Ebissa, Ta ekegn Ta iku; As aw, A ega Seyoum; Lakew, De esse Me sha
A icle
Women’s pa icipa ion and cos e iciency in mic o inance
ins i u ions: a Sub-Saha an s udy
Cogen Business & Managemen
P o ided in Coope a ion wi h:
Taylo & F ancis G oup
Sugges ed Ci a ion: Ebissa, Ta ekegn Ta iku; As aw, A ega Seyoum; Lakew, De esse Me sha (2024) :
Women’s pa icipa ion and cos e iciency in mic o inance ins i u ions: a Sub-Saha an s udy, Cogen
Business & Managemen , ISSN 2331-1975, Taylo & F ancis, Abingdon, Vol. 11, Iss. 1, pp. 1-16,
h ps://doi.o g/10.1080/23311975.2024.2304307
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/326005
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Cogen Business & Managemen
ISSN: 2331-1975 (Online) Jou nal homepage: www. and online.com/jou nals/oabm20
Women’s pa icipa ion and cos efficiency in
mic ofinance ins i u ions: a Sub-Saha an s udy
Ta ekegn Ta iku Ebissa, A ega Seyoum As aw & De esse Me sha Lakew
To ci e his a icle: Ta ekegn Ta iku Ebissa, A ega Seyoum As aw & De esse Me sha Lakew
(2024) Women’s pa icipa ion and cos efficiency in mic ofinance ins i u ions: a Sub-Saha an
s udy, Cogen Business & Managemen , 11:1, 2304307, DOI: 10.1080/23311975.2024.2304307
To link o his a icle: h ps://doi.o g/10.1080/23311975.2024.2304307
© 2024 The Au ho (s). Published by In o ma
UK Limi ed, ading as Taylo & F ancis
G oup
View supplemen a y ma e ial
Published online: 01 Feb 2024.
Submi you a icle o his jou nal
A icle iews: 1295
View ela ed a icles
View C ossma k da a
Full Te ms & Condi ions o access and use can be ound a
h ps://www. and online.com/ac ion/jou nalIn o ma ion?jou nalCode=oabm20
Banking & Finance | ReseaRch aR icle
Cogen Business & ManageMen
2024, VoL. 11, no. 1, 2304307
Women’s pa icipa ion and cos e iciency in mic o inance
ins i u ions: a Sub-Saha an s udy
a ekegn a iku ebissaa , a ega seyoum as awa and De esse Me sha lakewb
aJimma uni e si y, Jimma, e hiopia; be hiopian Ci il se ice uni e si y, addis ababa, e hiopia
ABSTRACT
his s udy aims o examine he ac o s in luencing he cos e iciency o mic o inance
ins i u ions in sub-saha an a ican coun ies. a en-yea unbalanced panel da a
consis ing o 128 mic o inance ins i u ions om 34 coun ies was used o he analysis.
he s udy included all mic o inance ins i u ions in sub-saha an a ica ha epo ed o
he Mic o inance in o ma ion exchange da abase and had inancial epo s o a leas
i e consecu i e iscal yea s be ween 2009 and 2018. he cos e iciency o he
ins i u ions was in es iga ed using a s ochas ic on ie app oach. he indings indica e
ha mic o inance ins i u ions ope a ing in sub-saha an a ican coun ies a e ope a ing
beyond he cos on ie line, wi h only 13% demons a ing e iciency. key de e minan s
o cos e iciency in mic o inance ins i u ions in ssa include he size o he ins i u ion,
cos pe bo owe , women’s pa icipa ion, ins i u ional ype, and coun y income
ca ego y. no ably, inc eased women’s pa icipa ion as boa d membe s, loan o ice s,
and bo owe s signi ican ly imp o es cos e iciency. Mic o inance ins i u ions a e
encou aged o ans o m hei cus om p ac ices in line wi h he apid changes in he
echnology and echniques o mode n inancial se ice p o isions and ensu e he
in ol emen o women in boa d membe s, supe iso y posi ions and bo owing
se ices. We also ecommend ha ins i u ions appoin women selec i ely on nomina ed
wo k posi ions.
IMPACT STATEMENT
he p ima y aim o he s udy is o in es iga e he associa ion be ween women pa icipa ion
and cos e iciency o mic o inance ins i u ions. he s udy examines he pa icipa ion o
women in mic o inance ins i u ions om i e engagemen pe spec i es, namely boa d
membe , managemen membe , loan o ice , egula s a and bo owing se ices. as
expec ed, women’s pa icipa ion on he boa d, as loan o ice and bo owe imp o e he
cos e iciency o mic o inance ins i u ions in ssa. he indings indica e ha allowing
women o be engaged in hese h ee posi ions can signi ican ly enhance he cos e iciency
o he mic o inance ins i u ions. i is ad isable o employ women selec i ely in designa ed
wo k posi ions wi hin mic o inance ins i u ions.
