INCLUSIVE INNOVATION AND RAPID

SOCIOTECHNICAL TRANSITIONS:

THE CASE OF MOBILE MONEY IN KENYA

Elsie Onsongo and Johan Schot

1

Inclusiveinnovationandrapidsociotechnicaltransitions:ThecaseofmobilemoneyinKenya1

ElsieOnsongo†,2andJohanSchot§,

†CentreforFrugalInnovationinAfrica,InstituteforSocialStudies,ErasmusUniversityRotterdam,theNetherlands§SciencePolicyResearchUnit,UniversityofSussex,UnitedKingdom

AbstractMobilemoney innovation is at the centre of a sociotechnical transformation in the financialservicessectorinKenyathatsawchangesintheregulatoryframework,inmarketshares,userpractices and social networks acrossmultiple regimes.While sociotechnical transitions oftentakeabout50years, themobilemoneyrevolution inKenyahas takenonly15years,andhasresulted in remarkable levels of financial inclusion of marginalised people. In this paper wecombinethemulti-levelperspective(MLP)fromthesustainabilitytransitionsliteratureandtheLadder of Inclusivity (LII) from the inclusive innovation literature to explain this rapidtransformation.Applyingbothframeworkstothecaseenablesustoelucidatepotentialareasforcross-fertilisationbetweenMLPandLII,thusrespondingtocallsfortheinclusionofasocialdimensioninthetransitionsperspective,whileexplaininghowprocessesofchangeenvisioned(butweaklydefined)inLIIcanbeexplored.Basedonourfindings,wehypothesisethatarapidsocio-technicalsystemtransitiontakes1)stronglandscapepressures,anopeningupofregimewhichmakes regimeactorswilling toactandstrongnichedevelopmentalloriented towardsthesamegoal,and2)regimeactorsmotivatedbythenormativegoaltotakerisksandrespondfavourablytothemixofdevelopmentsatallthreelevels.KeywordsSustainabilitytransitions,multilevelperspective,inclusiveinnovation,ladderofinclusivity,mobilemoney,financialservices1 Introduction

Policymakers and international organisations in the developing world now consider mobilemoneyinnovationasapromisingavenuefortacklingfinancialexclusionamonglow-incomeandvulnerable segments of the population (ADB, 2013; GSMA, 2015; IMF, 2016; World Bank,2013b). Various multinational financial inclusion-oriented networks such as the Better thatCashAlliance,theAllianceforFinancialInclusionandtheG20FinancialInclusionPeerLearningProgramareexploringhowlessonslearnedfromsuccessfuldeploymentsofmobilemoneycan

1ThispaperhasbenefitedfromdiscussionswithparticipantsoftheJenaEconomicResearchWorkshops,the7thInternationalSustainabilityTransitions(IST)Conference,the22ndInternationalSustainableDevelopmentResearchSociety(ISDRS)andtheAtlantaConferenceonScienceandInnovationPolicy.SpecialthanksgotoPeterWalgenbachforhelpfulcommentsandsuggestionsonanearlierdraft,andtoRaphieKaplinskyforopeninguptheinclusiveinnovationliterature.FinancialsupportfromtheGermanScienceFoundation(DFGRTG1411)isgratefullyacknowledged.2Correspondingauthor.onsongo@iss.nl

2

be shared amongmember states.One such exemplar of success isM-Pesa, amobilemoneyplatformdeveloped,pilotedanddeployedinKenyain2007,nowoperational insevenAfricancountries,and in India,RomaniaandAlbania.M-Pesawas launched in2007,andwas rapidlyadoptedinthefirstyearofoperation,reaching5.1millionactiveusersbyDecember.Between2007and2013,M-Pesasubscriptiongrewto17.1millionuserssupportedbyanagentnetworkof 65,547 outlets (Safaricom, 2013). In this duration, other competing mobile operatorsdeployed similar platformswhich cumulatively increasedmobilemoney subscription to 26.2millionusers—about74.5percentoftheadultpopulation(CBK,KNBS,&FSDKenya,2016).Thegovernment of Kenya now credits mobile money with extending access to formal financialservicestopreviouslymarginalisedcommunities(Ndung’u,2014).ThispaperexploreshowwecanexplaintherapidtransformationoffinancialservicesinKenya.Wedothisbyanalysingthe introductionofM-Pesafromamulti-levelperspective (MLP) (Ripand Kemp, 1998; Geels, 2002; Grin et al., 2010)which is designed to explain socio-technicalsystemtransitions.TheintroductionofM-Pesacanbecharacterizedassuchbecauseitsrapidgrowthmeantnotonlythedevelopmentofanewtechnology,butalsoatransformationofthecommercial banking business model, and in fact the financial services industry. This isevidencedbychanges in theregulatory framework,userpracticesandsocialnetworksacrossthedomainsofbothformalandinformalfinance,andinmarketsharesofincumbentfinancialservice providers. Besides cash, mobile money became the most popular person-to-persontransaction medium such that it accounted for 66.56 percent of the throughput volume inKenya’snationalpaymentsystem.While the fast and deep transformation of the financial services sociotechnical system wassurprisingandneedsexplanation,however,perhapsevenmoreremarkableistheinclusionofhithertocompletelyexcludedpoorpeople in thesystem.Whereas theMLPhasproventoberelevantforanalysingsociotechnicalsystemchange,ithasnotyetbeenappliedtocaseswhereissues of social inclusion are key. Even inMLP studies of transitions in developing countries,these cases have focused on transitions to environmental sustainability, mostly low carbontransitionsinAsia(Iizuka,2015;Rock,Murphy,Rasiah,vanSeters,&Managi,2009)andAfrica(Baker, Newell, & Phillips, 2014; Byrne, 2011; Romijn & Caniëls, 2011; van Eijck & Romijn,2008). A few calls have been made for including a social dimension into the transitionperspective (Meadowcroft, 2000; Swilling & Annecke, 2012) but this remains a largelyunexploredarea.TofillthisgapwearguethatinsightsfromtheinclusiveinnovationliteraturemaybeafruitfuladditiontothesustainabilitytransitionsliteratureandinparticulartheMLP.We use the Ladder of Inclusive Innovation (LII), a framework to assess levels of inclusivity(Heeksetal.,2013).Atthispointwedonotargueforamergerbetweenbothframeworks,onthecontraryweseethemascomplementary. Therefore inthispaperweconductaseparateanalysis of the M-Pesa case from each perspective and then identify areas where bothstrengtheneachother.Inouranalysisweusemultiplesourcesofevidence.Wecollecteddocumentaryevidencefromseveral sources: annual reports, press releases, legislation, regulatory guidelines and formalremarks fromthe twoconcerned regulatoryauthorities, i.e.CentralBankofKenya (CBK)andthe Communication Commission of Kenya (CCK)3.We also collected annual reports from theinnovating firms, Vodafone Group and Safaricom Ltd, and from the four largest commercialbanks in Kenya. Further evidence was collected from the print media and television videoclippingsofreportsonM-Pesa,thebankingsectorandinformalgroups,fromadvertisements,blogsandsocialmedia.Inaddition,weusethenationallyrepresentativeFinAccessHousehold

3NowCommunicationsAuthorityofKenya(CAK)

3

Surveys of 2006, 2009 and 2013 (n=4214, n=6598, n=6449 respectively) conducted by theCentral Bank of Kenya, Kenya National Bureau of Statistics (KNBS) and Financial SectorDeepening(FSD)Kenyatotrackaccesstoanddemandforfinancialservicesamongadults.Wealsousethe2015FinAccessGeospatialMappingSurveytoplotanddescribethegeographicalreach of banking and mobile money services. Our documentary evidence and quantitativedatasetsaresubstantiatedbysemi-structured interviewswithrepresentativesfromVodafoneLtd,SafaricomLtdandtheCentralBankofKenya(n=7).Inaddition,transcriptsfromfourfocusgroupswith informalfinancialgroups(chamas) inNairobiandWesternKenyawereused.Wealso observed M-Pesa functionality and real-time transactions at agent outlets in Nairobi.Finally,publishedresearchpapersandpolicybriefsonmobilemoneyprovidedusefulinsights.Drawingevidence frommultiple independentsourcesenabledus tocheck theconsistencyofourfindingsinaprocessoftriangulation(Yin,2009).For both the MLP and the LII, we provide detailed analytic explanations of the case. Theanalyticalprocedureentailedfirstwritingadetailednarrativeorchronicleofdevelopmentsinthe financial servicessectorand innovationofM-Pesa.Thedescriptivehistoricalaccountwasthenconverted intoanalyticaccountspresented intheoretically relevant language(George&Bennett,2005,p.211).Thiswasachievedbymovingbackandforthbetweenboththeoreticalframeworksandthedata,modifyingexplanationsinlightofthetheoreticalemphasisplacedbyeachapproachonvariousaspectsofthenarrative,effectivelyfocusingon“whatarethoughttobe particularly important parts of an adequate or parsimonious explanation” (George &Bennett,2005,p.215)ineachcase.The paper proceeds as follows: in sections 2 and 3we introduce theMLP and the inclusiveinnovationapproachrespectively.Insection4and5weapplytheframeworksseparatelytotheempiricalcase,andinsection6suggesthowtheycanbecross-fertilisedsothattheycanbetteraccountforsocialexclusionandinclusion.Weconcludebyconsideringtheimplicationsofourproposalsforfutureempiricalwork.2 Themulti-levelperspective(MLP)

Oneofthecentralapproachesthatdescribesandanalysessociotechnicalchangeprocesses isthemulti-levelperspective(MLP)(Geels,2002;Geels&Schot,2007;Rip&Kemp,1998;Smith,Stirling, & Berkhout, 2005). In this model, socio-technical transitions unfold throughdevelopments on three interdependent analytical levels that exist in a nested hierarchy: thesocio-technicalregimeatthemeso-level,thenicheatthemicro-level,andthelandscapeatthemacro-level.Thenotionoflevelsrefersthustoanalyticallevels,nottolevelsinageographicalsense.Theregimerepresentsthehighlyinstitutionalized,yetnotnecessarilyfullycoherentsetofrulesthatmutuallyconstructandareconstructedbyactors.Therulesareembeddedinskills,regulations,expectations,userpreferences,productsand infrastructuresetc;elementswhichtogether form a system of provision formobility, energy, food etc. The technical and socialelements in the regime co-evolve, align and stabilise with each other over time, creatingrigiditiesthatonlyenableorallowthedevelopmentofinnovationsinspecificdirections.Theseinnovationscanbeincrementalorrevolutionaryintermsofexploitingtheopportunitieswhichexistwithintheregime,howeverthatstaywithintheboundariesspecifiedbytherule-set.Radical or disruptive innovations go beyond the rule-set and therefore arise out oftechnological niches, or protected spaces (Kemp, Rip, & Schot, 2001; Smith & Raven, 2012)analytically located at the micro-level. The niche shields nascent innovations from thestructural pressures of the sociotechnical regime (Geels & Raven, 2006; Kemp, Schot, &

