbudget
preparation to allow ceilings to bereleasedearly enough for subnational planning.
Theseincludepreparationofmediumtermfinancialframeworksandbudgets,aswell
asarevisedannualbudgetcalendarandguidelines.
Figure7:Threereformtracks
Track1
Track2
Track3
Reforms in each track must also be able to interact and support the others. For
example,inCambodiaearlyphasesofdecentralizationfocusedoncapitalinvestment
fundsatvariouslevels.Thissituationissimilarin somewaysto Myanmars,where
effortstomoveforwardwithsubnationaldevelopmentfundsthreatentooutpacea
more holistic view of the system linking the central and local levels for improved
services.However,theMyanmarstatehasmorecapacitythanCambodiasdidatthe
startofitstransition,andthecountryisnotasaiddependent,soshapingtheentire
reform agenda around local development funds ignores the capacity that already
exists:
44

decentralisationisnotjustaboutlocalparticipatoryplanning,butalsoabout
gettingbasicservicestopeople.Todothat,meaningfulfiscaldecentralisation
isofthegreatestimportance.125
Many activities being pursued with subnational development funds overlap
substantiallywithresponsibilitiesofmajorsectorsthatarestillUnionresponsibilities:
these funds are used to build or support schools, clinics, and rural development.
Reformstobudgetingthatbettercoordinatetheselocalprojectswithsectorministry
plansorstrengthenministrycoordinationwithstate/regiongovernmentbeforethe
budgetispassedwillstrengthentheeffortsofbothsides.Theymayalsolaytheground
foradoptingamoredevolvedsysteminthefuture,inwhichlocalgovernmentshave
increaseddiscretionoversectorresponsibilities.
Finally, these tracks may converge over time. As the second part of this report
emphasized, the longterm scope of devolved government in Myanmar is unclear.
Additional responsibilities may be devolved to state/region government over time,
and/orelectedlocalgovernmentsmaybeintroducedatloweradministrativelevels.A
threetrackapproachallowsfortheplanningandfinancialmanagementcapacitiesof
theselocalgovernmentsandadministrationstobebuilt,evenasthefutureextentof
devolved government or even federalism remains unclear. An important reminder
here is that these reforms must be undertaken with great care in areas where
governanceauthorityoverpublicservicesisstillcontested.
Coordinating,integrating,oreventuallyevencompletelycombiningthesethreetracks,
isamassivepolicyinitiative,butthereiscurrentlynoclearpolicyhomeforthisactivity
withintheMyanmargovernment.Whilesomepiecemealreformshavebeendriven
outofthePresidentsOffice,theyhavenotbeencloselycoordinatedwiththeplanning
andbudgetinginstitutionsintheMinistryofFinanceandRevenueandPlanningand
EconomicDevelopment.Atthesametime,theFinancialCommissionhasremained
largely a forum for compiling budget proposals, rather than setting fiscal policy
betweenUnion andstate/regiongovernment.Andtherehasbeenlittleinteraction
between the ongoing fiscal decentralization reforms and the broader policy
discussionsovertheconstitution,naturalresourcewealth,andthepeaceprocess.
4.2 Institutionsandprocessesforsuccessfulfiscaldecentralization
Myanmars political and reform environment presents great opportunities because
therearestrongincentivesforreformtosucceed.Butthereisanurgentneedfora
nationalpolicycoordinationcapacitythatsofarisnotbeingexercisedconsistentlydue
toafocusonpiecemealandrapidreformssuchasanincreasedstate/regionbudget
share.Decentralizationcandemonstratebenefitsquicklysuchaslocaldevelopment