1. In oduc ion
access o a o dable adi ional inancial se ices emains limi ed in sub-saha an a ica (ssa) coun ies, pa -
icula ly o low-income indi iduals, he uneduca ed, and women (Mlachila e al., 2016). a Wo ld Bank
epo shows ha 80% o adul s in de eloping economies ha e no access o o mal inancial ins i u ions o
sa e hei money while hey a e using in o mal and cos ly me hods o sa e (Paza basioglu e al., 2020).
Mos low-income people, women, and small-scale business en e p ises ha e no easy access o a o dable
inancial se ices o e ed by la ge banks and insu ance companies in many de eloping economies.
Mic o inance ins i u ions (MFis) a e designed o se e hose people excluded om adi ional inancial se -
ices due o hei low economic s a us, less c edi wo hiness (azad e al., 2016) and cos ly o moni o ing
and adminis e ing small loans ha p io i ized by poo es class (chu chill, 2018; adesse aba e e al., 2014).
© 2024 he au ho (s). Published by in o ma uK Limi ed, ading as aylo & F ancis g oup
CONTACT a ekegn a iku ebissa a ekegn[email p o ec ed] Jimma uni e si y, Jimma, e hiopia.
a ekegn a iku ebissa is an academic s a a Wallaga uni e si y, nekem e, e hiopia.
h ps://doi.o g/10.1080/23311975.2024.2304307
his is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons a ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0/), which
pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. he e ms on which his a icle has been
published allow he pos ing o he accep ed Manusc ip in a eposi o y by he au ho (s) o wi h hei consen .
ARTICLE HISTORY
Recei ed 18 augus 2023
Re ised 22 Oc obe 2023
accep ed 8 Janua y 2024
KEYWORDS
Mic o inance; women;
cos e iciency; gende
di e si y; sub-saha an
a ica
SUBJECTS
gende & De elopmen ;
sus ainable De elopmen ;
Finance; Business,
Managemen and
accoun ing; gende
s udies - soc sci
REVIEWING EDITOR
Da id McMillan,
Uni e si y o s i ling,
Uni ed kingdom
2 . . eBissa e al.
Despi e hei con ibu ions in c ea ing inclusi e inancial se ices, MFis in ssa a e s uggling wi h inan-
cial ine iciency and noneconomic p oblems. a s udy conduc ed by geb emichael and gessesse (2016) con-
cludes ha he echnical e iciency o MFis in ssa is below a e age, a only 48.9%. Below a e age cos
e iciency is also ecognized in he MFis o he egion. acco ding o abdulai and ewa i (2016), he cos
e iciency o MFis in he egion is limi ed o 40.09%. a b ie epo eleased by he cgaP1 esea ch eam
e lec s he exis ence o mul idimensional challenges acing he ins i u ions in he egion. he ins i u ions
a e cha ac e ized by holding asse s wi h poo quali y, smalle loan size, less access o capi al ma ke , weak
go e nance, absence o anspa ency, and ope a ing in uns able economic and poli ical condi ions (gliso ic
e al., 2012). hese challenges migh magni y he ine iciency o MFis in he egion.
as he challenges s and, howe e , he con ibu ions o women o he cos e iciency o MFis emain
e ile esea ch g ound in ssa li e a u e. acco ding o MiX (2021) epo , women can engage in i e di -
e en ac i i ies in MFi ope a ions. hey could ac as a bo owe , boa d membe , loan o ice , manage
and no mal s a (pe sonnel) in he day- o-day ac i i ies o he ins i u ions. hough hese engagemen s
explain why he oles o women a e mul idimensional in he ope a ions o MFis, he exis ing empi ical
li e a u e has ailed o add ess he signi icance o he oles, pa icula ly om he loan o ice , manage
and no mal pe sonnel aspec s.
he possible eason o his gap is he exis ence o one side pe cep ion wi h s akeholde s o he ma e .
Because o he disad an aged g oup conside a ions, women a e ecognized as po en ial clien s o MFis
a he han po en ial se an s o he ins i u ions. his has p essed schola s o ocus only on bo owe s’ and
boa d membe s’ gende di e si y aspec s in hei s udies. as a esul , many empi ical s udies ha e ocused
on assessing he oles o women only om bo owe s’ and boa d membe s’ gende di e si y pe spec i es
(adusei, 2019; abdulai & ewa i, 2016; Mo i e al., 2015; he mes e al., 2011). his s udy a emp s o ill his
gap by assessing he signi icance o i e oles o women in he cos e iciency o MFis in ssa coun ies.