4

Hoogma, 1998).Micro in this context refers to a low degree of institutionalisation which isevident in the low stability of the rules, uncertainty about development trajectories, and arelativesmallsocialnetworkbehindit(comparedtotheregime).Thesociotechnicalregimeandniches exist within a broader exogenous environment at themacro-level referred to as thesociotechnical landscape (Geels, 2004; Rip & Kemp, 1998). The landscape consists of granddevelopments and structures in technology and infrastructure, politics, cultural values andworldviews, demographical developments, ecological developments and other discoursesoccurringattheglobalscale.Thesedevelopments—whichcannotbedirectlychangedatwillbyactors—exert pressure on the sociotechnical system, influencing the trajectory of both theregimesandtheniches.ForaMLPanalysisitiscrucialtodeterminethetimeperiodandplaceoftheanalysis,e.g.theunitofanalysis.Forexampleonecouldanalysetheregimeinacityforone decade or in a country for five decades.What counts as relevant landscape (exogenousenvironment)mightdifferforeachcase.Inourcasewefocusonfinancialservicesregimeandtheriseofthemobilemoneynichefrom2000upuntil2015,withsomeconsiderationofwhathappenedbefore.ThecorecontributionoftheMLPisthatthedynamicinterplayofelementswithinandbetweenthelevelsresultsintheoccurrenceofasociotechnicaltransitions.Internaldynamicswithintheregimemaygeneratemisalignmentsamongtheelements,creatingtensionsandinstability. Inaddition, developments at the landscape and/or the niche level may put pressure on theregime, creating or exacerbating tensions and instability. These tensions create windows ofopportunityforthebreakthroughofinnovationsarisingoutofthetechnologicalniche.Sinceitsconception,thescopeandrobustnessoftheframeworkhasevolvedasitslimitationscontinuetobeaddressed,forinstance,facilitatingabetterunderstandingofthedelineationofthe levels and regimes and operationalizing regime transformation using insights from neo-institutional thinking (Fuenfschilling & Truffer, 2014; Raven & Verbong, 2007) and formalevolutionary modelling (Safarzyńska, Frenken, & van den Bergh, 2012), a more explicitconsideration of agency, power and politics in the framework by incorporating moreperspectives from institutional theory (Fuenfschilling & Truffer, 2016) and political economy(Bakeretal.,2014),andincorporatingspatial(globalandlocal)dimensionsbydrawinginsightsfrom economic geography (Coenen, Benneworth, & Truffer, 2012; Hansen& Coenen, 2015).MLP-based studies also have begun to analyse the agency of civil society actors and usersshaping transition processes (Schot, Kanger, & Verbong, 2016; Seyfang & Smith, 2007).However, issues of social inclusion and exclusion have not been systematically addressed.Although the MLP offers a useful framework for understanding socio-technical change, wesuggest that it needs conceptual renewal to account for sustainability transitions to socialinclusion. Insights from the inclusive innovation literature discussed in the next sectionmayprovideafruitfulstartingpoint.3 InclusiveInnovationThe notion of inclusive innovation has drawn the attention of scholars, governments anddevelopmentpartnersalike inrecentyears (Cozzens&Sutz,2014;Heeks,Foster,&Nugroho,2014; IDRC, 2011; Kaplinsky, 2011a; Paunov, 2013; UNCTAD, 2014; World Bank, 2013a).AlthoughthisareaofresearchisnascentaccordingtotheWorldBank(2013a),itshedslightonthe idea that innovation can be used to tackle social exclusion. In doing so, the inclusiveinnovation literaturemobilises ideas on social exclusion, poverty, inequality, and innovationsystems.

5

Fordiscussingsocialexclusion,itisimportanttoidentifywhotheexcludedare.Intheinclusiveinnovation literature the marginalised or excluded have often been associated witheconomic/income poverty, with referential terms such as the poor being used pervasively(Chataway,Smith,&Wield,2005;Cozzens&Kaplinsky,2009;G.George,McGahan,&Prabhu,2012; Heeks, Amalia, Kintu, & Shah, 2013). Poverty is measured in Purchasing Power ParitytermsasusedinMillenniumDevelopmentGoalstargetswhichclassifyindividualslivingbelow$1.25 per day as absolutely poor, and those living above $1.25 but below $2.50 per day aspoor. The term Base of the Pyramid (BoP) has also been adopted in reference to thesepopulations (Kaplinsky, 2011b). However, as Sen (2000) asserts, poverty and by extension,exclusion,aremultidimensionalconstructs,citingSilver’s (1994,p.541) listofthingsthattheexcluded are unable to access: “a livelihood; secure, permanent employment; earnings;property, credit,or land;housing;minimalorprevailingconsumption levels;education, skills,andculturalcapital;thewelfarestate;citizenshipandlegalequality;democraticparticipation;public goods; the nation or the dominant race; family and sociability; humanity, respect,fulfilment and understanding”. Social exclusion therefore has economic, social, political,individual,spatialandgroupdimensions.New forms of innovation are emerging in developing countries to cater to the welfare ofcommunities experiencing these exclusions. Such innovative activities are referred to asinclusive innovation,butalsobyvariousterminologies.Foranoverviewwerefertotable3.1.Because Heeks et al. (2013) capture of broad variety of these various types in a Ladder ofInclusiveInnovation(LII)wedecidedtousethisapproachforouranalysis.TheLII(depictedinFigure 1) conceptualises the following steps through which greater levels of inclusivity areachievedthroughinnovation:

• Level 1/Intention: an innovation is inclusive if the intention of that innovation is toaddresstheneedsorwantsorproblemsoftheexcludedgroup.

• Level 2/Consumption: an innovation is inclusive if it is adopted and used by theexcludedgroup.

• Level3/Impact:aninnovationisinclusiveifithasapositiveimpactonthelivelihoodsoftheexcludedgroup.

• Level 4/Process: an innovation is inclusive if the excluded group is involved in thedevelopment of the innovation. This step can be broken down further to gaugeinclusion of the excluded groups in each of stages of the innovation process, i.e.invention, design, development, production, distribution. Heeks et al. (2013)complicatethisstepfurtherbyconsideringthelevelofparticipationofexcludedgroupsintheinnovationprocessrangingfromexcludedgroupsjustbeinginformedabouttheinnovativeprocess,totheexcludedcontrollingtheinnovativeprocess.

• Level 5/Structure: an innovation is inclusive if it is createdwithin a structure that isitself inclusive,i.e.theunderlyinginstitutionsthatmakeuptheinnovationsystemareinclusive.

• Level 6/Post-Structure: an innovation is inclusive if it is created within a frame ofknowledgeanddiscoursethatisitselfinclusive.

Eachlevelrepresentsagradualdeepeningandbroadeningoftheextentofinclusionachievedthroughinnovationprocesses.

6

Figure1.Typesofinclusiveinnovation

Producers/actorsintheinnovationprocess/siteofinnovation

OriginTargetanddrivingforce

Relatedterms Remarks Examples

Grassrootsinnovation

! Innovators/entrepreneursinthe

informalsector(Gupta,

2013)Publicinstitutions

! Communitybasedorganisations

andco-operatives(Seyfang&

Smith,2007)

! Non-profitorganisations,activists

Bottom-up Socialneed,

ideology

Socialinnovation(Seyfang

&Smith,2007)

! Associatedwithenvironmental

sustainability(Seyfang&Smith,

2007),‘ecologicalethics’(Guptaet

al.,2003)

! Innovatinggroupsaresmall,low-

profile,voluntary,citizen-led

(Chanan,2004)

! Drawfromtraditionalknowledge

(Guptaetal.,2003)

! TheHoneyBee

Network(India)

Socialinnovation ! Communities,localgovernments,

universities,firms

! Public-privatepartnerships

Top-down&

Bottom-up

Socialneed,

ideology

Socialtechnologies

(Fressoli,Smith,&Thomas,

2011;Hanlin&Muraguri,

2009);Social

entrepreneurship,social

ventures

! Innovationgearedtowards

problem-solvingandsustainable

businessmodelsratherthan

profitability

! ‘social’usedastheoppositeto

‘economic’(Cozzens&Sutz,2014)

! Themicrofinance

model(Grameen

Bank,Bangladesh)

! ArogyaGhar

ClinicsforMass

Care(India)

Pro-poorinnovation

! Firms:largefirms,

multinationals/transnationals

! Universityresearch(Alzugaray,

Mederos,&Sutz,2012)

! Public–privatepartnerships

(Chataway,Hanlin,Mugwagwa,&

Muraguri,2010)

Top-down

(Bottom-up

too?)

Profit ‘Belowtheradar’

innovation(Kaplinsky,

2011b);Bottomofthe

pyramid,BRI(Cozzens&

Sutz,2014)

! Refertoinnovationsforandby

thepooringeneral.Thereisno

clarityintheliteratureabout

howthistermisused.

Frugalinnovation ! Westernmultinationals(Agarwal

&Brem,2012)

! MNCsinemergingeconomiese.g.

China,India(Zeschky,

Widenmayer,&Gassmann,2011)

Top-down Profit

(low-cost

manufacturing)

Frugalengineering,

Reverseinnovation,

Constraint-based

innovation,Cost

innovation,Jugaad(Rao,

2013;Zeschkyetal.,2011);

Disruptiveinnovation

(Christensen,1997)

! Westerncorporationshavebeen

overtakenbycorporationsin

emergingmarketsthatunderstand

theresource-constrainedcustomer

better(Agarwal&Brem,2012)

! Frugalinnovationproducts

originallydesignedfordeveloping

marketshaveenteredmarketsin

industrialisedcountries(Zeschkyet

al.,2011)

India’sTataNanocar;

ChineseMiniMagical

Childwashing

machinebyHaier

Informalsectorinnovation

! Entrepreneursintheinformal

sector(MicroSMEs)(Cozzens&

Sutz,2014;Kraemer-Mbula&

Watu,2010)

Bottom-up Profit ! Typicallyuserecycledmaterials

! Characterisedbyclusters

! Aknowledgemixbetween

indigenousknowledgeandforeign

knowledgefromformalsector

(Cozzens&Sutz,2014)

JuaKalienterprisesinKenya(McCormick,

1987),Suame

MagazineinGhana

(Mytelka&Farinelli,

2000)

7

Figure2.Theladderofinclusiveinnovation(LII)depictingdifferentlevels(orgradations)ofinclusion.Source:Heeksetal.(2013).

Inthenexttwosectionswewillanalyseourcase-studyfrombothaMLPandaLIIperspective.

3. Caseanalysisfromthemultilevelperspective

Pre-Developments;theestablishmentofaformalmoneyregimeinKenya

One of the most significant landscape factors that influenced the development of the Kenyanfinancialservicesindustrythroughoutthe20thcenturywastheestablishmentofthecolonialstateinthe1890s.TheBritishcolonialgovernment introducedformal financialservices intheformofcommercial banking by foreign multinational banks, a currency system based on foreigncurrencies4asthemediumofexchange,andalegalcodegoverningthesectorincludingregulationthrough a state-owned Central Bank (Ochieng’ & Maxon, 1992). Until then, native Africancommunities relied on an already well-established barter trade system, and only moved on tocowrie shells, preciousbeads and cloth later in the19th century. Since theestablishmentof theBritishprotectorateuntiltheattainmentof independencein1964,formalfinancialservicesweredominated by British banks concentrated in major towns, focused on financing trade, andexclusively served the settler immigrant community. Apart from the Post Office Savings Bankwhich accepted African savings since 1910, these banks had virtually no involvement with thenativeAfricanpopulation.Thesedevelopmentsbirthedaformalfinancialservicesregime.

4FirsttheIndiancurrencytherupeewasadopted,anditwasreplacedwithpoundsterlingasthesovereignlegaltenderon1April1906byorderofthecolonialstate

8

The formal regime started todiversifywith theentryofnewmarket actors in the1960s:banksfromSouthAfrica, IndiaandTurkeywhichexpandedthescopeofservicetoAfricanfarmersandbusinessmen from whom they only accepted deposits but were reluctant to advance loans(Maxon, 1992). Nevertheless, the exclusionary nature of formal financial services was alreadydeeply entrenched into the sector, and efforts by the new Kenyan government to reform theinherited regime after independence by taking control of the monetary and fiscal policy,introducing a Kenyan currency and establishing state-owned community banks or buying outforeignbanksyieldedsluggishresults.In themeantime,nativeAfricansdeveloped informalmechanisms tomanage fiatmoney.As theindigenouswayoflifewasinherentlycommunal,correspondingculturalvaluesandpracticeswerecarried forward into currency management, particularly the reliance on chamas, or informalgroups—whichBouman (1983,p.268)describedas “thepoorman’sbank”—to facilitate saving,borrowing and insurance. The semi-coherent networks of informal financial groups andhouseholds, and the informal indigenous institutions that governed their behaviour constitutedthe informal regime.Nomadicandremotecommunitieshowever remainedcompletelyexcludedfromeithertheformalorinformalregime.ThingsstartingtochangeinKenyainthe1980sduetotheemergenceofthemicrofinancetrend.The international donor community responded by supporting non-governmental microfinanceagencies such as the Kenya Rural Enterprise Programme (K-REP), KenyaWomen Finance Trust(KWFT)andFauluKenya(Hospes,Musinga,&Ong’ayo,2002),anddevelopedandtestedvariousmicrofinance models. This microfinance movement debunked the myth that the poor areincapable of saving and making investments due to erratic and low-volume incomes. Anotherimportant landscape development—the launch of the Millennium Development Goals in 2000mandatinggovernmentsand theprivate sector toaddresspovertyand social exclusion,exertedpressureontheformalregimetoextendservicestothepoor,andmotivatedfurtherexpansionofmicrofinanceoptions.By the startof thenewmillennium, the formal regimehadsedimentedaroundmarket-orientedcommercialbankingastheprimaryguidingprinciple.TheCentralBankofKenyastrictlyregulated46bankslocatedincitiesandlargetownsbasedonregulatorystandardsfocusedonefficiencyandcompetition, financial risk management and sector stability. These standards represented corerulesoftheformalregime.Theformalregimewassupervisedbyvariousglobalregulatorybodies,among them the Swiss-based Basle Committee of Banking Supervisionwhich promulgated newminimum capital requirements for banks in 1988. Such landscape-level regulatory forcesinfluenced national and firm-level strategies for providing financial services, e.g. by limitingavenues for serving low incomemarket segments due to the (then perceived) high risk profilesthey carried. These developments also favoured industrialised economies and wealthierhouseholds indevelopingcountries,andasaconsequence,poorcountriesandhouseholdswerefurtherdisenfranchised.Landscapedevelopmentsinglobalfinancewereaccompaniedbydevelopmentsininformationandcommunicationtechnologies.Earlyinthe2000s,thelaunchoftheWorldWideWebandspreadofinternetconnectivityintroducednewtransactionalprotocolsinformalfinancialservices,e.g.ATMnetworks, internetbanking, creditanddebit card facilities,and real-time interbankconnectivity.These services entrenched the use of electronic payment mechanisms in the sector, which inKenyaculminatedwiththeintroductionofthereal-timegrosssettlementsystem(RTGS)in2005,anincrementalimprovementoftheexistingfundstransfersystem.Asformalfinanceincreasinglyreliedon information technology,expansionof thebankingnetworkbeyondurbanareaswhere