125
PakKimchoeun,FiscalDecentralisationinCambodia,3.
45

projectsandnewstaffinsectordepartmentsbutitisnotaquickwinreform.It
requires a consistent and clear effort to establish objectives, coordinate many
stakeholders,anddevelopandimplementoverarchingpolicies.Someofthenecessary
componentsofMyanmarslongertermfiscaldecentralizationstrategyareoutlinedin
Annex4.
Manyinstitutionalmodelshavebeenadoptedtomanagedecentralizationreform.In
somecountries,aMinistryofLocalGovernmentorsimilarinstitutionmayhavethis
responsibility.However,acommonissueisthatsuchinstitutionstendtobeableto
act on administrative and even some political decentralization, but have little
influenceoverfiscalmatters,whicharedominatedbyMinistriesofFinance.Egypthad
aMinistryofLocalDevelopmentpriortotherecentuprisings,butitwasunabletogain
traction on fiscal decentralization proposals developed there. In Afghanistan, a
Directorate for Local Government introduced ambitious policy documents on local
governance, but the Ministry of Finance retained a tightly centralized budget,
preventingthekindofreformsadvocatedinthispaper.
Onesolutiontothisproblemistoleadreformfromaninterministerialtaskforce,as
was adopted in the early 2000s in Kenya to manage a national decentralization
program. Other countries focus on developing specialpurpose, somewhat more
neutral,andhighlevelmechanismstooverseeandcoordinatedecentralization.126
Themechanismneedstobalancetherequirementsofpoliticaldecisionmakingand
soundtechnicalanalysisandpolicydevelopment.Asdecentralizationisacrosscutting
issuethatinfluencespublicfinancialmanagementofallkinds,itneedsparticipation
ofallkeybudgetaryinstitutions.Suchaninstitutionmaytaketheformoftemporary
ormorelastinginstitutionssuchasanationalcommission,usuallybasedinoraround
thePrimeMinisterorPresidentsOffice.
Whatever the structure for driving policy, the existence of some form of Finance
Commissionorfiscalforumtodefinetheintergovernmentalfiscalsystemiscrucial.
Aswiththepoliticallevel,theseinstitutionstakemanyforms,eitherpermanentor
intermittent, and independent or governmentled. Lessons gathered across many
countrieswithsuchinstitutionssuggestthattheyshouldbepermanentbodieswitha
limited membership; but include the Minister of Finance, Minister of Interior or
equivalent, representatives of local governments, and more independent advisory
membersfromacademia,civilsocietyandtheprivatesector.127
Box6:Selecteddecentralizationpolicycoordinationmechanisms
Ethiopia:RegionalAffairsDepartmentintheOfficeofthePrimeMinister.

126
Smoke,StrategicFiscalDecentralizationinDevelopingCountries,103.
Steffensen,FiscalDecentralizationandSectorFundingPrinciplesandPractices,33;RoyKelly,
unpublishedpresentation,DukeUniversity(August2008).
127
46

Uganda: Decentralization Secretariat reporting to the Permanent Secretary for Local
Government, but staff are not civil servants. Local Government Finance Commission is
independentandreportstothePresident,influencinglocalrevenueandtransferdesign.
Cambodia: Subnational Democratic Development Program is designed and managed by a
permanent interministerial National Commission on Democratic Development (NCDD)
chairedbytheMinisterofInterior.TheNCDDhasitsownbudget,supportingsecretariat,and
subcommittees on functions and resources, fiscal and financial affairs, and personnel of
subnationaladministrations.128
India:Anindependentadvisorybodywith5membersperiodicallyconstitutedeveryfiveyears
tofocusoncentralstate level allocationsofrevenues andgrants.Subnational(state)level
financecommissionswereintroducedinthe1990s.
Canada:Provincesmeetperiodicallyinintergovernmentalfirstministersforum.
South Africa: Permanent autonomous and independent financial and fiscal commission
includingprovincialandlocalrepresentatives,withfulltimesecretariat.
Itwouldbeunwiseforthispapertoprovidecomprehensiverecommendationsona
specific decentralization policy management arrangement for Myanmar. However,
twogeneralobservationsmightbemade.First,givenMyanmarscontextofpolitical
decentralization leading administrative and fiscal dimensions and its pending high
levelquestionsabouttheconstitution,thepeaceprocess,andnaturalresources,itis
likelythatahigherlevelpoliticalbodysuchasanationalcommissionorsecretariatis
moresuitedthanaLocalGovernmentorHomeAffairsMinistryledmodel.
Second,onceagainMyanmarhassomeimportantbuildingblocksinplace.Thereisan
existing constitutional Financial Commission that meets annually to consider the
budget.However,theconstitutionactually allowsforapotentiallymoresignificant
policymaking role for the Financial Commission: namely to carry out the duties
assigned by the Pyidaungsu Hluttaw through the promulgation of law for the
emergence of a substantial financial system. 129 The Financial Commission already
includessomekeycentralandsubnationalplayerstheVicePresidents,theAuditor
General,theMinisterofFinanceandtheChiefMinistersofthestatesandregions.
Inadditiontoahigherlevelbodytoguidepolicywithwiderparticipation,theFinance
Commissioncouldtobegiveninlawinaccordancewiththeconstitutionadefined
policy role in designing precisely the fiscal frameworks for deconcentration and
devolution introduced in this paper. Essentially the Financial Commission needs to
movefrombeingaforumforintergovernmentalbargainingtoapolicysettingbody.
To do so would require substantial investment in a technical secretariat, and
preferably legislative requirements for it to consult widelyboth within and beyond
governmentonpolicydecisionsitmakes.