his s udy aims o assess he associa ion be ween women’s oles and he cos e iciency o MFis in
ssa. speci ically, he associa ion be ween gende di e si y in bo owe s, boa d membe s, manage s, loan
o ice s and egula pe sonnel and he cos e iciency o MFis was widely add essed. hus, his s udy is
he i s in i s scope ac oss ssa empi ical s udies o add ess he i e oles o women in he e iciency o
MFis. On he o he hand, he indings o he s udy enhance he gene alizabili y o he s udy esul s om
wo aspec s. Fi s , a la ge numbe o MFis ope a ing in ssa coun ies we e conside ed o a longe
pe iod (2009–2018). second, an ad anced da a analysis echnique, ha is, he s ochas ic on ie app oach
(sFa), was employed o analyze he collec ed da a. Mo eo e , he s udy indings inspi e policymake s and
he go e ning body o MFis o ecognize he con ibu ions o women in he cos e iciency o MFis om
i e scena ios o gende di e si y.
he s udy esul s e eal ha he size o he mic o inance ins i u ion, cos pe bo owe , women’s
engagemen , o ms o mic o inance ins i u ions and coun y income ca ego y a e majo de e minan s o
he cos e iciency o he ins i u ions. in pa icula ly, women’s ep esen a ion in boa d membe s, bo ow-
ing se ices and appoin men s o loan o ice posi ions signi ican ly imp o e he cos e iciency o MFis
in ssa. s uc u ally, he emaining con en s o his s udy a e p esen ed in i e majo sec ions. Rela ed
heo e ical and li e a u e e iews a e p esen ed in he second sec ion. he esea ch me hodology and
design a e p esen ed in he hi d sec ion. he ou h sec ion p esen s he empi ical indings, and he i h
sec ion summa izes he conclusions and manage ial implica ions o he s udy indings. Policy implica-
ions and sugges ions o u u e esea ch a e p esen ed in he six h sec ion.
2.Theo e ical and empi ical li e a u e e iew
e iciency measu es he success o a i m in eaching i s op imal a ge a minimized cos in he gi en
ime ame. e iciency encompasses wo complemen a y inancial goals o a i m: cos minimiza ion and
p o i maximiza ion (s ai i, 2010). i ealizes he maximum possible ou pu s by spending ewe po en ial
esou ces in business ope a ions. cos e iciency measu es he achie emen o a i m in a aining an op i-
mal ou pu le el a he lowes possible cos (nguyen & Pham, 2020; hassan & sanchez, 2009) ela i e o
i s bes -p ac iced i m ope a ing in simila condi ions. he cos e ec i eness o MFis is he ocus o his s udy.
his sec ion p esen s he e iewed li e a u e om heo e ical and empi ical aspec s o i m e iciency.
Di e en heo ies ha e eme ged o illus a e he associa ion be ween wo k o ce gende di e si y and
cOgen BUsiness & ManageMen 3
i m e iciency. he “ alue-in-di e si y pe spec i e” is he mos men ioned pe spec i e in ecen s udies.
he pe spec i e p oposes ha gende di e si y in he wo k o ce b ings al e na i e iews and knowledge
o he decision-making p ocess by inc easing he a ailabili y o in o ma ion (abou-el-sood, 2021; B ahma
e al., 2020; Xie e al., 2020). in line wi h he iew o he “ alue-in-di e si y” hypo hesis, abou-el-sood
(2021) concluded ha he p esence o emales on a boa d has a meaning ul implica ion in making
in es men decisions. he au ho s a ed ha emale di ec o s ake in o cau ion bank esou ces, he
s eng h o i s capi al base and he essen ial e u n and ela ed isk o accep o ejec in es men wi h
less isky o isky oppo uni ies. as a esul , emale di ec o on a boa d b ings e hical/socie al pe spec-
i es and new esou ces o he decision-making p ocess.
Fu he mo e, Xie e al. (2020) a gued ha gende di e si y p omo es he inno a ion e iciency o a
eam by gene a ing in o ma ion and c ea ing social bene i s. he au ho s jus i y ha women a e compe-
en in acili a ing communica ion, sha ing in o ma ion, c ea ing a common unde s anding and building
social bene i s. he p esence o h ee o mo e women on a boa d enhances he inancial pe o mance
o a i m (B ahma e al., 2020). he au ho s belie e ha women a e empowe ed o ha e enough mana-
ge ial powe , be e in o ma ion and ime o in luence decision-making as hey hold execu i e posi ions.