9

supportinginfrastructuresuchaselectricityandbroadbandinternetexistedbecamedifficult.The2006 FinAccess Survey reveals that the usage of formal financial services was twice as high inurbanareas(at32.0percent)asinruralareas(14.6percent).Formalfinancewasalsousedlargelybyhighandmiddle incomehouseholdsrelativetopoorerhouseholds inthe informalregime.23percentofunbankedindividualsattributednothavingabankaccounttotheirinabilitytoaffordtomaintainit.Despite technological, policy,market and cultural lock-ins in the sociotechnical system, tensionswithintheformalregimebeguntoappear.Thepressuretoprioritisepovertyalleviationandsocialinclusion arising from the MDGs, the microfinance movement and the national developmentstrategy Vision 2030 weighed on the Kenyan government, and by extension, the Ministry ofFinanceandCentralBankofKenyatoprioritisefinancialinclusion.Extendingformalfinancetothepoor presented a significant regulatory dilemma: balancing potential institutional innovationsagainst financial sector stability and security pursuant toBasle III requirements. Further tensionrosefromtheinabilityofcommercialbankstorespondtothenewmandatetoservethepoorduetobusinessmodelsthatcouldnoteffectivelysupporthigh-volumemicro-transactionsandspatiallylimitedinfrastructure.Newlandscapedevelopmentspoisedtotakeadvantageofthewindowofopportunitycreatedinthesysteminsteadsprangfromthetelecommunicationssector. Intheperiodbetween1998andtheearly2000s,thetelecommunicationsindustryinKenyahadundergonesubstantialreformthatstimulated the entry of and competition among several mobile operators into a previouslymonopolistic sector. As a consequence the cost ofmobile phone services dropped significantly,andbytheendof2006thenumberofmobilesubscribershadgrownto9.3million(CCKAnnualReport,2007-2008)5.Competitioninthesectorstimulatedinnovation,andoneofthevalue-addedservices6introducedintothesectorwasairtime-sharingservices:Me2youfromZainKenyaLtdandSambazafromSafaricomLtd7.Theseairtime-sharingservicesareseenasaprecursortothemobilemoneyplatformsubsequentlydevelopedinthesameyear,ascustomersusedtheseservicesasarudimentarymoneytransfersystem(Jack&Suri,2011).Thebeginningsofthemobilemoneyniche(2003-2007)

InresponsetotheMDG’scall forprivate-sectorparticipationinpovertyalleviationindevelopingcountries,VodafoneUK—whichpart-ownsSafaricomLtd,thelargestmobileoperator inKenya—ventured todevelop a technological solution to financial exclusion.Vodafone in 2003wona £1millionmatchedawardfromtheFinancialDeepeningChallengeFund(FDCF),aprogrammebytheU.K. government’s Department for International Development (DFID). The establishment of amobile money niche was marked by the start of open brainstorming workshops in 2003,spearheadedbyVodafoneinNairobiandDaresSalaam.Theworkshopswhichinvolveddifferenttypesofactors8wereusedasopportunitiesforlearningaboutlocalchallengesinfinancialservices.Thenetworkofactorseventuallygravitated intoapartnershipbetweenVodafoneUK,SafaricomKenya Ltd, Faulu Kenya—a microfinance organisation previously involved in developing newmodels of informal finance, and Commercial Bank of Africa. At the outset, expectations were59.3millionmobilesubscriberswasabout56%oftheadultpopulation.SafaricomLtdcontrolled72%ofthemarket6Thecorebusinessofmobileoperatorswasvoice,SMS,andeventuallydata(mobilebroadband).Otherservicesnotcategorisedintheseclasseswerereferredtoasvalue-addedservices.7Kamau,Macharia2010,‘Zainwoossubscriberswithvalueaddedservices’,TheStandard,10August.Availablefromhttp://www.standardmedia.co.ke.[28October2013].8Participatingactorsincludedbanks,microfinanceorganizations,technologyservicessuppliers,nongovernmentalorganizationswithaninterestinmicro-credit,andrepresentativesfromthetelecomsandfinancesectorregulators

10

articulated towards providing solutions for microfinance loan repayment. Several productpropositions were considered and abandoned due to feasibility issues, among then the use ofPoint of Sale systems andmagnetic stripe cards to facilitate payment. Eventually, expectations“firmeduparoundthedesignandtestofaplatformthatwouldallowacustomertoreceiveandre-payasmallloanusinghisorher[mobile]handset”,accordingtoaVodafoneexecutive(Hughes&Lonie,2007,p.68).Additionally, thenoveltyhadtobe frugalandeasilyaccessibleamongthepoor.AnSMS-basedapplicationnamedM-Pesawasdevelopedbasedonbasicmobilephones.M-Pesawasthenpilottestedamonglow-incomeusersoveraperiodof7monthsbetweenOctober2005andMay2006.ThetestresultedinnumerousproductimprovementsmadebeforenationallaunchinMarch2007.Intheperiodleadinguptothelaunch,VodafoneandSafaricomnegotiatedM-Pesa’s fit with prevailing regulatory rules, for instance, its compliance with fraud and anti-moneylaunderingguidelines,andcustomerdepositprotection.M-Pesa was launched in 2007 under the key customer proposition Send Money Home, whichexploitedthenormofmakingfrequentremittancesthroughfamilialsocialnetworks.Subscriptionsreached1.3millionusersbytheendoftheyear.Thisrapidgrowthoccurredbothintheformalandtheinformalregimes,andasaresult,M-Pesabeguntoformalisemoneytransfersoccurringacrosstheregimes.Actorsandinstitutionsintheboththeformalandtheinformalregimesreactedtothelaunch anddramatic uptake of the service in differentways, creating newdynamicswithin andacrossregimes.SOUGHTFROMCBKHollowingoutoftheformalandinformalregimes(2007–2013)

Between 2007 and 2013, M-Pesa subscription grew from 5.1 million to 17.1 million registeredusers, supportedbyanagentnetworkof65,547 (Safaricom,2013).Thebusinessmodeldiffusedacross the telecommunications sector in Kenya as other mobile operators entered the mobilemoneymarket:ZapbyZainKenyain20099,YuCashbyEssarin2009,IkoPesabyOrangein2010andTangazaPesabyMobilePay Ltd in2011.Cumulatively,mobilemoney subscription reached26.2millionregisteredsubscribersin2013,i.e.about82percentoftheadultpopulation.However,Safaricom’sM-Pesacontinuedtodominatethemarket.By2013,17,000agentssupportedmobilemoney, surpassing the total number of Postbank branches, post offices, bank branches andautomated teller machines by nearly five times (Mas & Radcliffe, 2011). An equivalent of 15percent of Kenya’s gross domestic product (GDP) worth of person-to-person transactions wasconductedviamobilemoney.DuetothebreakthroughofM-Pesa,hollowingoutatthemeso-levelmanifestedthroughdramaticdisturbances in market segmentation across the formal and informal regimes, preferences andbehavioural norms among the banked and the unbanked, banking sector policy in the formalregime, innovation agendas in several sectors in theeconomy, and the general configurationofsocialgroupsandnetworksofactorsinthesystem.With respect to market segmentation, it is evident that M-Pesa adoption brought about aredistribution of households across the formal and informal regimes, and a blurring of theboundaries between these regimes (see Figure 2 for an illustration). Proportions of householdsincorporating mobile money into their financial management practices increased. As aconsequence, households exclusively relying on either formal or informal services reducedprogressivelysuchthatby2013,thosethatusedformalservicesonlydroppedfrom12percentto3percent,while informalonlydroppedfrom33percentto9percent.Participationintheformalregime however increased from 28 percent in 2006 to 64 percent in 2013 as more unbankedpeopleopenedbankaccounts,henceexpandingthescopeoftheregime.Incontrast,participation

9RebrandedAirtelMoneyin2011

11

in the informal regime decreased from 49 percent in 2006 to 42 percent in 2013 as peoplemigratedfromchamastomobilemoneyandformalfinancialservices,resultingintheshrinkingoftheinformalregime.Ininterviews,membersofinformalgroupsreportedsubstitutingM-Pesaforregular chamameetings, andborrowingdirectly from friends and family on theirmobile phonenetwork rather than from the chama. Additional multi-regime dynamics are evident throughblurringregimeboundaries.Householdsincreasinglystraddledfinancialservicesavailableinboththe formal and informal regimes. For instance, there was an increase in households thatsimultaneouslyusedbankaccounts,chamas andmobilemoneyservices (29percent),and thosethat combined bank accounts and mobile money (30 percent). Interestingly, by 2013 only 2.2percentofhouseholdscombinedmobilemoneyandchamas.Proportionsofabsolutelyexcludedhouseholdsreducedfrom39.3percentto25.4percent.

Figure3.Areagraphpresentingdataontheshiftintheuseoffinancialservicesbeforeandafterthelaunchofmobilemoneytechnologyin2007.Whiletheproportionofexcludedusersreducesprogressivelyin2009and2013,thereisagradualerosionofboundariesbetweentheformalandinformalregimesasusersadoptmobilemoney.Householdsincreasinglyparticipateinmultipleregimes.Source:OwncomputationsfromFinAccessNationalHouseholdSurveys

In the period immediately afterM-Pesa launch (2007-2009) however, competition between thenovelty and incumbent banking platforms was fierce. Growth in commercial banking wasnegligible, a phenomenon that incumbent banks reacted to aggressively especially when theircustomers started to divert funds from their bank accounts to their mobile money accounts(Kendall,Maurer,Machoka,&Veniard,2011).Asaconsequence,M-Pesafacedcontestationfromcommercial banks which cited fraud andmoney laundering as risks to financial sector stabilityposedbyM-Pesadiffusion.AnarticleinTheNairobiStar,alocalnewspaper,headlined“BigBanks