128
129
Niazi,DeconcentrationandDecentralizationReformsinCambodia,12.
ConstitutionoftheRepublicoftheUnionofMyanmar,Art.230.
47

ThispaperhasstressedthroughoutthatMyanmarcanstrengthenitsdecentralization
processinanincrementalway,whilealsoworkingtowardsresolutionofalongerterm
vision. In this sense, these policy coordination institutions do not need to start by
determiningeverything,asinthetextbookmodelpresentedintheintroduction.In
fact,
Developingaprocesstobuildconsensusabouttheappropriateextent,form,
andpaceoffiscaldecentralizationisatleastasimportantasandinitially
oftenmoreimportantthandefininganendproductsystemthatconforms
tothenormativedecentralizationmodel.Thisisparticularlycriticalinhighly
centralizedcountrieswherethereisnounifiedvisionofdecentralization
Suchaprocess,ifproperlystructured,representativeofthekeystakeholders,
ledbyatrustedorneutralagency,broadlyperceivedasfair,andadequately
empowered,canhelpdefinearealisticstartingpointforfiscal
decentralization.130
Themostimportantrecommendationsofthispaper,therefore,arenotjustabouta
specific decentralization proposal, but rather the need to form coordinating
institutions and a policy process to manage these evolving fiscal decentralization
reforms.Therearemanywaystoapproachdecentralization,andwiththepaceand
uncertaintyofMyanmarstransition,amuchmoredramaticprocessisapossibility.
Regardlessofthepathtaken,theanalysisinthispaperisintendedtocontributetoa
moreinformeddiscussionabouthowbesttofulfilthehopesofallMyanmarspeoples,
nomatterwheretheylive,foramoredevelopedandpeacefulcountry.