F om ano he pe spec i e, esou ce dependency heo y sugges s ha a i m accumula es ad iso y
and guidance bene i s and ge access o po en ial esou ces and communica ion channels when i s
boa d is a composi e o gende (Duppa i e al., 2019). in hei s udy, B ahma e al. (2020) and Duppa i
e al. (2019) sugges ed ha gende di e si y on a boa d enhances he inancial pe o mance o a i m
by p e en ing an indi idual o g oup o indi iduals wi h simila opinions om domina ing he
decision-making p ocess. i also inc eases manage ial accoun abili y and he quali y o moni o ing
oles. hese sugges ions also ha e meaning ul implica ions o gende di e si y om an agency heo y
pe spec i e. his means ha women’s pa icipa ion on a boa d and in execu i e posi ions could min-
imize agency p oblems by p e en ing con lic s o in e es and decep i e ope a ions in a i m. in line
wi h he pos ula e o agency heo y, many empi ical s udies ha e con i med he p esence o a posi i e
ela ionship be ween b oad gende di e si y and he inancial pe o mance o a i m (B ahma e al.,
2020; Duppa i e al., 2019).
howe e , social iden i y heo y has e ol ed wi h con adic o y opinions. he heo y sugges s ha
indi iduals may use age and gende as a ibu es o c ea e hei pe sonal ca ego y (in-g oup) and o he
social g oups (ou -g oup) wi h he desi e o ei he sha e o deny exis ing ac s ( ep e & loy, 2017). he
summa ized li e a u e in he s udy o ali e al. (2014) ealized ha social ca ego iza ion maximizes he
di e ence be ween in-g oup and ou -g oup. hese social a angemen s e ode g oup cohesion, smoo h
communica ion and coope a ion. in-g oup membe s a e conside ed as mo e us wo hy, hones and
coope a i e han ou -g oup membe s. his esul s in less collabo a ion wi h ou -g oup membe s and
may lead o in ensi e agency p oblems.
Fo una ely, ali e al. (2014) could no ind adequa e e idence ha con i ms he p esence o a nega-
i e ela ionship be ween boa d gende di e si y and he inancial pe o mance o i ms, consis en wi h
he p oposi ions o social iden i y heo y. he un o una e is he absence o adequa e empi ical in es i-
ga ions ha jus i y he iew o a o emen ioned heo ies om ssa mic o inance ins i u ions pe spec i es.
Ou s udy p ima ily aimed o ill he li e a u e gap om ssa coun ies lookou s.
On he o he hand, he indings o he exis ing empi ical li e a u e a e also help ul o use in he s udy
o he e iciency o MFis. a ield epo conduc ed in e hiopia con i ms he di icul y o achie ing no able
ou each and a aining cos e iciency a a ime due o he exis ence o a sys ema ic ade-o ( adesse
aba e e al., 2014). O he s udy conduc ed in a ica coun ies on mic o inance ins i u ions’ inancial sus-
ainabili y and ou each dep h con i ms he exis ence o a adeo be ween he wo aspec s (chu chill,
2018). in addi ion, he mes e al. (2011) p o ided s ong e idence ha shows he p esence o nega i e
associa ion be ween ou each o he poo and e iciency o MFis. acco ding o chu chill (2018; he mes
& lensink, 2011), mic o inance ins i u ions a e igno ing social mission o ensu ing hei inancial sus ain-
abili y. hus, he ou each o mic o inance ins i u ions is becoming an oppo uni y cos o e iciency o
he ins i u ions.
Fu he s udies a e ealizing ha de e minan s o he e iciency o MFis in ssa coun ies a e mul idi-
mensional. acco ding o O eng-abayie e al. (2011), an ins i u ion’s age, ou each, p oduc i i y and cos
pe bo owe play signi ican oles in de e mining he economic e iciency o MFis. speci ically, he
4 . . eBissa e al.
au ho s con i m he exis ence o posi i e ela ionship be ween cos pe bo owe and he e iciency o
MFis. he au ho s pe cei e ha mic o inance ins i u ions in es hei ime and esou ces on moni o ing,
aining and ad ising cus ome s imp o e e iciency le el han hose do no ; hus, MFis ope a ing in
ghana ha e he oppo uni y o enjoying economics o scale. howe e , in es iga ion o abdulai and
ewa i (2016) was no success ul o sugges he associa ion be ween cos pe bo owe and cos e i-
ciency o MFis in ssa.
in ano he s udy, abdulai and ewa i (2016) sugges ed ha o al asse s, ope a ing expense o asse s
a io, a e age loan balance pe sa e , he pe cen age o emale bo owe s and bo owe s pe s a mem-
be a e he key de e minan s o cos e iciency o MFis. Pa icula ly, he au ho s sugges ed ha lending
o mo e women magni ies he cos ine iciency o MFis in ssa. likewise, he mes e al. (2011) a gues ha
MFis wi h la ge women bo owe s and lowe a e age loan balances a e less e icien . howe e , agos inho
e al. (2021) con i ms ha p o iding inancial se ices o women and disad an aged people makes MFis
inancially mo e p o i able and sus ainable. his inding o i ies he iew o Fadikpe e al. (2022) and
he mes and lensink (2011) which s a ed ha women bo owe s ou pe o m male bo owe s in epaying
a loan and eliabili y.
conside able s udy indings also exis in he li e a u e ega ding he oles o i m owne ship s uc-
u e ( o m) and ope a ing loca ion in e iciency o MFis. he s udy inding o adesse aba e e al.