Formalonly

InformalonlyMobileonly

Formal&Informal

Formal&Mobile

Informal&Mobile

Formal&Informal&Mobile

Excluded

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2006 2009 2013

Prop

or<on

ofu

se

Year

Householduseoffinancialservices

2007M-Pesalaunch

12

PlottoKillM-Pesa”10reportedthatfourlargebanksformedanadhoccommitteetotryandgetM-Pesastopped,pitchingtotheMinisterofFinancethatM-Pesaissimilartoaponzischeme.Atthesametime, thebankingsector lobby—KenyaBankersAssociation (KBA)—pressuredCBKtoauditthelegalstructureofM-Pesawiththeexpectationthatitwouldberegulatedundertheprudentialguidelinesforcommercialbanking.Suchregulationwouldeffectivelyhomogeniseitwithbanks.TheMinisterofFinanceorderedCBKto“studytheschemeandpronouncepolicytosafeguarddepositors”,stating:“Wewanttoprotect

wananchi from the sharks who want to makemoney from the misfortune of others”11, further

creatinguncertaintyaboutM-Pesaattheregimelevel.CBKandtheNationalTreasuryproceededtoconductanin-depthauditandlegalevaluation.TheauditestablishedthatM-Pesaserviceswerewas distinct from banking since 1) M-Pesa does not accept deposits that are repayable ondemand,oracceptmoneyforcurrentaccountpurposesthat isusedforpaymenttocheques,orlendorinvestcustomerfundsandthereforenorelatedrisksareinherent;2)customerfundsareheld ina trust inoneof the leadingbanks;3) therearenocredit risksasM-Pesaagentsprepayfundstoreplenishtheirfloatsbeforetransactingwithcustomers.Further,CBKandtheTreasuryina public circular committed to regularly monitor risk factors associated with the platform, anddeclaredM-Pesa as trustworthy. In so doing, CBK quelled contestation from incumbent formalregime actors and restored public confidence in mobile money services. Subsequently, CBKadopted an “experiment first, then regulate” approach (Mas & Ng’weno, 2010, p. 355), andexpandeditsmandatetocreateanenablingenvironmentthatwelcomesinnovationinthebankingsector.Whilelaunchingamobilephonebankingsolution,theCBKgovernor“urge[d]thebankingsector to seek innovative ways of leveraging on existing technologies and infrastructures toprovideaffordableandinclusivefinancialservicestoKenyans”12.Inordertoremainrelevant,commercialbanks13redesignedtheiraccountstobecomecheapertooperatebyloweringoreliminatingaccountopeningcosts,transactioncostsandledgerfees,andundertook a wave of branch opening in new rural and low-income locations where M-Pesaadoption had demonstrated a market potential (Johnson, 2012 Financial landscapes). By 2009,data from the FinAccess survey showed that 75 percent of the urban population cited mobilemoneyagentsastheclosestfinancialserviceprovider,ratherthancommercialbankbranchesorATMs.A significant landscape shock that cementedM-Pesa’s legitimacy in theevolving formal and theinformalregimeswasthe2007-2008post-electioncrisis inKenya.Chronicpolitical instabilityandcivil unrest in January 2008 followed the disputed presidential and parliamentary election. Theunrest caused the shutdownof all formal financial services for oneweek, resulting in a chronicshortage of cash in urban areas and in violence-ridden rural areas in the country.M-Pesa as amobile phone-based service provided the only channel for money transfer to isolated anddisplacedcommunitiesinruralareasandtobesiegedurbanareas.ForthefirsttimesinceM-Pesalaunch, the volume of transfers from rural to urban areas rose sharply as rural families made

10Mbugua,J.:Kenya:BigBanksinPlottoKillM-PESA.TheNairobiStar(December23,2008),http://allafrica.com/stories/200812230962.html(accessed01/02/2017)11Wahome,M.:M-PESAunderinvestigation.TheEastAfrican(December9,2008),http://www.theeastafrican.co.ke/news/2558-500218-v0rfibz/index.html(accessed01/02/2017).Atthetime,pyramidschemeshadproliferatedinKenyaandmanypeoplehadlostmoneyintheseschemes.M-Pesafacedtheriskofbeingbundledtogetherwithpyramidschemesbothfromapolicyanduserperspective.12Ndung’u,N.,GovernorofCentralBankofKenyaremarksattheLaunchofEquitelBankmobilebankingservice,September3,2008.13EquityBank,FamilybankandCooperativeBank

13

remittancestosupporttheirurbancounterpartsduringthecrisis(Morawczynski&Pickens,2009).Therecessionthatfollowedtheelectionbroughtaresurgenceof informalfinance inruralareas,where newly formed informal groups served a dual purpose: communal support for displacedfamiliestoestablishsmallbusinessesasasourceoflivelihood,andtoaidreconciliationeffortstoquellpoliticaltensionsincommunities(Langat,2014).ManypolicychangesoccurredintheformalregimeinperiodafterM-Pesalaunch,andtherewasaclearshiftinpolicyprioritiesfromaheavyfocusonfinancialsectorstabilityandaminorfocusonfinancial inclusiontoabalanceof thetwo.CentralBankofKenyaspeededupthereleaseof theMicrofinance Act (2006) to regulate credit services and payments occurring within themicrofinance sphere inwhichM-Pesawas pilot-tested. It also exempted value added tax on allfinancial services in2009to further lowercustomercosts14.Newagencybankingguidelines thatwould enable banks operate an agent network similar to M-Pesa’s networks were released in2009.Additionally,araftofnewlegislationandregulatorymeasuressettingparametersformobilebanking, regulating electronic money transfer, instituting anti-money laundering guidelines andrestructuringthenationalpaymentssystemweredebatedinpublicandinparliament,andpassedinto law. Two landscape factors influenced this tightening of regulatory frameworks in thefinancialservices industry inKenya:the2008global financialcrisis inwhichstricterregulationattheglobalandlocallevelwaswidelyseenasaremedy,andtheissuanceofBaselIIIPrinciplesforSound Liquidity Risk Management and Supervision in 2009 which recommended new liquiditylevelsinformalfinancetoensurefinancialstability.From 2009 onwards, new networks of actors centred on M-Pesa and other mobile moneyprovidersformedinboththeformalandinthe informalregime.First,commercialbanksformedallianceswithmobileoperatorstointegratemobilemoneyservicestotraditionalbankingservices.ThemostcelebratedalliancewasbetweenSafaricom’sM-PesaandEquityBank,thebankwiththelargest customer base in Kenya. This partnership—referred to as “the financial inclusion holyalliance”byGSMA15duetotheirfocusontheunbanked—launchedM-KESHO,acobrandedsuiteof financial products16. By 2013, all commercial banks had integratedmobilemoney services intheirproducts.OthernotablealliancesincludedM-PesaandVisaandMastercard,andM-PesaandWesternUniontofacilitateinternationalmoneytransfers.Atthenichelevel,thecombinationoflandscapepressuresforsocial inclusionandthesuccessofM-Pesa at the regime level—which yielded positive expectations on mobile phone-basedsolutions—gave rise to a deluge of foreign and local investments into mobile phone-basedinnovationinKenya.Newpay-as-you-goinitiativesexperimentingonnewformsofsocialinclusionusingM-Pesaresultedintheformationofnewandlargernetworksofactors.Onesuchinitiativewasthenot-for-profitGrundfosLIFELINKwaterproject17managedbyDanida,theUnitedNations,thegovernmentofKenya,WorldVisionandRedCross.40demonstrationprojectsimplementedinKituiDistrictin2009exploredsustainableruralwatersupplymodelsinwhichhouseholdspaidforsafewaterusingM-Pesa.AnotherinitiativeisM-KOPASolar,apay-per-usemodelforsolar-based14However,thegovernmentintroduceda10percentexcisedutyonfeeschargedformoneytransferservices,anditappliedtomobilephoneproviders,banks,andothermoneytransferagencies.Thistaxwaspassedontotheconsumer.ThevalueofM-Pesapaymentsreducedmarginallyasaconsequence.Seehttp://www.cgdev.org/blog/taxing-kenya%E2%80%99s-m-pesa-picks-pockets-poor15GSMAblogpost,M-KESHOinKenya.http://www.gsma.com/mobilefordevelopment/m-kesho-in-kenya16M-KESHOwaseventuallyabandonedbybothcompaniesduetoprofit-sharingdisagreements.http://www.businessdailyafrica.com/M-Kesho-growth-stalls-over-hitch-on-profit-sharing--/-/539552/1373474/-/8e1xlj/-/index.html17GrundfosLifelinkprojectsinKenya:http://www.grundfos.com/cases/find-case/grundfos-lifelink-projects-in-kenya.html

14

energyprovision tooff-gridcustomersusingM-Pesaandothermobilemoneyservices tocollectrevenues18. Mobile money was seen as providing the most efficient and cost effectivetransactional rail for overlayingmyriad services forwidely dispersed and geographically remotehouseholds. Further, the differentiation between similar mobile money products into moredistinct businessmodels becomes clearer: themobile network operatormodel, for instanceM-Pesa,thebank-ledmodelforinstanceEquityBank’smobilevirtualnetworknamedFinserveAfrica,andtheindependentmodelforinstanceTangazaPesabyMobilePayLtd.Dynamicstabilityofthenewregime

At the end of 2013, diffusion of mobile money services stabilised at approximately 26 millionregistered users.M-Pesa is now considered theworld’smost successfulmobilemoney transferservice19,and isoperational insevenAfricancountries. AnarticlebyTheEconomist IntelligenceUnit suggestively titled“OutofAfrica, intoRomania”20 signifies theniche-landscape influenceofM-PesaasitstartedtoberolledoutincountriesoutsideAfrica,i.e.India(2013),Romania(2014)andAlbania(2015).Furtherniche-landscapeinfluencesareapparentintheestablishmentofglobalniches where voluntary “multinational financial inclusion-oriented networks” are formed(Villasenor,West,&Lewis,2015,p.3),providingspacesforpolicymakersindevelopingcountriesto learn and share knowledge towards enhancing access to basic financial services, and setquantifiabletargetstodriveprogresstowardsfinancialinclusion.Intheseforums,mobilemoneyisconsidered central to country-level programmes, and M-Pesa’s role in Kenya’s success inextendingdigitalfinancialservicestotheunbankedisusedastheexemplar.ExamplesincludetheAlliance for Financial Inclusion—a global knowledge-sharing network for financial inclusionpolicymakersfrom89memberstates,BetterthanCashAlliance(2012)composedof14developingcountrygovernments,privatesectoractorsandphilanthropicfoundationsseekingtotransitiontoelectronic payments, and G20 Financial Inclusion Peer Learning Program (2012). Additionally,numerousexplicit and implicitmentionsofusingmobilemoney technologiesweremadeduringthe post-2015 Sustainable Development Goals summit sub-meeting Digitising Payments andInclusiveFinancetoAchievetheSustainableDevelopmentGoals(DigitalCurrencyCouncil,2015)21.Mobilemoneyiswidelyconsideredasapromisingavenueforpromotingfinancialinclusioninthedevelopingworld.Attheregimelevel,integrationoftheformalandtheinformalregimescontinues.Thebankedareforming chamas to manage their investments where they are adopting more sophisticatedinvestmentinstrumentssuchasrealestate,thestockmarketandoffshoreopportunities.Informalgroupsontheotherhandcontinuetoformalisetheiroperationsbyrelyingonmobilemoneyandformalbankaccounts.Mobilemoneysupportsseamlessremittancesacrosspre-existingfinancialdevices. As this is an on-going transition, the potential for incremental innovation and furtherreconfigurationstillexists.

4. CaseanalysisfromtheInclusiveInnovationPerspective

18M-KOPASolar:http://solar.m-kopa.com/about/company-overview/19InaBrookingsFinancialandDigitalInclusionscorecardthatcoveredmultiplecountries,Kenyascoresthehighest,i.e.89percentoftotalpossiblepoints(Villasenor,West,&Lewis,2015)20TheEconomistIntelligenceUnit.OutofAfrica,intoRomania.(May13,2014)http://www.eiu.com/industry/article/1451806929/m-pesa-out-of-africa-into-romania/2014-05-1321Babcock,L.:WhatDotheUN’sNewSustainableDevelopmentGoalsMeanfortheBlockchain?(September18,2015)https://www.digitalcurrencycouncil.com/professional/what-do-the-uns-new-sustainable-development-goals-mean-for-the-blockchain/

15

InthissectionwerecasttheM-PesacasethroughtheLII,highlightinghowdifferent(higher)levelsofinclusionwereachieved.Level1/InclusionofIntention

M-Pesa was born out of the aspiration to foster financial inclusion by deepening the financialsector. By 2006, only 26 percent of the Kenyan adult population had access to formal financialservicesprovidedbycommercialbanks,microfinanceinstitutions,SACCOsandotherprudentiallyregulated institutions, while 32 percent used informal financial services. 39.3 percent wasabsolutely excluded from any form of financial service, formal or informal. Thus, the unbankedmadeup74percentofthepopulation.OuranalysisofdatafromtheFinAccessNationalSurveysrevealsthattheunbankedwere located in low-incomeurbanareasandruralareaswhere,apartfromadeficiencyinformalfinancialservices,therewasadeficiencyinotherinfrastructuresuchasgridelectricity, pipedwater andmobility systems. In addition, theunbanked consistof ahigherproportionofhouseholdsrunbywomen,andthesehouseholdsreliedonsubsistencefarmingorwere dependent on financial or resource contributions frommembers in their social networks.Theirmedianmonthlyincomerangedbetween$11and$30,andtheyhadrelativelylowerlevelsof educational attainment. Figure 3 summarises these multidimensional characteristics of theunbankedvis-à-visthebanked.