130
Smoke,StrategicFiscalDecentralizationinDevelopingCountries,107.
48
Annex1:Political,administrativeandfiscaldimensionsof
decentralization
Political decentralisation involves the transfer of decisionmaking power and
accountabilitytolocallevels.Itofteninvolvessomeformofdevolutionthetransfer
ofresponsibilitiestolocalgovernmentsthathavebeengrantedsignificantautonomy.
In democratic decentralization these local governments are accountable to local
populationsthroughelectionsand/orothermeans.
Administrative decentralization focuses on distributing managerial responsibilities
(for example, for delivering a given public service) among different levels of
government or administration. It may also take the relatively modest form of
deconcentration in which officials at lower levels are given more authority or
discretion,butremainaccountabletotheirchiefsatthecentre.Decentralizationcould
alsobecombinedwiththedevolutionofexecutiveauthoritytolocalgovernments.
Delegationistheshiftingoffunctionstosemiorwhollyindependentorganizations
outside thecoregovernment sector (forexample to independent authorities, or in
privatization,toprivatefirms.
Fiscaldecentralizationdescribesthewayinwhichtheexpenditureresponsibilitiesare
assignedandcorrespondingresourcesareprovided.Theseresourcesmaybeprovided
bydeconcentratingcontrolovercentralfundstolowerlevels,ordevolving tolocal
government,amorecomprehensivesystemofplanningandbudgetingsupportedby
assignmentoflocalrevenues,centrallocaltransfers,andpossiblylocalborrowing.
ExcerptedfromNixonetal.,StateandRegionGovernmentsinMyanmar.
49
Annex2:Designingtransferschemes
1. Definethepurposeofeachscheme
Consideringthesystemobjectives,thepurposeofthisspecifictransferscheme
shouldbedefined.Itisprobablyeasiesttolimittheobjectivesofagivenschemeto
preservesimplicityandtransparency:Willthisschemefocusonprovidinggeneral
purposefundsforlocalservicedeliveryordevelopment,supportnationalpolicy
objectives,orequalizeconditionsamongdifferentpartsofthecountry?
2. Definethetransfermechanismsorwindows
Dependingonthepurposesofthetransfer,severaldesignchoicesneedtobemade.
Thefirstiswhatconditionswillbeattachedtothetransfer.Thesemightrequirethat
thetransferistobeusedforaspecificsector(health,education),purpose(local
roads),orexpenditurecategory(recurrentorcapital).Ontheotherhand,thegrant
maybeunconditionalorgeneralpurpose,foruseatthediscretionofthelocalunit.
Eachtypeoftransferhasadvantagesanddisadvantages.
Advantagesanddisadvantagesofconditionalandunconditionaltransfers
Conditionaltransferscansupportspecificnationalpolicygoalsorstandardsof
service,stimulateservicedeliveryinpriorityareas,orcompensatelocalareasfor
spillovers.Ontheotherhandtheycanbecomplicatedandexpensivetoadminister,
andmaydistortlocalprioritiesandconfuseaccountabilityforthedeliveryof
services.
Unconditionaltransfersbuilduplocalautonomyandplanningcapacity,can
strengthenlocalaccountabilityandpoliticaldecentralization,andarerelatively
simpletoadminister.However,theymaycrowdoutlocalspendingonlocalservices,
andrequiregoodlocalfinancialmanagement.
Ingeneral,conditionalgrantsshouldbeusedwhenapolicyobjectivecannotbemet
byothermeans,anditisgoodtoconsideranexitstrategytophasethemoutin
favouroflessconditionalmechanisms.Thereshouldbeabalancebetween
conditionalandmorediscretionaryfundsinthelocalbudget.
Second,thetransfermaybedeliveredasalumpsumorinsomematchingformthat
variesdependingonlocalexpendituresforthesamepurpose.Ifitismatching,the
grantshouldbedefinedasopenendedorwithafixedceiling.Inaddition,transfers
areincreasinglybeingdesignedthatincorporateaperformancebasedcomponent,
bywhichtheamounttransferredvariesaccordingtolocalgovernmentsperformance
inagivenareaforexamplefinancialmanagement,revenuemobilizationor
developmentoutcomes.Suchperformancebasedsystemsrequirecareful

considerationinordertoincentivizeandsupportcapacityforgoodperformance
withoutpunishingdisadvantagedjurisdictions.131
3. Determinehowthetransferpoolisestablished(verticalallocation)
Thetransferpoolisthetotalamountofresourcestobetransferredtolowerlevels
throughthescheme,beforeconsideringhowitistobedividedup.Thispoolmaybe
determinedinanumberofways.
Thepoolmaybedeterminedonanadhocbasisorbedefinedbyaprogram,
forexamplethroughdonorlocaldevelopmentfunds.
Thepoolmaybeadefinedshareofthetotalnationalbudgetorrevenues,or
someportionofaspecificrevenuestream.
Itmaybebasedonthereimbursementoftheexpendituresoflocal
governments,mostcommonforconditionaltransfers.
Inmoreunusualcasesthepoolmaybebasedonestimatingtheexpenditure
needsoflocalgovernment(forexampletomeetastandardlevelofservice
delivery).Asitmustbebasedonexpenditureandrevenueestimatesthis
processcanbecomplicated.
Thedeterminationofthepoolofresourcesshouldberelatedtoitspurpose.For
exampleanequalizationtransferpoolfocusedonjustlydistributingnaturalresource
revenuewouldbeaportionofthatspecificrevenuestream,abroadersocio
economicequalizationpoolmightbeaportionofthenationalbudget,andan
equalizationgrantaimedatpromotingnationalservicestandardsmightbe
determinedonareimbursementbasis.Ingeneral,adhoctransfersaretheleast
decentralizingform,astheytendtorelyonpoliticalnegotiationsatthecentreand
canbeunpredictableandunderminelocalplanningandbudgeting.
4. Determinehowthepoolisdividedamongsubnationalunits(horizontal
allocation)
Oncethemechanismandpoolaredetermined,itremainstoestablishtheallocation
ordivisionoftheresourcesamongtheeligiblesubnationalunits.Thisdivisioncanbe
determinedinfourorfivebroadways,alsoaccordingtothepurposesofthe
transfer:
Oftenthedivisionofthepoolisdeterminedthroughanadhocorcaseby
caseprocessthroughpoliticalnegotiationsorexecutiveorder.When
transfersformpartofadevelopmentassistanceprogramitwillusuallyhave
itsownpredeterminedallocationcriteria,oftenastandardgrantamount.