(2014) e ealed ha coope a i e o m o MFis is mo e e icien in managing cos s han he specialized
o m o MFis. in a simila ein, hassan and sanchez (2009) concluded ha o mal MFis (banks & c edi
unions) a e echnically mo e e icien han in o mal MFis (no - o -p o i o ganiza ions & non inancial
ins i u ions). howe e , he s udy indings o agos inho e al. (2021) and geb emichael and gessesse
(2016) con adic he a o emen ioned conclusions: nonbank inancial ins i u ions (nBFis) and nongo -
e nmen al o ganiza ions (ngOs) a e inancially mo e e icien han c edi union o bank o ms o MFis.
a ela ed s udy pe o med by ilahun (2021) s a es ha he legal s a us and loca ion o MFis signi i-
can ly de e mine he inancial pe o mance o inancial ins i u ions in ssa. hese indings sugges he
exis ence o a conside able ole o a i m owne ship s uc u e, legal s a us and loca ion in he e i-
ciency o mic o inance ins i u ions. in e es ing esea ch indings and epo s a e also a ailable ega d-
ing he ole o a coun y’s income ca ego y in he e iciency o MFis. acco ding o hassan and sanchez
(2009) MFis ope a ing in sou h asia a e mo e e icien han hose in la in ame ica and Mena coun-
ies. Mo eo e , gliso ic e al. (2012) added ha MFis in ssa egions pe o m wo se han MFis in o he
egions ega ding asse quali y and cos managemen .
empi ical s udies assessing he impac s o go e ning mechanisms applied in MFis ha e also
d awn meaning ul esea ch insigh . Fo ins ance, Mo i e al. (2015) claim ha a ibu es o boa d
membe s ha e a signi ican impac on he pe o mance o mic o inance ins i u ions. hey s a ed
ha he ou each pe o mance o MFis is imp o ed when he boa d membe s become mo e inde-
penden , include emales, and clea sepa a ion o du ies exis s be ween he chie execu i e o ice
(ceO) and he chai pe son. simila s udy indings a e e ealed in he s udy o kye eboah‐coleman
and Osei (2008). in ela ed s udy, adusei (2019) s a ed ha he e ec o boa d gende di e si y on
echnical e iciency o mic o inance ins i u ions depends on he size o he ins i u ion. acco ding
o he opinion o he au ho , he p esence o emale on he boa ds in smalle mic o inance ins i-
u ion hu s echnical e iciency o he ins i u ion and ice e sa. in o he wo ds, la ge mic o i-
nance ins i u ions a e bene i ed om boa d gende di e si y han he smalle one in ela ion o
echnical e iciency.
howe e , exis ing empi ical e idence ega ding o he oles o women in he cos e iciency o MFis is
no insigh ul o hose ope a ing in ssa coun ies. Mo e ocus has been paid o assessing he oles o
women only om bo owe and boa d membe poin s o iew (adusei, 2019; chu chill, 2018; Mo i e al.,
2015). ne e heless, women a e engaged in day- o-day ope a ions o MFis as manage s, loan o ice s and
egula pe sonnel. hus, in addi ion o add essing o he de e minan s, his s udy aims o unde s and he
ela ionship be ween women’s pa icipa ion in manage ial, egula pe sonnel and loan o ice posi ions
and he cos e iciency o MFis in ssa.
cOgen BUsiness & ManageMen 5
3. Ma e ials and me hods
3.1. Sampling and da a sou ce
his s udy aims o in es iga e he cos e iciencies o MFis in ssa coun ies om he pe spec i e o
women’s pa icipa ion. Mic o inance ins i u ions ope a ing in ssa coun ies we e selec ed based on he
a ailabili y o adequa e da a. in addi ion, he ollowing h ee condi ions we e used o sample he MFis.
Fi s , a mic o inance ins i u ion licensed o ope a e in a speci ic ssa coun y wi h a speci ied o m o
inancial ins i u ions was conside ed. second, mic o inance ins i u ions ha p esen ed annual inancial
epo s o he global mic o inance in o ma ion exchange (MiX – ma ke )2 da abase be ween 2009 and
2018 we e selec ed. hi d, hose MFis ha did no p esen annual epo s o he MiX–ma ke da abase
o a leas i e concu en iscal yea s wi hin he indica ed ime ame we e excluded om he s udy.
subsequen ly, we ex ac ed seconda y da a o 128 MFis ope a ing in 34 sub–saha an a ica (ssa) coun-
ies and ob ained an unbalanced panel da ase o 930 obse a ions.
3.2. Va iable de ini ion and measu emen
he a iables used in his s udy we e g ouped as dependen a iable ( o al cos ), inpu p ices, ou pu
alues, i m-speci ic and coun y-speci ic ac o s. he de ailed de ini ion and measu emen o each a i-
able a e p esen ed in able 1 wi h essen ial ema ks.