46%54%60%

40%

Male Female

Gender

Unbanked Banked

12%

54%

35%

4%

47% 49%

Mostvulnerable Vulnerable LeastVulnerable

Vulnerability

Unbanked Banked

7%

13% 12%

5%9%

3% 1%

6%

17% 17%

10%

29%

15%

6%

Educa<onlevels

Unbanked Banked

22%28%

49%59%

18% 23%

Havemobilephone

Usefamily/friend'sphone

Noaccesstophone

Accesstomobilephone

Unbanked Banked

16

Figure4.Bargraphsdifferentiatingthecharacteristicsoftheunbankedvis-à-visthebanked.Vulnerabilityreferstothefrequencywithwhichhouseholdsgowithoutenoughfoodtoeat:themostvulnerableoftengowithout food.Data forgender,education levelsandagegroupsarefromthe2006FinAccessSurvey.Dataonlivelihood, incomegroupsandvulnerabilityarefromthe2013FinAccessSurvey.While their lack of access to banking services was in itself harmless to households, and somehouseholdsdidnotconsidersuchservicesas inherently important22,such instrumentalexclusionled to capability deprivations that prevented them from optimally participating in economicactivityatthelocalandnationallevel,andlimitedtheirownopportunitiesforincomegrowth.Thisexclusionwas furtherentrenchedbyprudential regulations fromtheCentralBankofKenya thatgavelittleroomforcommercialbankstoextendbankingproductsforthepoor.Asaconsequence,thesecommunitieswerepassivelyexcludedbypoliciesthatinadvertentlylimitedparticipationofthepoorinformalfinancialservices.Vodafone UK undertook an initiative to provide a solution for the unbanked alternative tocommercialbanking.TheiraspirationtoaddresssocialexclusioninfinancialservicesisevidencedbytheirefforttoacquirecapitalfromtheU.K.government’sDFIDFinancialDeepeningChallengeFund(FDCF)aimedatfinancialinclusionprojects.Vodafone’sHeadofSocialInnovationdescribedthe target population segment as “customers who were unbanked, unconnected, often semi-literate, and who faced routine challenges to their physical and financial security” (Hughes &Lonie,2007,p.69),categoricallystatingthat“wewerespecificallytargetingtheunbanked”(p.69).The outcome of this initiative was M-Pesa, a mobile phone based money transfer platformlaunchedinKenyain2007.Intheperiodleadinguptoandimmediatelyafterthelaunch,M-Pesafacedobjection from commercial bankswhichdemanded its regulation as a financial institutionundertheBankingAct.TheCentralBankofKenyahowevershieldedM-Pesadevelopmentandroll-out,citingfinancial inclusionasthehigher immediateobjectiveovertheinterestsofthebankingsector.Thereforeitisclearthatfromtheoutset,M-Pesawasdevelopedwiththeintentiontoincludethemarginalised. As we shall see in the next section, the business model and the platform weredesignedanddevelopedwiththeneedsoftheexcludedinmind,thusconfirmingthattheintenttoincludewastheoverarchingguidingprinciple.Level2/InclusionofConsumption

To ensure that the platform was accessible and user-friendly for the unbanked as the targetmarket, the developers took into account their social-economic circumstances, lifestyle and

2262%oftheunbankedbelievedthat“youcaneasilyliveyourlifewithouthavingabankaccount”,while47%believedthat“bankstakeadvantageofpoorpeople”

6%

26% 24%16% 15%

7% 4% 1% 1%1%

13%21% 19% 22%

13% 9%2% 1%

MonthlyincomeUnbanked Banked

60%

10% 9%19%

3%

58%

18% 12% 10%2%

Agric EmployedOwnBusinessDependent Other

Sourceoflivelihood

Unbanked Banked

17

environment. Theunbankednormally usedor had access to a basic SIM-basedmobile handset,lived greater distances from brick-and-mortar bank branches23, spoke Swahili and vernacularlanguages, and normally conducted high-volume, low-value transactions. Therefore, anytechnological solution needed to account for these factors. M-Pesa was designed as a simple,menudriven,SMS-basedapplicationthatwouldrunontheSIMtoolkit,available inbothEnglishandSwahili,tobesupportedbysmalloutletsandkiosksasagentsintradingcentresinruralareasandlowincomeurbanareas(Hughes&Lonie,2007).Withinthreeyearsoflaunch,theservicehadover10.3millionregisteredusers.Halfofthe17,700M-Pesa agents in operation were located outside urban centres. Notably, M-Pesa was initiallywidely adopted by the banked during the Send Money Home marketing campaign run bySafaricomLtdtostimulatecashtransfersfromurbanhigh-incomeindividualstotheirrelativesandfriends in ruralareas.However, theservicespreadtopoorerhouseholds in remoteareasoverashortperiodoftime.JackandSuri (2014) intheirsurveyof3,000randomlyselectedhouseholdsacross Kenya found that their proximity to mobile money agents grew fourfold betweenSeptember2008andDecember2009.Data fromthe2015FinAccessGeospatialMappingSurveyreveals that 73 percent of Kenyans livedwithin a three kilometre radius of a financial servicesoutlet,and78percentoftheseoutletsaremobilemoneyagents.Jack and Suri (2011) show thatM-Pesa’s reach down the socio-economic spectrum is reflectedalong a number of dimensions. First, the share of M-Pesa use among unbanked householdssteadily increased, doubling from 25 to 50 percent between 2008 and 2009. Second,while therepresentationof all segmentsof the incomedistribution in theprofile of users has grown, theproportional growthhasbeenhighestamong thoseat thebottom. Third, the representationofruralhouseholds,typicallypoorerandlessintegratedintotherealeconomyalsoincreased.Whileapredictablyhighshare, three-quarters,ofurbanhouseholdsusedM-Pesa in2009,theshareofruralhouseholdsusingtheserviceagaindoubledfromjust29percentin2008tonearly60percentin2009.Similarpatternswereobservedwith regard toeducationalattainment:bettereducatedpeople weremore likely to useM-Pesa, but growthwas higher among the less-well educated.Finally, M-Pesa reach among women increased, such that the proportion of female users rosefrom38percentin2008to44percentby2009.JackandSuri(2011)furtherarguethatM-Pesa’spatternoftechnologyadoptionmirrorsthatofotherproductandservice innovations,whichareoftenfirstusedbythebetter-off.However,thespeedatwhichtheservicehasreachedlesswell-offusersandtheirapparentlyhighvaluationof it isunprecedented,especially in thedevelopingworld.Dueto itsadoptionamongtheunbankedand ituseamongthebanked,M-PesahascontributedtangiblytowardspromotingfinancialinclusionanddeepeningofthefinancialsectorinKenya.Level3/InclusionofImpact

AsM-Pesadiffusedamongthebankedandtheunbanked,itsimpactsbeguntobefeltacrossbothdomains.Economicimpactsamongruralcommunitiesandtheurbanpoorareespeciallypalpable.Firstly, theuseofM-Pesaminimised thevulnerabilitiesandnegative impactsof financial shocksarising fromevents suchas crop failure, illness, job lossor violence thatpoorerhouseholds areparticularly susceptible. M-Pesa use enabled such households to smooth consumption throughaccess to creditand insurance fromtheir socialnetworkson theplatform (Jack&Suri,2014).AFinancial Diaries study revealed that M-Pesa enabled poor households to make micro-savingsmoreeffectively, andplan for short-termconsumptionneeds inamannerpreviouslydifficult toachieve through mattress-banking (Zollman, 2014). The same study also showed that mobile

23Inthe2006FinAccessSurvey,68%oftheunbankedreportedthatthenearestbankwas‘veryfar’away

18

money use reduced the proportion of low-risk, low-return assets held by households forprecautionary purposes such as jewellery, enabling them to invest in potentially higher risk buthigher returnassets suchasa smallbusinessoragriculturalequipment,withpotentiallyprovideoverall long-term income enhancing impacts. In addition, M-Pesa enabled such households todiversifysourcesofincomee.g.fromremittancesfromfriendsandrelativesthatwouldotherwisebe difficult to access due to lack ofmoney transfermechanisms (Malkamäki, Johnson, &Nino-Zarazua, 2009). Low-income households reported income increased by asmuch as 30 percentfrom transfers from people in their social networks (Morawczynski & Pickens, 2009). Anotherstudyshowedthatduetohigherdisposableincome,registeredusersofM-Pesahadaveragedailyexpenditure67percenthigherthanthosenotusingtheservice(McKay&Pickens,2010).Astudyconducted inKibera, the largesturbanslum inKenya,andMurangaandKitui, tworuraldistricts, revealed additional economic impacts ofM-Pesa: the expansion of the local economythroughincreasedmoneycirculationandlocalemployment;increasedphysical,financialandfoodsecurity;anenablingoffinancial,humanandsocialcapitalaccumulation;andeaseintransactionsand increasedquality control in thebusinessenvironment (Plyler,Haas,&Nagarajan,2010).M-Pesaincreasedthevolumeofmoneyflowinginandoutofthesecommunities,speedingupmoneyvelocitywhichboostedlocalconsumption.Asaconsequence,therewasanincreasedvolumeandvarietyof foodandagricultural inputs in localmarketsthatbolsteringfoodsecurity.MoreyouthandwomenfoundemploymentintherapidlyexpandingagentnetworkasbusinessesbecameM-Pesa agents in addition to their core business. In terms of physical security, users reportedreducedmuggingsandthefts,andwomenreportedbeingabletouseM-Pesatoaccumulatecashsecurity.WhiletheimpactofM-Pesaamongthepooraredramatic,M-Pesahasalsopositivelyimpactedthe‘banked’, providing them with an alternative fast, convenient money transfer and m-bankingalternateorcomplementtoformalbankaccounts.Level4/InclusionofProcess

Here we evaluate the inclusivity the M-Pesa innovation process. At the conception stage,inclusivity is evident in Vodafone’s involvement of multiple actors during the initial effort tounderstand local challenges and search for a feasible solution. M-Pesa was then conceived incollaboration with several local organisations: the local Vodafone affiliate, the microfinanceinstitutionandalocalbank.Later,participationoftheunbankedduringtheproductdevelopmentprocesshelpedtoestablishthefeasibilityofthebusinessmodelandthetechnologicalchoices.UserswerefirstinvitedintotheinnovationprocessatthefirsttrialoftheM-Pesaprototype.Theoriginaldesignpropositionwasaplatformthatenablescustomerstoreceiveandrepaymicroloanstomicrofinanceinstitutionsusingtheirmobilehandset.ThemicrofinanceinstitutionFauluKenyaprovidedthefirstsetof500pilot-testusers(Hughes&Lonie,2007).ThroughparticipationofthissubsetofpotentialM-Pesaadopters,theM-Pesateamlearnedaboutthedifferentiationofusers according tomobilephone literacy, the importanceof user training,andways tosimplify thecomplexityof transactionsbuilt into theplatform.Asusers familiarisedthemselves with the technology and transaction procedures, they creatively carried out newproceduresontheplatformnotinitiallyenvisioned,forexamplerepayingloansonbehalfofotherclientsinreturnforservicesrendered,payingforgoodsandservices,usingM-Pesaasanovernightsafeforcash,etc.ThisentrepreneurialbehaviourshowedM-Pesa’spotentialforopportunitiesforinnovation beyond the bounds ofmicrofinance. It wasmore revolutionary as amoney transfersystem than as a credit disbursement and repayment system. Consequently the microfinanceproposition and Faulu Kenya as amain actorwere abandoned andM-Pesa took a new turn to

19

remittances services. These new user applications were subsequently incorporated into theM-Pesa platform, thereby initiating an iterative process of user innovation and M-Pesa productimprovements and extensions before and after the official launch. After launch, for instance,value-addedservicessuchasbillpaymentsandsavingproductswereintroducedontheplatformafterSafaricomLtdnoticedthatuserswereusingM-Pesatomakepayments inretail storesandsavingmoneyontheplatform.ThedistributionofM-Pesathroughsmallbusinesseslocatedinlow-incomeurbanareasandruralareas acting asM-Pesa agents is a furthermanifestation of user involvement. These businesseswere run by the individuals in the excluded segment, and these entrepreneurs interactedintimatelywithpotentialusers in their localities, therebyprovidinguseful feedbacktoSafaricomduringtheroll-outofM-Pesa.ItisclearthereforethatM-Pesa’sinnovationprocesswasinclusivetotheextentthattheexcludedwere consulted and collaborated in the development of the businessmodel during the design,development, and distribution phases of the innovation process. However, as productdevelopmentwasentirelyownedbySafaricomandVodafone,andFauluKenyainitially,usershadnodirectcontrollingstakeinthetrajectoryofdevelopment.Level5/InclusionofStructure

Prior to thedevelopmentand launchofM-Pesa, the financial servicessectorwas fundamentallyexclusive,asitonlycateredtotheneedsofmiddle-andhigh-incomewell-educatedhouseholdsinurbanareas. Infact,theexcludedwerenotconsideredaspotentialactors intheformalfinancialservicessectorduetotheirlowincomelevelsandfrequentlow-valuetransactions,thusthelackofincentives to develop products tailor-made to their needs. Commercial banks limited theparticipationofthepoorthroughunaffordableminimumrequirementstooperateabankaccountsuch as minimum opening balances, the maintenance of a minimum operating balance,demonstrationofa regular sourceof income,etc.Bydesign, these requirementswerepassivelyexclusionary.