131
ForathoroughpresentationofperformancebasedgrantdesignprinciplesseeUNCDF,
PerformanceBasedGrantSystems:ConceptsandInternationalExperience.
51
Thedivisionmaybebasedonthederivation,ororiginofthefunds.Inthis
casesomeproportionofagivenrevenuestreamisdirectedbacktowhereit
wascollected.
Itmaybebasedontheconditionalreimbursementofexpendituresfora
specificpurposebyeachlocalgovernment.
Itmaybebasedonanobjectiveformulabasedallocationusingobjectiveand
verifiableallocationfactorsthatareassigneddifferingweights.Sometimes
performancecriteriaareincludedintheseallocationfactors,sothattransfers
rewardgoodlocalgovernmentperformance.
Adhocorpoliticallynegotiatedtransfersarecommon,butoftenhaveproblemswith
theirlackoftransparency,participationandobjectivityleadingtoperceptionsof
bias.Asaresulttheyoftenrelytoomuchontheequalsharesprincipletoavoid
controversy.Therecentdistributionofpovertyreductiondevelopmentgrantsin
Myanmarisanexample:giving1billionKyatperstate/regionappearsfair,butdoes
notreflecttherelativeexpenditureneedsofeachstateandregion.Further
adjustments(forexamplegivinganadditional2billionKyattoChinState)may
appeararbitraryandarestillnotdirectlyrelatedtodefinedneeds.Ontheother
hand,someadhocplanningmaybeneededinthecaseofexceptionalorlarge
capitalinvestments.
Thederivationprincipleisoftenusedtoensurelocalgovernmentsareassuredthata
portionofthefinancialresourcesraisedthereareavailablefortheiruse.Sucha
principleisparticularlyimportantinpoliticalagreementsovernaturalresource
revenuesorinsharingoutnationallycollectedtaxes.Thetransferdoesnothaveto
beofallresources,assomeresourcesmaybereservedfornationalpurposesorto
makeequalizingtransferstolessadvantagedregions.
Reimbursementtransfersaredistributedaccordingtotheexpenditureofthelocal
governmentonaparticularserviceforexampleteacherssalaries.Theyaimto
ensurenationalpolicygoalsarepursuedorservicesdeliveredatanationalstandard.
Possibleproblemsarethatanopenendedcommitmentwilldiscouragelocal
revenuecollectionforthesameservice,andthiscanbeaddressedbycappingthe
grantpoolordefiningamatchingratio,bywhichthegrantisbasedonthelevelof
localrevenuesused.Withmatchinggrants,caremustbetakentoavoiddistortingor
crowdingoutotherpreferredformsoflocalspending,ortosupportthewealthiest
regionsmorethanpoorerones.
Formulabasedtransfershavetheadvantagesoftransparency,clearlinkagetothe
objectivesofthetransfer,andflexibility.Sincetheyareobjectivelydetermined,the
impactofdifferentallocationfactorsandweightscanbemodeledinadvance,and
canbeadjustedovertimemanycountriesevaluatetheweightsandallocations
usedthrougharegularinstitutionalprocess.Formulabasedtransferscanbeshaped
toservearangeofobjectivesincludingequalization,supportfornationalpolicy
goals,localdevelopment,andincentivizingimprovedperformance.
52
Annex3:SchedulesTwoandFiveoftheMyanmarConstitution
RegionorStateLegislativeList(ScheduleTwo)
1.FinanceandPlanningSector
(a)TheRegionorStatebudget;
(b)TheRegionorStatefund;
(c)Landrevenue;
(d)Exciseduty(notincludingnarcoticdrugsandpsychotropicsubstances);
(e)Municipaltaxessuchastaxesonbuildingsandlands,water,streetlightingsand
wheels;
(f)ServicesoftheRegionorState;
(g)Sale,leaseandothermeansofexecutionofpropertyoftheRegionorState;
(h)DisbursementofloansinthecountryfromtheRegionorStatefunds;
(i)InvestmentinthecountryfromtheRegionorStatefunds;
(j)Localplan;and
(k)Smallloansbusiness.
2.EconomicSector
(a)EconomicmattersundertakenintheRegionorStateinaccordwithlawenacted
bytheUnion;
(b)CommercialmattersundertakenintheRegionorStateinaccordwithlaw
enactedbytheUnion;and
(c)CooperativemattersundertakenintheRegionorStateinaccordwithlaw
enactedbytheUnion.
3.AgricultureandLivestockBreedingSector
(a)Agriculture;
(b)Protectionagainstandcontrolofplantsandcroppestsanddiseases;
(c)Systematicuseofchemicalfertilizersandsystematicproductionanduseof
naturalfertilizers;
(d)Agriculturalloansandsavings;
(e)Dams,embankments,lakes,drainsandirrigationworkshavingtherighttobe
managedbytheRegionorState;
(f)Freshwaterfisheries;and
(g)Livestockbreedingandsystematicherdinginaccordwiththelawenactedbythe
Union.
4.Energy,Electricity,MiningandForestrySector
(a)Mediumandsmallscaleelectricpowerproductionanddistributionthathavethe
righttobemanagedbytheRegionorStatenothavinganylinkwithnationalpower
grid,exceptlargescaleelectricpowerproductionanddistributionhavingtherightto
bemanagedbytheUnion;
(b)Saltandsaltproducts;
(c)CuttingandpolishingofgemstoneswithintheRegionorState;
(d)Villagefirewoodplantation;and
(e)Recreationcentres,zoologicalgardenandbotanicalgarden.
53
5.IndustrialSector
(a)IndustriesotherthanthoseprescribedtobeundertakenbytheUnionlevel;and
(b)Cottageindustries.
6.Transport,CommunicationandConstructionSector
(a)Ports,jettiesandpontoonshavingtherighttobemanagedbytheRegionor
State;
(b)RoadsandbridgeshavingtherighttobemanagedbytheRegionorState;and
(c)SystematicrunningofprivatevehicleswithintheRegionorState.
7.SocialSector
(a)Mattersontraditionalmedicinenotcontrarytotraditionalmedicinepolicies
prescribedbytheUnion;
(b)SocialwelfareworkswithintheRegionorState;
(c)Preventiveandprecautionarymeasuresagainstfireandnaturaldisasters;
(d)Stevedoring;
(e)HavingtherightofmanagementbytheRegionorState,thefollowing:
(i)preservationofculturalheritage;
(ii)museumsandlibraries.
(f)Theatres,cinemasandvideohouses;and
(g)Exhibitionssuchasphotographs,paintingsandsculptures.
8.ManagementSector
(a)Developmentmatters;
(b)Townandhousingdevelopment;and
(c)Honorarycertificatesandawards.
TaxesandFeesCollectedbyRegionorStates(ScheduleFive)
1. Landrevenue.
2. Exciserevenue.
3. Watertaxandembankmenttaxbasedondamsandreservoirsmanagedbythe
RegionorStateandtaxonuseofelectricitygeneratedbysuchfacilitiesmanagedby
theRegionorState.
4. TollfeesfromusingroadsandbridgesmanagedbytheRegionorState.
5. (a)Royaltycollectedonfreshwaterfisheries;(b)Royaltycollectedonmarine
fisherieswithinthepermittedrangeofterritorialwater.
6. Taxescollectedonvehiclesonroadtransportandvesselsoninlandwaterway
transport,inaccordwithlaw,inaRegionoraState.
7. Proceeds,rentfeesandotherprofitsfromthosepropertiesownedbyaRegionor
aState.
8. Fees,taxesandotherrevenuescollectedonservicesenterprisesbyaRegionora
State.
9. FinesimposedbyjudicialcourtsinaRegionoraStateincludingRegionTaya
HluttaworStateTayaHluttawandtaxescollectedonserviceprovisionandother
revenues.
54