3.3. Model speci ica ion
in his s udy, he s ochas ic on ie app oach (sFa) p oposed by Ba ese and coelli (1995) is applied o
examine he cos e iciency o MFis. he s ochas ic on ie app oach is a pa ame ic app oach ha allows
esea che s o analyze panel da a o s ochas ic p oduc ion, cos and/o p o i unc ions. he app oach
uses an econome ic model and speci ies he dis u bance e m in e ms o ine iciency e m and he
idiosync a ic e o . Unlike he nonpa ame ic app oaches (such as da a en elopmen analysis (Dea)), sFa
p o ides an e ec i e es ima ion o he e iciency le el by sepa a ing ine iciency e ms om o he s o-
chas ic shocks (Ba ese & coelli, 1995). Pa ame ic and nonpa ame ic app oaches a e also di e en in
esea ch indings consis ency. acco ding o nguyen and Pham (2020) a gumen , cos e iciency sco es
es ima ed unde he sFa a e mo e consis en han unde he Dea model. Mo eo e , he sFa has he
empi ical ad an age o allowing esea che s o in oduce coun y-speci ic o i m-speci ic a iables in o
he s ochas ic on ie model (s ai i, 2010) o u he in es iga ion. On he o he hand, he pa ame ic
app oach is less sensi i e o mul icollinea i y p oblems and ou lie s e ec s. howe e , he sFa is c i icized
o he unc ional o m speci ica ion and no mal dis ibu ion assump ions imposed on he e icien on-
ie models. acco ding o Dong e al. (2014) sugges ion, unc ional o m misspeci ica ions would subjec
o inaccu a e e iciency sco e es ima ion.
as p esen ed in his sec ion, he anslog on ie unc ion o cos e iciency es ima ion is de i ed
om he o al cos o sampled MFis. o al cos ( c) e iciency measu es he minimum possible inpu cos
incu ed by a i m o achie e he a ge ed maximum ou pu ela i e o i s bes pe o me . he c unc-
ion associa es he p ice incu ed o use inpu s—such as labo , unds, and physical capi al o p oduce
ou pu s—such as ne loan po olios and o he ea ning asse s.
TC z
ij ij
ij
=
()
+
ε
(1)
ε
ij ij ij
u= +
(2)
he s ochas ic c unc ion o
MFIi
ope a ing in speci ic coun y j ac oss ime pe iod is de ined in
e ms o he explana o y a iables
zij
and he dis u bance e m
ε
ij
. he dis u bance e m is u he
6 . . eBissa e al.
di ided in o andom shock ij
()
and ac ual ine iciency e m uij
()
. o do his, he anslog s ochas ic cos
unc ion is adop ed in his s udy.
ln
TC
PPC ln Labo
PPC ln Fund
PPC
ij ij
= +
+
αα α
01 2 +
()
+
()
+
ij
ij ij
ln Loan ln OEA
ln Labo
PPC
αα
α
34
5
1
2 +
+
[]
+
ij ij
ij
ln Fund
PPC Loan
2
6
2
7
2
8
1
2
1
2
1
2
α αα
ln ll n O E A
ln Labo
PPC ln Fund
PPC l
ij
ij ij
[]
+
+
2
910
αα
*nn Labo
PPC ln L an ln Labo
PPC ln O
ij
ij
ij
()
+
*o *
α
11 EE A
ln Fund
PPC ln Loan ln Fund
PPC
ij
ij
ij
()
+
()
+
αα
12 13
*
()
+
( )()
++
ij
ij ij ij
ln OEA ln Loan ln OEA
T ln Lab
**
α
αα
14
15 16
oo
PPC T ln Fund
PPC T ln Loan T
ij ij
ij
+
+
()
* **
αα
17 18 ++
()
+ ++
α
α
19
20
2
1
2
ln OEA T
Tu
ij
ij ij
*
(3)
he s ochas ic on ie app oach assumes ha he o al cos de ia es om he a ge ed cos because
o he andom dis u bance e m
ij
and he ine iciency e m
uij
(Ba ese & coelli, 1995). he dis u -
bance e m
()
ij ep esen s a unca ed andom e o due o measu emen e o om explana o y a i-
ables and is assumed o be independen and iden ically dis ibu ed om
uij
wi h n (0,
σ
2
). he
ine iciency e m
(
u
ij
) ep esen s he nonnega i e andom a iable ha es ima es he ine icien e ec s
and is assumed o ollow an asymme ic hal no mal dis ibu ion in which bo h he mean u and he
a iance
σ
u
2
a e a ied. Fu he mo e, pa ame iza ion echniques sugges ed by Ba ese and coelli
(1995) and used in lu e al. (2018; s ai i, 2010) o
σ
2
and
σ
u
2
a e applied in his s udy: hese a e
σσ
22
=
+
σ
u
2
and
γσ σ σ
= +
u u
2 22
/ ( ).
he coe icien o pa ame e
γ
lies be ween 0 and 1. he a iance o he ine iciency e ec s is null
when he coe icien o pa ame e
γ
equals ze o (Ba ese & coelli, 1995). a small ine iciency e ec is
ecognized as
γ
being close o ze o and a la ge ine iciency as
γ
being close o one (s ai i, 2010).