Figure5.Graphpresentingtherestructuringofmarketsegmentsinthefinancialservicesindustry.Thefiguresrepresenthouseholdsexclusivelyusingeachofthefinancialservices.Dataarefromthe2006,2009and2013FinAccessSurveys.

AfterthelaunchofM-Pesain2007,theproportionofhouseholdsabsolutelyexcludedreducedto31.3percentin2009and25.4percentin2013,whiletheproportionofthebankedincreasedfrom27.4percentin2006to41.3percentin2013(seeFigure4).Theproportionofhouseholdsrelyingexclusivelyoninformalmechanismsoffinancialmanagementreducedfrom33.3percentin2006

27.4

28.8

41.3

33.3

34.1

7.8

5.8

25.4

39.3

31.3

25.4

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

2006

2009

2013

Formal Informal MobileMoney Excluded

20

to7.8 in2013as theyadoptedM-Pesaandopenedbankaccounts.Thesenumberssuggest thatmobilemoneywasmorepervasivelyadoptedbyhouseholdsusinginformalfinancialservicesthanbytheabsolutelyexcluded.Therefore,althoughgainsonfinancialinclusionwereachievedtotheextentthat64percentofthepopulationwerenowpartoftheformalfinancialsector,thesegainswerenotequallydistributedamongthosethatwerepreviouslyexcludedfromtheformalsector.Thus,M-Pesaresultedinsubordinatedinclusion.DatafromtheFinAccess2013surveyrevealsthattheabsolutelyexcludedarethepoorestofthepoor,i.e.themostvulnerable(9percentofthepopulation)andthevulnerable(50percentofthepopulation),i.e.thosewho“gowithoutfoodoften”andthosewho“gowithoutfoodsometimesorrarely” respectively. This category of households lives below the poverty line, and often live inregionswithlimitedcoverageofbothbankingandtelecommunicationinfrastructureaspreviouslyillustrated in Figure3. Theseeconomic and infrastructural deficiencies,whichare chronic in thenorthernpartsofthecountry,aremirroredinthefinancialinclusionandexclusionmapsinFigure7. Additionally, the cost of owning and operating a basic mobile phone is prohibitive forhouseholds at the base of the pyramid, further preventing their participation inmobilemoneyservices.FinAccess2013data shows thatonly15percentofexcludedhouseholdsownamobilephone. In their studyof800households in this segment inKenya, infoDev (2012) found that20percentofpoorhouseholdsmadesacrificestosustaintheirmobilephonesubscriptionservicebydivertingasmuchasKsh72($0.71USD)weeklyfromfood,clothes,utilitiesorbusfare.Overandabove these basic subscription services, the transaction fee structure ofmobilemoney servicesmay also be unaffordable for the excluded. For instance, a $1USD cashwithdrawal onM-Pesacarriesan11percentfee,andthelowestwithdrawallimitatanM-Pesaagentis$0.5.Whilethesevaluesmayseemlow,theymaybeburdensomeforahouseholdsubsistingonlessthanadollaraday.Duetothesefactors,people inthissegmentmayexperiencesustainedexclusionevenaftertheadventofmobilemoney.TheseobservationsdrawattentiontothedeficienciesofM-Pesaasaninclusiveinnovationthathaverepercussionsonthelevelofinclusivityachievedatastructurallevel.

Inclusionmap Exclusionmap

Figure6.Mapspresentingthelevelsoffinancialinclusionandexclusionin2016.Thedarkerareasofthemapontheleftrepresenthigherlevelsofinclusion,whilethedarkerareasonthemapontherightrepresenthigherlevelsofexclusion.Theinclusionmapshowsthatdeeperlevelsof

21

inclusionofupto93percenthavebeenachievedinthewestern,southernandcoastalpartsofthecountrywhichhavehigherpopulationdensitiesandhigherlevelsofurbanisation.Deeperlevelsofexclusionofupto52percentareprevalentintheNorthernpartsofcountrySource:FinAccessHouseholdSurveyReport(2016)

However, the development of M-Pesa as a technology for financial inclusion stimulated theopeningupoftheunderlyingfinancialservicesstructure,andtheprioritisationofthemissiontoexpand theaccess frontier. By innovatively using existing technologies to extend services to thepoor,M-Pesademonstratedtoincumbentactorsthatfinancialinclusionisanachievableobjective.At thepolicy level for instance, reviewsof theBankingAct, implementationof theMicrofinanceAct anddraftingof theNationalPaymentSystem regulationsbecamemotivatedby theneed to“createanenablingenvironmentforthese institutionstoofferawiderangeoffinancialservicesand products using a variety of innovative delivery channels to different segments of thepopulation”24.Commercialbankshaveorientedtheircorporatestrategiestocapturethebaseofthe pyramid market by lowering account operating costs and geographically expanding theirbankinginfrastructure.Forinstance,thecorporatepurposeofthelargestbankbycustomerbaseEquityBank is to“transformthe livesand livelihoodsofourpeoplesociallyandeconomicallybyavailingthemmodern,inclusivefinancialservicesthatmaximizetheiropportunities”(EquityBankGroupAnnualReport, 2014,p.1). Similarly,KenyaCommercialBank, the largestbygeographicalreach in Eastern Africa, states that its sustainability long term strategy is to “promote financialinclusion in our region” (Kenya Commercial Bank Group integrated report and financialstatements, 2014, p.15). This prioritisation of an inclusive agenda in formal financial servicesindicatestheemergenceofamoreinclusivestructure.

Level6/Post-structuralinclusion

Asseenintheprecedingsection,sustainedexclusionsinfinancialservicesinKenyapersistinthewakeofmobilemoney.Thisraisesthequestionofwhetherandhowmobilemoneycanaddressthese exclusions, or whether there is need for pursuing a different trajectory of inclusiveinnovation toextend formal financial servicesbeyondthereachofmobilemoney.Nevertheless,discourse at the structural level places the financial inclusion agenda centre stage, and this isevident in the rhetoric of the government and the Central Bank, private sector actors likecommercialbanks,and incivil societyorganisations suchasnot-for-profitorganisations.Wecanthereforeconcludethatpost-structuralinclusionisanon-goingprocess.4 Cross-fertilisingtheframeworksWehaveappliedbothframeworks independentlytothecaseofmobilemoneyinnovationinthefinancial services industry Kenya to empirically elucidate potential areas for cross-fertilisationbetweenMLPandLII.AnimportantcontributionofLIItoMLPisthattheformerallowstoanalyseinclusionsandexclusionsof thenicheand regime inadifferentiatedway.This leads toabetterunderstandingofthetransitiondynamicsandthelevelofsocialinclusionthatwillbeattained.Forexample, itwillbepossible toassess thestabilityof theregimes in termsof its inclusivenature,butalsowhethernicheprocessesareontheirwaytoachievinganinclusivesocio-technicalsystemortriggeringatransitiontoaninclusivesociety.Theprogressive stepsof inclusivitypresented in theLII canbe transported intoMLPanalyses intwoways: Firstly,matching steps in the ladder tophases in the transitionpathway to inclusion.Treated as progressive steps towards social inclusion, the steps in the LII can be applied as24Ndung’u,N.:RemarksoftheGovernoroftheCentralBankofKenyaduringthebreakfastlaunchofFauluKenyaasthefirstlicenceddeposit-takingmicrofinanceinstitutioninKenya,Nairobi,June162009.

22

milestonesinthetransitiontomeasureprogressassociatedwiththedevelopmentofaniche.Forinstance, if thenicheexpansion includesexcludedgroupsbuthashadnomeasurable impactontheirlivelihoods,thetransitionprocessmaybedeemedincompletefromaninclusionperspective.In such a case, the LII would enable the identification of steps or phases that are pending toachievehigher levelsof inclusion.Suchanapproachdeliberatelyreorientstheanalyticalexerciseto social inclusion as the ultimatemeasure of successful system transformation. In theM-Pesacase,thenoveltywasdesignedfortheunbanked(Level1);wasadoptedbyboththebankedandtheunbanked(Level2)andhadtangibleeffectsoneconomicandsocialfactorsinthelivesoftheunbankedpoor (Level3); involvedusers inseveralsteps in the innovationprocess (Level4);andprogressively resulted in a change in policy priorities and product development trajectories incommercial banking to cater for the needs of excluded (Level 5, on-going); and stimulateddiscourse in financial servicesat the localandglobal level toaccount for social inclusion innewdevelopmenttrajectories(Level6,on-going).ByevaluatingthesemilestonesinthecontextofMLPanalysis, the on-going transition to financial inclusion in Kenya may be deemed successful.Secondly, thesteps to inclusivitymaybe imported intoMLPanalysisby identifying thedepthofsocialinclusionachievedatthepointatwhichthenewregimeisputintoplace.TheLIIcanthusbeusednotonlytoassessthepotentialforinclusionbutasanexposttooltogaugetheachievementofinclusivenessresultingfromunfoldingtransitiondynamics.ThecontributionofMLPtoLII literatureisthat itprovidesaframeworktoanalysewhyandhowinnovationsmove up the ladder, andwhich rolemarginalised actors can play. Insights into keyniche formation processes of learning, expectations and network dynamics will help to gain adeeperunderstandingoftheshiftfromInclusionofIntention(Level1)toInclusionofimpact(Level2), and inclusion of consumption (Level 3). The MLP’s formulation of sociotechnical changehowever also provides insights that may prove useful for the conceptualisation of structuralreform as generally discussed in the inclusive innovation literature and presented in the LII.According toHeeks et al. (2013, p. 5), “deep inclusion requires that the underlying institutions,organisationsand relations thatmakeupan innovation systemare inclusive. Thismight requireeither significant structural reformof existing innovation systems, or the creation of alternativeinnovation systems.” Structural change is at the centre of sociotechnical transitions. As seenearlier, the notion of structure is elaborated through sociotechnical systems that consist ofmaterial and social elements such as policies, culture, markets and technologies which areinterdependent, and co-evolve over time into stable configurations referred to as regimes.Sociotechnical transition,orstructural reform,occurswhentherigiditiesandpathdependenciesof regimes are overcome, causing a shift from one configuration to another. Deep inclusiontherefore is achieved when previously exclusive regimes are replaced, transformed orreconfigured.In sum, compatibilities between the MLP and the LII are summarised in Figure 5. ProcessesoccurringduringfirstfourlevelsoftheLIIalignwithprocessesatthenichelevelintheMLP.Level5 of the LII, although in a reduced form, aligns with the regime level where structural changeoccurs.Level6maybealignedwiththesociotechnicallandscape.