10. InterestsfromdisbursedbyaRegionorState.
11. ProfitsreturnedfrominvestmentofaRegionorState.
12. TaxescollectedonextractionofthefollowingitemsfromtheforestsinaRegion
oraState:
a. Taxescollectedonallotherwoodsexceptteakandotherrestricted
hardwoods;
b. Taxescollectedonfirewood,charcoal,rattan,bamboo,birdnests,
cutch,thanetkha,turpentine,eaglewoodandhoneybasedproducts.
13. Registrationfees.
14. Taxesonentrainments.
15. Salttax.
16. RevenuereceivedfromtheUnionFundAccount.
17. ContributionsbydevelopmentaffairsorganizationsinaRegionorState
concerned.
18. Unclaimedcashandproperty.
19. Treasuretrove.
55
Annex4:ElementsforMyanmarsDecentralizationStrategy
DecentralizationStrategyforCentreState/Regions
Functional
Assignments
FiscalFramework
Intergovernmental
FiscalSystem
Deconcentrationof
sectorbudgets
State/RegionTransfer
System
WealthSharing
ScheduleIIClarification
Revenue
Policyand
Administration
ResourceRevenue
UnbundlingServices
State/RegionPolitical
Autonomy
BroadeningState&
RegionResponsibilities
ResourceManagement
Functions
KeySocialServices
Clarifying
Accountability
Relationships
CivilServicein
State/Region
Political
Decentralization&
Autonomy
IntegrationwithPeace
Process
ParliamentaryDialogue
Autonomyofhluttaw,
accountability
ofCM
References
Amin,KhaledZ.,andRobertD.Ebel.EgyptianIntergovernmentalRelationsandFiscal
Decentralization.EgyptPublicExpenditureReview.Washington,DC:WorldBank,
January2006.
Andrews,Matthew.TheLimitsofInstitutionalReforminDevelopmentChangingRules
forRealisticSolutions.Cambridge:CambridgeUniversityPress,2013.
Bahl,RoyW.,andJorgeMartinezVazquez.SequencingFiscalDecentralization.World
BankPublications,2006.
Bissinger,Jared.SubnationalGovernmentsandBusinessinMyanmar.Yangon:TheAsia
Foundation/CentreforEconomicandSocialDevelopment,February2014.
Boex,Jamie.DemocratizationinEgypt:ThePotentialRoleofDecentralization.Policy
Brief.Washington,DC:UrbanInstitute,February2011.
BurmaNewsInternational.DecipheringMyanmarsPeaceProcess.ChiangMai:BNI,
March2014).
Charbit,Claire.GovernanceofPublicPoliciesinDecentralisedContexts.OECDRegional
DevelopmentWorkingPapers.Paris:OrganisationforEconomicCooperation
andDevelopment,July19,2011.http://www.oecd
ilibrary.org/content/workingpaper/5kg883pkxkhcen.
ConstitutionoftheRepublicoftheUnionofMyanmar.MinistryofInformation,
GovernmentofMyanmar,2008.
Eaton,Kent,KaiKaiser,andPaulJ.Smoke.ThePoliticalEconomyofDecentralization
Reforms.Washington,D.C:WorldBank,2011.
gert,Balzs,andHansjrgBlchliger.DecentralisationandEconomicGrowthPart2:
TheImpactonEconomicActivity,ProductivityandInvestment.OECDWorking
PapersonFiscalFederalism,June3,2013.http://www.oecd
ilibrary.org/taxation/decentralisationandeconomicgrowthpart2theimpact
oneconomicactivityproductivityandinvestment_5k4559gp7pzwen.
FactorsAffectingFiscalDecentralizationinCambodia.ParliamentaryInstituteof
Cambodia,November2012.
GovernmentoftheUnionofMyanmar.FrameworkforEconomicandSocialReform,
December2012.
Jones,Lee.ThePoliticalEconomyofMyanmarsTransition.JournalofContemporary
Asia44,no.1(January2,2014):14470.
Lindblom,CharlesE.TheScienceofmuddlingthrough.PublicAdministrationReview
19,no.2(Spring1959):7988.