Mo eo e , a linea homogenei y assump ion is imposed on he inpu p ices o labo and unds, and he
o al cos by no malizing hem in e ms o he p ice o physical capi al (PPc) be o e aking hei loga-
i hms (lu e al., 2018; s ai i, 2010). in his case, he cos e iciency (ce) sco es o each MFi a e es ima ed
using he unc ion:
CE u
ij ij
=1/ exp( )
.
acco ding o Ba ese and coelli (1995), a one-s ep s ochas ic on ie model can be used o iden i y
p edic o s o he e iciency o a i m. he s ochas ic on ie app oach uses he maximum likelihood es i-
ma ion echnique o p edic he pa ame e s included in he on ie model. in his s udy, he ollowing
al e na i e model is o mula ed o assess he de e minan s o he cos ine iciency o MFis a e es ima -
ing he ine iciency sco es h ough he anslog s ochas ic cos unc ion.
u ln i msize nROA lnCos pe b o Boa
ij ij ij ji
=+ ++ +
ββ β β β
01 2 3 4
l dg Mgm g Femaleb
lnFemales a Femal
ij ij ij
ij
++
++
ββ
ββ
56
78
eeo Income Type z
ij ij ij ij
+ ++
ββ
9 10
(4)
whe e, ln is he na u al loga i hm unc ion,
uij
is cos ine iciency sco e, i msize is o al asse s o MFi, ROA
is he e u n on asse s, Cos pe b o is cos pe bo owe , Boa dg is boa d gende di e si y in %, Mgm g is
managemen gende di e si y in %, Femaleb is he p opo ion o women bo owe s in %, Females a is
numbe emale pe sonnel and Femaleo is emale loan o ice in %. income is income ca ego y o he
MFi’s coun y; ype is legal s a us o MFi and z
ij
is he dis u bance e m in ine iciency de e minan s
es ima ion.
cOgen BUsiness & ManageMen 7
4. Resul s and discussion
4.1. A ibu es o cos e iciency o MFIs in SSA
a s ochas ic on ie app oach was employed o assess he a ibu es and de e minan s o he cos e i-
ciency o MFis in ssa. an unbalanced penal da a se was ob ained om he annual epo s o 128
selec ed mic o inance ins i u ions. he da a se co e s en-yea annual epo s be ween 2009 and 2018.
he da a analysis begins in his sec ion by classi ying i in o wo main sec ions. he i s sec ion p esen s
he basic a ibu es o cos e iciency and he second sec ion p esen s he ac o s a ec ing he cos e i-
ciency. he a ibu es o cos e iciency a e analyzed based on e idence p esen ed in able 2 and able
3 (see he (appendix).
MFis ope a ing in he ssa egion ealize cos e iciency below he expec ed a e age alue, which is
only 13.07% demons a ing e iciency. Rema kable e iciency sco e di e ences a e obse ed among he
sampled MFis. he cos e iciency o MFis ope a ing in he egion a ies be ween 87 and 2%. On a e -
age, MFis ope a ing in angola a e mo e cos e icien han hose ope a ing in o he coun ies o ssa.
MFis ope a ing in angola a e anked i s by ealizing he la ges cos e iciency sco e (87%), ollowed
by Mali and he congo Republic o sco ing 24 and 17% on a e age, espec i ely. in con as , MFis om
Uganda, Malawi and sudan a e he mos cos -ine icien inancial ins i u ions, ealizing less han 10%
cos e iciency sco es on a e age. in gene al, coun y-speci ic a ibu es such as economic policy and
poli ical s abili y a e he mos likely easons o he exis ence o such cos e iciency sco e disc epancies
among he MFis in he egion.
he e is no signi ican di e ence in he cos e iciency le el among o he ypes o MFis, excep
o he bank o m o MFis. he cos e iciency o he bank o m o MFi (9.4%) was less han ha o
he o he o ms o MFi, on a e age. simila ly, he e is no ema kable cos e iciency sco e a ia ion
among he income ca ego ies o he MFi coun ies. howe e , uppe -middle income coun ies we e
less e icien in cos managemen (9.56%) han o he income ca ego y coun ies. Rega ding he ime
end, cos e iciency sco es a ied be ween 12.5 and 13.59%, wi hou showing a signi ican sco e
di e ence be ween 2009 and 2018. his means ha almos uni o m cos e iciency le el is obse ed
om pe iod o pe iod. his implies ha MFis in ssa ha e he endency o ollow ou ine and cus-
omized inancial se ice p o ision s a egies. hus, he ins i u ions a e cha ac e ized by slow o
weak sel -adap a ion o human esou ce e o m, new ope a ing echniques and inno a i e inancial
echnology.