Nicheexperiments

Theregime

Thelandscape

NetworkingLearningExpectations

Socio-technicalregimechange

Wesuggestlevel6relatestolandscapechange,thislevelisnotexploredinthisarticle

23

Figure7.RecastingtheLIIoverthethreelevelsoftheMLP

5 ConclusionWe began this paper seeking to understand why M-Pesa—and mobile money innovation ingeneral—hasledtoarapidfinancialservicessociotechnicalsystemchange.OuranalysisusingthemultilevelperspectiveofsociotechnicaltransitionshasrevealedthatM-Pesa’ssuccessisaresultofa confluence of events and processes occurring over a long time in the incumbent financialservices system that consisted of a formal and an informal regime. Over time, a multitude ofpressures from different levels—which were fundamentally oriented towards social inclusion—were exerted on the incumbent system, among them - the microfinance movement, themillennium development goals and the ICT revolution. Tensions arose within the incumbentsystem, some arising from incremental improvements in funds transfer systems, shifts inpreferences,andchangesinregulatorypriorities.Thesetensionsopenedupanopportunityfortheemergenceofamobilemoneyinnovation.Still it isremarkablehowfasttheM-Pesanicheexpanded,andtransformedboththeformalandtheinformalregimecreatinganewmixedregime.Suchtransitionprocessesoftentakemoretime,often50years,asit isarguedinthetransitionliterature.Hereweseeachangein30yearsifweincludetheriseofamicrofinanceniche,andonly15yearsifwefocusonmobilemoneyonly.Thisfast transition can be explained when we take into account the fact that the confluence oflandscapepressures,theopeningupoftheregimeandthenichedevelopedwereentirelyorientedto social inclusion. There is an additional factor, however. Regime actors exhibited a strongwillingness toexperimentwithnew technological and institutional configurations. Policy-makerswerepreparedtotakeariskinanuncertainenvironmenttoachieveasocialobjective,andforcecommercialbankstodothesameinordertosurvivethehollowingoutoftheirbusiness.Basedonthese outcomes we would like to suggest two propositions. For a rapid socio-technical systemtransitions,1) it isnecessarytohavestronglandscapepressures,anopeningupofregimewhichmakesregimeactorswilling toactandstrongnichedevelopmentallorientedtowards thesamegoal,inthiscaseinclusion.Inclusionwasthusnotonlyanoutcomebutalsoafactorintheprocess.In other words, transition might slow down or falter because a strong overriding orientationtowards a similar direction is not present at all three levels; 2) for an inclusive socio-technicaltransition,regimeactorsmustbewillingtotakesignificantrisksandrespondfavourablytothemixofdevelopmentsatall three levels. Themotivation to takebig risks is likely tobehigherwhenregimeactorshaveanormativeinterestinthesocialgoalorientingthetransition.

24

Further research on mobile money as an inclusive innovation could use these propositions toexplorewhythesuccessinKenyahasnotbeenreplicatedinothercontextswithseeminglysimilarinstitutional environments. According to World Bank’s Global Findex Database 2014, mobilemoney usage amongneighbouring countries is notably lower: 37.07 percent in Somalia, 35.1 inUganda, 32.36 in Tanzania, and 18.1 in Rwanda. In India and RomaniawhereM-Pesa has beendeployed, usage rates are 2.35 percent and 0.46 percent respectively. Despite deep mobilepenetration, repeated efforts to launch M-Pesa in South Africa have been unsuccessful, whilemobile money platforms in Nigeria have yet to break through (Villasenor et al., 2015). Aslandscapepressurestofosterinclusionaresimilaracrossthesecountries,wehypothesisethatthecountry-leveldynamicsatthemicro-andmeso-levelsaredifferent:regimeactorsarelessorientedtowards inclusion, and they have a low risk appetite to experimentwith new technologies andnewrulesinthefinancialsystem.AcaseinpointistheNigerianCentralBank’sbank-ledapproachthat barsmobile operators from offeringmobilemoney services (Guidelines onMobileMoneyServices in Nigeria, 2015, p.4), and anecdotally this approach is faulted for limiting uptake ofmobilemoneyservices25.Afinalnoteaboutinclusion.WehaveestablishedthatM-Pesaisinclusiveonlytoacertainextent.Theextremepoororvulnerablecommunitieshavesofarbeenincapableofaccessingmainstreamformalfinancialservicesthroughmobilemoneyduetotheunaffordabilityoftheservice,thelackof capabilities to access and use mobile phone technology, and the absence of mobile moneyinfrastructureinremotepartsofthecountry.Thebulkofthesechallengesarelinkedtosystemicexclusionof rural communities thatwouldbeaddressedviaalternativepolicy interventions thatenabletheadoptionoftechnology,forinstance,provisionofbasiceducationanddevelopmentofinfrastructure for utilities like electricity and roads. The challenge to extending access usingmarket-based inclusive innovations is to reduce the cost of access and ownership to enableparticipationoftheextremepoor.Furtherresearchcanexplorewhetherthecurrentregimeisableto foster inclusion among extremelymarginalised groups, orwhether it is locked in the currentlevel of inclusion, and thus needs another socio-technical system change for overcoming thisbarrier.6 ReferencesADB.(2013).FinancialInclusionandIntegrationthroughMobilePaymentsandTransfer

(ProceedingsofWorkshopon‘EnhancingFinancialIntegrationthroughSoundRegulationofCross-BorderMobilePayments:OpportunitiesandChallenges’).Tunisia:AfricanDevelopmentBank.

Agarwal,N.,&Brem,A.(2012).Frugalandreverseinnovation-LiteratureoverviewandcasestudyinsightsfromaGermanMNCinIndiaandChina.In201218thInternationalICEConferenceonEngineering,TechnologyandInnovation(ICE)(pp.1–11).https://doi.org/10.1109/ICE.2012.6297683

Alzugaray,S.,Mederos,L.,&Sutz,J.(2012).BuildingBridges:SocialInclusionProblemsasResearchandInnovationIssues.ReviewofPolicyResearch,29(6),776–796.https://doi.org/10.1111/j.1541-1338.2012.00592.x

25Llewellyn-Jones,L.:WhymobilemoneytransformedKenya,butfailedtotakeoffinNigeria.(February11,2016)https://iea.org.uk/blog/why-mobile-money-transformed-kenya-failed-to-take-in-nigeria(accessedFebruary01,2017)

25

Baker,L.,Newell,P.,&Phillips,J.(2014).ThePoliticalEconomyofEnergyTransitions:TheCaseofSouthAfrica.NewPoliticalEconomy,19(6),791–818.https://doi.org/10.1080/13563467.2013.849674

Bouman,F.J.A.(1983).Indigenoussavingsandcreditsocietiesinthedevelopingworld.Retrievedfromhttp://library.wur.nl/WebQuery/wurpubs/77112

Byrne,R.P.(2011,June2).Learningdrivers:ruralelectrificationregimebuildinginKenyaand

Tanzania(doctoral).UniversityofSussex.Retrievedfromhttp://sro.sussex.ac.uk/6963/CBK,KNBS,&FSDKenya.(2016).The2016FinAccessHouseholdSurveyonfinancialinclusion.

Nairobi,Kenya:FSDKenya.Chataway,J.,Hanlin,R.,Mugwagwa,J.,&Muraguri,L.(2010).Globalhealthsocialtechnologies:

Reflectionsonevolvingtheoriesandlandscapes.ResearchPolicy,39(10),1277–1288.https://doi.org/10.1016/j.respol.2010.07.006

Chataway,J.,Smith,J.,&Wield,D.(2005).ScienceandtechnologypartnershipsandpovertyalleviationinAfrica.InternationalJournalofTechnologyManagement&Sustainable

Development,5(2),103–123.https://doi.org/10.1386/ijtm.5.2.103_1Christensen,C.M.(1997).TheInnovator’sDilemma:WhenNewTechnologiesCauseGreatFirmsto

Fail.HarvardBusinessPress.Coenen,L.,Benneworth,P.,&Truffer,B.(2012).Towardaspatialperspectiveonsustainability

transitions.ResearchPolicy,41(6),968–979.https://doi.org/10.1016/j.respol.2012.02.014Cozzens,S.,E.,&Kaplinsky,R.(2009).Innovation,povertyandinequality:Causecoincidenceorco-

evolution?InB.-Å.Lundvall,K.J.Joseph,C.Chaminade,&J.Vang(Eds.),Handbookofinnovationsystemsanddevelopingcountries:Buildingdomesticcapabilitiesinaglobal

setting(pp.57–82).EdwardElgarPublishing.Cozzens,S.,&Sutz,J.(2014).Innovationininformalsettings:reflectionsandproposalsfora

researchagenda.InnovationandDevelopment,4(1),5–31.Fressoli,M.,Smith,A.,&Thomas,H.(2011).FromAppropriatetoSocialTechnologies:Some

EnduringDilemmasinGrassrootsInnovationMovementsforSociallyJustFutures.PresentedattheGlobalNetworkfortheEconomicsofLearning,Innovation,andCompetenceBuildingSystems(GLOBELICS)internationalconference,BuenosAires,Argentina.

Fuenfschilling,L.,&Truffer,B.(2014).Thestructurationofsocio-technicalregimes—Conceptualfoundationsfrominstitutionaltheory.ResearchPolicy,43(4),772–791.https://doi.org/10.1016/j.respol.2013.10.010

Fuenfschilling,L.,&Truffer,B.(2016).Theinterplayofinstitutions,actorsandtechnologiesinsocio-technicalsystems—AnanalysisoftransformationsintheAustralianurbanwatersector.TechnologicalForecastingandSocialChange,103,298–312.https://doi.org/10.1016/j.techfore.2015.11.023

Geels,F.W.(2002).Technologicaltransitionsasevolutionaryreconfigurationprocesses:amulti-levelperspectiveandacase-study.ResearchPolicy,31(8–9),1257–1274.https://doi.org/10.1016/S0048-7333(02)00062-8

Geels,F.W.(2004).Fromsectoralsystemsofinnovationtosocio-technicalsystems:Insightsaboutdynamicsandchangefromsociologyandinstitutionaltheory.ResearchPolicy,33(6–7),897–920.https://doi.org/10.1016/j.respol.2004.01.015

Geels,F.W.,&Raven,R.(2006).Non-linearityandExpectationsinNiche-DevelopmentTrajectories:UpsandDownsinDutchBiogasDevelopment(1973–2003).TechnologyAnalysis&StrategicManagement,18(3–4),375–392.https://doi.org/10.1080/09537320600777143

Geels,F.W.,&Schot,J.(2007).Typologyofsociotechnicaltransitionpathways.ResearchPolicy,36(3),399–417.https://doi.org/10.1016/j.respol.2007.01.003

George,A.L.,&Bennett,A.(2005).CaseStudiesandTheoryDevelopmentintheSocialSciences.MITPress.

26

George,G.,McGahan,A.M.,&Prabhu,J.(2012).InnovationforInclusiveGrowth:TowardsaTheoreticalFrameworkandaResearchAgenda.JournalofManagementStudies,49(4),661–683.https://doi.org/10.1111/j.1467-6486.2012.01048.x

GSMA.(2015).StateoftheIndustryReportonMobileMoney.London,UnitedKingdom:GroupeSpecialeMobile(GSM)Association.Retrievedfromhttp://www.gsma.com/mobilefordevelopment/wp-content/uploads/2016/04/SOTIR_2015.pdf

Gupta,A.K.(2013).TappingtheEntrepreneurialPotentialofGrassrootsInnovation.StanfordSocialInnovationReview.Retrievedfromhttp://www.ssireview.org/articles/entry/tapping_the_entrepreneurial_potential_of_grassroots_innovation

Gupta,A.K.,Sinha,R.,Koradia,D.,Patel,R.,Parmar,M.,Rohit,P.,…Vivekanandan,P.(2003).Mobilizinggrassroots’technologicalinnovationsandtraditionalknowledge,valuesandinstitutions:articulatingsocialandethicalcapital.Futures,35(9),975–987.https://doi.org/10.1016/S0016-3287(03)00053-3

Hanlin,R.,&Muraguri,L.(2009).Improvingaccesstohealthtechnologiesbythepoor:thesocialcontextinTanzanianbednetproductionanddelivery.InternationalJournalofTechnologyManagement&SustainableDevelopment,8(3),237–248.https://doi.org/10.1386/ijtm.8.3.237/1

Hansen,T.,&Coenen,L.(2015).Thegeographyofsustainabilitytransitions:Review,synthesisandreflectionsonanemergentresearchfield.EnvironmentalInnovationandSocietal

Transitions,17,92–109.https://doi.org/10.1016/j.eist.2014.11.001Heeks,R.,Amalia,M.,Kintu,R.,&Shah,N.(2013).Inclusiveinnovation:definition,

conceptualisationandfutureresearchpriorities(IDPMDevelopmentInformaticsWorkingPaperSeriesNo.53).UniversityofManchester,UK.