LocalDevelopmentInternational.TheRoleofDecentralisation/DevolutioninImproving
DevelopmentOutcomesattheLocalLevel:ReviewoftheLiteratureandSelected
Cases.DFID/LocalDevelopmentInternational,November2013.
LocalPublicSectorInitiative.MeasuringtheLocalPublicSector:AConceptualand
MethodologicalFramework.TheUrbanInstitute,December2012.
MartinezVazquez,Jorge,andAndreyTimofeev.DecentralizingEgypt:NotJustAnother
EconomicReform.InternationalStudiesProgramWorkingPaper.Atlanta,GA:
AndrewYoungSchoolofPolicyStudies,December2008.
Mutebi,AlexM.MyanmarsCivilBureaucracyandtheNeedforReform.InMyanmar:
BeyondPoliticstoSocietalImperatives,140.Singapore:InstituteofSoutheast
AsianStudies,2005.
Muwonge,Abdu,andRobertD.Ebel.IntergovernmentalFinancesinaDecentralized
World.InMunicipalFinances:AHandbookforLocalGovernments,editedby
CatherineFarvacqueVitkovicandMihalyKopanyi,139.Washington,DC:World
Bank,2014.
Niazi,TariqH.DeconcentrationandDecentralizationReformsinCambodia.Philippines:
AsianDevelopmentBank,2011.
Nixon,Hamish,andJamesonBoex.ImplementingGIROAsSubnationalPlanningand
FinancePolicy:OptionsandInternationalPractice.WorldBank:PolicyNote
PreparedforGovernmentofAfghanistan,2010.
Nixon,Hamish,CindyJoelene,KyiPyarChitSaw,ThetAungLynn,andMatthewArnold.
StateandRegionGovernmentsinMyanmar.Yangon:TheAsiaFoundation/MDRI
CentreforEconomicandSocialDevelopment,September2013.
PakKimchoeun.FiscalDecentralisationinCambodia:AReviewofProgressand
Challenges.CDRIWorkingPaperSeries50.PhnomPenh:CDRI,2011.
Parks,Thomas,andWilliamCole.PoliticalSettlements:ImplicationsforDevelopment
PolicyandPractice.OccasionalPaper.TheAsiaFoundation,July2010.
Pritchett,Lant,andVaradPande.MakingPrimaryEducationWorkforIndiasRuralPoor:
AProposalforEffectiveDecentralization.SocialDevelopmentPapers.
Washington,D.C.:WorldBank,June2006.
PyidaungsuHluttaw.ReportofConstitutionalReviewJointCommitteesFindingsand
Assessment.NayPyiTaw:GovernmentoftheRepublicoftheUnionofMyanmar,
31January2014.
Robertson,Bart.OverviewoftheMyanmarBudget.Yangon:ActionAid,April2014.
Smoke,Paul.FiscalFederalisminDevelopingCountries:AReviewofCurrentConcepts
andPractices.NewYork:UnitedNationsResearchInstituteforSocial
Development,2001.
58