Heeks,R.,Foster,C.,&Nugroho,Y.(2014).Newmodelsofinclusiveinnovationfordevelopment.InnovationandDevelopment,4(2),175–185.

Hospes,O.,Musinga,M.,&Ong’ayo,M.(2002).AnEvaluationofMicro-FinanceProgrammesinKenyaasSupportedThroughtheDutchCo-FinancingProgramme.Retrievedfromhttp://agris.fao.org/agris-search/search.do?recordID=NL2012019474

Hughes,N.,&Lonie,S.(2007).M-PESA:MobileMoneyforthe‘Unbanked’TurningCellphonesinto24-HourTellersinKenya.Innovations:Technology,Governance,Globalization,2(1–2),63–81.https://doi.org/10.1162/itgg.2007.2.1-2.63

IDRC.(2011).InnovationforInclusiveDevelopment-ProgramProspectusfor2011-2016.Iizuka,M.(2015).Diverseandunevenpathwaystowardstransitiontolowcarbondevelopment:

thecaseofsolarPVtechnologyinChina.InnovationandDevelopment,5(2),241–261.https://doi.org/10.1080/2157930X.2015.1049850

IMF.(2016).FinancialDevelopmentinSub-SaharanAfrica:PromotingInclusiveandSustainableGrowth.InternationalMonetaryFund.Retrievedfromhttp://www.imf.org/~/media/websites/imf/imported-full-text-pdf/external/pubs/ft/dp/2016/_afr1605.ashx

infoDev.(2012).MobileUsageattheBaseofthePyramidinKenya.Washington,DC:TheWorldBank.Retrievedfromhttp://www.infodev.org/en/TopicPublications.34.html

Jack,W.,&Suri,T.(2011).MobileMoney:TheEconomicsofM-PESA(WorkingPaperNo.16721).NationalBureauofEconomicResearch.Retrievedfromhttp://www.nber.org/papers/w16721

Jack,W.,&Suri,T.(2014).RiskSharingandTransactionsCosts:EvidencefromKenya’sMobileMoneyRevolution.AmericanEconomicReview,104(1),183–223.https://doi.org/10.1257/aer.104.1.183

Kaplinsky,R.(2011a).Innovationforpro-poorgrowth:fromredistributionwithgrowthtoredistributionthroughgrowth.PresentedattheConferenceinHonourofSirRichardJolly

27

-‘Fromstructuraladjustmenttohumandevelopment:impactonpovertyandinequality’,Brighton.Retrievedfromhttp://www.ids.ac.uk/F6671340-0A2A-11E1-8E5C005056AA4739

Kaplinsky,R.(2011b).SchumachermeetsSchumpeter:Appropriatetechnologybelowtheradar.ResearchPolicy,40(2),193–203.https://doi.org/10.1016/j.respol.2010.10.003

Kemp,R.,Rip,A.,&Schot,J.W.(2001).Constructingtransitionpathsthroughthemanagementofniches.InR.Garud&P.Karnoe(Eds.),PathDependenceandCreation.PsychologyPress.

Kemp,R.,Schot,J.,&Hoogma,R.(1998).Regimeshiftstosustainabilitythroughprocessesofnicheformation:Theapproachofstrategicnichemanagement.TechnologyAnalysis&StrategicManagement,10(2),175–198.https://doi.org/10.1080/09537329808524310

Kendall,J.,Maurer,B.,Machoka,P.,&Veniard,C.(2011).AnEmergingPlatform:FromMoneyTransferSystemtoMobileMoneyEcosystem.Innovations:Technology,Governance,Globalization,6(4),49–64.https://doi.org/10.1162/INOV_a_00100

Kraemer-Mbula,E.,&Watu,W.(2010).InnovationandtheDevelopmentAgenda.OECDPublishing.

Langat,A.(2014).CommunityBanksHelpReconciliationinKenya’sRiftValley.ACR,InstituteforWarandPeaceReporting,(399).Retrievedfromhttps://iwpr.net/global-voices/community-banks-help-reconciliation-kenyas-rift-valley

Malkamäki,M.,Johnson,S.,&Nino-Zarazua,M.(2009).Theroleofinformalfinancialgroupsin

extendingaccessinKenya.FinancialSectorDeepeningKenya.Retrievedfromhttp://www.fsdkenya.org/pdf_documents/10-05-07_Role_of_informal_financial_groups_in_Kenya.pdf

Mas,I.,&Ng’weno,A.(2010).ThreeKeystoM-PESA’sSuccess:Branding,ChannelManagementandPricing.JournalofPaymentsStrategyandSystems,4(4).

Mas,I.,&Radcliffe,D.(2011).Mobilepaymentsgoviral�:M-PESAinKenya.CapcoInstitute’sJournalofFinancialTransformation,32,162–182.

Maxon,R.M.(1992).TheEstablishmentoftheColonialEconomy.InW.R.Ochieng’&R.M.Maxon(Eds.),AnEconomicHistoryofKenya.EastAfricanPublishers.

McCormick,D.(1987).FundisandFormality:VerySmallManufacturersinNairobi.McKay,C.,&Pickens,M.(2010).BranchlessBanking2010:Who’sServed?AtWhatPrice?What’s

Next?(CGAPFocusNote).ConsultativeGrouptoAssistthePoor.Retrievedfromhttp://www.cgap.org/publications/branchless-banking-2010-who%E2%80%99s-served-what-price-what%E2%80%99s-next

Meadowcroft,J.(2000).SustainableDevelopment:aNew(ish)IdeaforaNewCentury?PoliticalStudies,48(2),370–387.https://doi.org/10.1111/1467-9248.00265

Morawczynski,O.,&Pickens,M.(2009).PoorPeopleUsingMobileFinancialServices�:ObservationsonCustomerUsageandImpactfromM-PESA.WorldBank,WashingtonDC.Retrievedfromhttps://openknowledge.worldbank.org/handle/10986/9492

Mytelka,L.,&Farinelli,F.(2000).LocalClusters,InnovationSystemsandSustained

Competitiveness(UNU-INTECHDiscussionPaperSeriesNo.5).UnitedNationsUniversity-INTECH.Retrievedfromhttp://econpapers.repec.org/paper/unmunuint/200005.htm

Ndung’u,N.(2014).IncreasingtheleveloffinancialinclusioninKenya.RemarksbytheGovernoroftheCentralBankofKenya,Nairobi,Kenya.Retrievedfromhttp://www.bis.org/review/r140516b.htm

Ochieng’,W.R.,&Maxon,R.M.(1992).AnEconomicHistoryofKenya.Nairobi:EastAfricanPublishers.

Paunov,C.(2013).InnovationandInclusiveDevelopment:ADiscussionoftheMainPolicyIssues.OECDScience,TechnologyandIndustryWorkingPapers,(2013/01),63.https://doi.org/10.1787/18151965

Plyler,M.,Haas,S.,&Nagarajan,G.(2010).CommunityLevelEconomicEffectsofM-PESAinKenya:InitialFindings.FinancialServicesAssessmentProject,MicrofinanceOpportunities.Retrievedfromhttp://www.microfinancegateway.org/p/site/m/template.rc/1.9.44854/

28

Rao,B.C.(2013).Howdisruptiveisfrugal?TechnologyinSociety,35(1),65–73.https://doi.org/10.1016/j.techsoc.2013.03.003

Raven,R.,&Verbong,G.(2007).Multi-RegimeInteractionsintheDutchEnergySector:TheCaseofCombinedHeatandPowerTechnologiesintheNetherlands1970–2000.TechnologyAnalysis&StrategicManagement,19(4),491–507.https://doi.org/10.1080/09537320701403441

Rip,A.,&Kemp,R.P.M.(1998).TechnologicalChange.In:RaynerS.,MaloneE.L.(editors).InHumanChoiceandClimateChange.Vol.II,ResourcesandTechnology(pp.327–399).Columbus,Ohio:BattellePress.Retrievedfromhttp://doc.utwente.nl/34706/

Rock,M.,Murphy,J.T.,Rasiah,R.,vanSeters,P.,&Managi,S.(2009).Ahardslog,notaleapfrog:GlobalizationandsustainabilitytransitionsindevelopingAsia.TechnologicalForecastingandSocialChange,76(2),241–254.https://doi.org/10.1016/j.techfore.2007.11.014

Romijn,H.A.,&Caniëls,M.C.J.(2011).TheJatrophabiofuelssectorinTanzania2005–2009:Evolutiontowardssustainability?ResearchPolicy,40(4),618–636.https://doi.org/10.1016/j.respol.2011.01.005

Safaricom.(2013).SafaricomAnnualReport&GroupAccounts2013.Retrievedfromhttp://www.safaricom.co.ke/about-us/investor-relations/annual-reports-pdf

Safarzyńska,K.,Frenken,K.,&vandenBergh,J.C.J.M.(2012).Evolutionarytheorizingandmodelingofsustainabilitytransitions.ResearchPolicy,41(6),1011–1024.https://doi.org/10.1016/j.respol.2011.10.014

Schot,J.,Kanger,L.,&Verbong,G.(2016).Therolesofusersinshapingtransitionstonewenergysystems.NatureEnergy,1,16054.https://doi.org/10.1038/nenergy.2016.54

Sen,A.K.(2000).SocialExclusion:Concept,Application,andScrutiny.AsianDevelopmentBank.Retrievedfromhttp://www.adb.org/publications/social-exclusion-concept-application-and-scrutiny

Seyfang,G.,&Smith,A.(2007).Grassrootsinnovationsforsustainabledevelopment:Towardsanewresearchandpolicyagenda.EnvironmentalPolitics,16(4),584–603.https://doi.org/10.1080/09644010701419121

Silver,H.(1994).Reconceptualizingsocialdisadvantage:threeparadigmsofsocialexclusion.InG.Rodgers,C.G.Gore,&J.B.Figueiredo(Eds.),SocialExclusion:Rhetoric,Reality,Responses.InternationalInstituteforLabourStudies.

Smith,A.,&Raven,R.(2012).Whatisprotectivespace?Reconsideringnichesintransitionstosustainability.ResearchPolicy,41(6),1025–1036.https://doi.org/10.1016/j.respol.2011.12.012

Smith,A.,Stirling,A.,&Berkhout,F.(2005).Thegovernanceofsustainablesocio-technicaltransitions.ResearchPolicy,34(10),1491–1510.https://doi.org/10.1016/j.respol.2005.07.005

Swilling,M.,&Annecke,E.(2012).JustTransitions:ExplorationsofSustainabilityinanUnfairWorld.Claremont,SouthAfrica�:Tokyo,Japan:UnitedNationsUniversityPress.

UNCTAD.(2014).Innovationpolicytoolsforinclusivedevelopment(Investment,EnterpriseandDevelopmentCommission,sixthsessionNo.TD/B/C.II/25).Geneva,Switzerland.Retrievedfromhttp://unctad.org/meetings/en/SessionalDocuments/ciid25_en.pdf

vanEijck,J.,&Romijn,H.(2008).ProspectsforJatrophabiofuelsinTanzania:AnanalysiswithStrategicNicheManagement.EnergyPolicy,36(1),311–325.https://doi.org/10.1016/j.enpol.2007.09.016

Villasenor,J.,West,D.M.,&Lewis,R.J.(2015).The2015BrookingsFinancialandDigitalInclusionProjectReport.Washington,DC:TheBrookingsInstitute.Retrievedfromhttp://www.brookings.edu/research/reports/2015/08/financial-digital-inclusion-2015-villasenor-west-lewis

WorldBank.(2013a).China-Inclusiveinnovationforsustainableinclusivegrowth(WorkingPaperNo.82519)(pp.1–95).Washington,DC:WorldBankGroup.Retrievedfrom

29

http://documents.worldbank.org/curated/en/2013/10/18639876/china-inclusive-innovation-sustainable-inclusive-growth

WorldBank.(2013b).GlobalFinancialDevelopmentReport2014:FinancialInclusion.Washington,DC:WorldBankPublications.

Yin,R.K.(2009).Casestudyresearch:designandmethods.SagePublications.Zeschky,M.B.,Widenmayer,B.,&Gassmann,O.(2011).FrugalInnovationinEmergingMarkets.

Research-TechnologyManagement,54(4),38–45.https://doi.org/10.5437/08956308X5404007

Zollman,J.(2014).KenyaFinancialDiaries–Shilingikwashilingi,thefinanciallivesofthepoor.FinancialSectorDeepening,Kenya.