.StrategicFiscalDecentralizationinDevelopingCountries:LearningfromRecent
Innovations.InLocalDynamicsinanEraofGlobalization,editedbyShahid
Yusuf,WeipingWu,andSimonEvenett,1019.WorldBank,2000.
Steffensen,Jesper.FiscalDecentralizationandSectorFundingPrinciplesandPractices.
DANIDA,November2010.
VanZyl,Albert.WhatIsWrongwiththeConstituencyDevelopmentFunds?BudgetBrief.
InternationalBudgetPartnership,2010.
Wild,Leni,andMartaForesti.WorkingwiththePolitics,2013.
http://www.odi.org.uk/sites/odi.org.uk/files/odiassets/publicationsopinion
files/8587.pdf.
WorldBank.RepublicoftheUnionofMyanmarPublicFinancialManagement
PerformanceReport.PublicExpenditureandFinancialAccountabilityReport.
Yangon:WorldBank,May2013.
59

MyanmarTowardsARoadmaptoReform en Red

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MyanmarTowardsARoadmaptoReform en Red

Uploaded by

Copyright:

Available Formats

FiscalDecentralizationinMyanmar:TowardsaRoadmapforReform

Hamish Nixon is an independent governance specialist focusing on the areas of

hybrid or nonstate systems, and decentralizing central government services has

Channel 1: Line ministries or departments that remain under Union

an unconditionalblock grant forordinaryfunctionsassigned tothatlevel,

coordinate many stakeholders, and develop and implement overarching policies.

There are important differences between devolution and deconcentration. In

Improving vertical fiscal balance (ensuring sufficient resources for

an unconditional block grant for those ordinary functions assigned to

You might also like