in gene al, MFis in he ssa egion a e ei he slowly adop ing o no adop ing inancial inno a ions
and new s a egies, ega dless o he exis ing acuum o imp o e cos e iciency by almos 87%. hese
esul s ha e ano he insigh ul implica ion. i shows he p esence o oppo uni y o imp o ing cos e i-
ciency up o 87% by implemen ing inancial inno a ion and new se ice p o ision s a egies. hese
e o ms con ibu e o he cos e iciency o ins i u ions by minimizing po en ial esou ce was age and
ensu ing e icien u iliza ion o exis ing asse s. hese indings a e consis en wi h he s udy indings o
gliso ic e al. (2012) and O eng-abayie e al. (2011).
4.2. Cos ine iciency o MFIs
he cos on ie model is s a is ically signi ican and accep able o analysis o h ee easons (see able
3). Fi s , he chi-squa e es o ze o coe icien a ia ion in he model was ejec ed a he 1% signi icance
le el (
x2
= 2043.13). his implies ha he explana o y a iables ha e signi ican ly explained he exis ing
a ia ions in he cos e iciency model and ha he coe icien s o he pa ame e s a e highly di e en
om ze o. second, he alue o sigma-squa ed (
σ
2
= 0.4216, 8.0330) was signi ican a he 1% signi i-
cance le el, implying ha he es ima e o he pa ame e s is highly signi ican . hi d, he es ima ed alue
o gamma (γ = 0.7694, 77%) is also highly signi ican a he 1% signi icance le el, which implies ha a
signi ican amoun o a ia ion is de i ed om he ine iciency o he MFis, while he a iance due o
andom e o is small. in addi ion, he coe icien o e a (η) (0.0337) is close o ze o, and he alue is
signi ican , implying ha he e is no signi ican di e ence be ween he esul s o he ime-in a ian and
ime- a ying decay on ie models in his s udy. howe e , a ime- a ying decay on ie model is p e-
e ed o unde s and he ea u es o changes in he cos ine iciency o he MFis ac oss ime.
14 . . eBissa e al.
Table 2. summa y s a is ic o cos e iciency sco es.
Cos e iciency sco es (%)
Mean s d.De
Panel A: Types o MFI
nBFi 14.22 0.131
Bank 9.40 0.065
ngo 14.81 0.092
C edi union/coope a i e 11.66 0.069
o e all e iciency 13.07 0.099
Panel B: Income ca ego y o MFI’s coun y
uppe -middle income 9.56 0.043
Lowe -middle income 13.62 0.110
Low income 12.46 0.083
o e all e iciency 13.07 0.099
Panel C: Time end (2009–2018)
2009 12.50 0.109
2010 12.70 0.105
2011 13.25 0.106
2012 12.77 0.068
2013 12.96 0.098
2014 13.31 0.105
2015 13.29 0.104
2016 13.37 0.111
2017 13.15 0.083
2018 13.59 0.089
o e all e iciency 13.07 0.099
Panel D: Top Th ee Coun ies wi h mos cos e icien MFIs
sco e (%) Rank
angola 87% 1
Mali 24% 2
Congo Republic 17% 3
egyp 17% 3
Panel E: Coun ies wi h smalles cos e icien MFIs
uganda 8% 3
Malawi 7% 2
sudan * 2% 1
sou ce: au ho s’ compu a ion.
*MFis ope a ing in sudan ha e sco ed he smalles cos e iciency esul , which means ha hey a e e icien only o 2%.
cOgen BUsiness & ManageMen 15
Table 3. s ochas ic on ie eg ession esul s.
Dependen a iable – o al cos
Cos ine iciencyPanel a: inpu , ou pu s and C oss e ms
no a ion Pa ame e Coe z- alue
ln(Labo /PPC)
α
1
0.5622 3.17***
ln(Fund/PPC)
α
2
15.1082 2.55**
ln(Loan)
α
3
0.2265 2.01**
ln(oea)
α
4
0.0108 0.11
½ ln(Labo /PPC)^2
α
5
0.1024 11.82***
½ ln(Fund/PPC)^2
α
6
−20.2819 −2.97***
½ ln(Loan)^2
α
7
0.0514 11.81***
½ ln(oea)^2
α
8
0.0079 1.55
ln(Labo /PPC)* ln(Fund/PPC)
α
9
−2.3964 −7.73***
ln(Labo /PPC)* ln(Loan)
α
10
−0.0523 −4.17***
ln(Labo /PPC)* ln(oea)
α
11
0.0052 0.64
ln(Fund/PPC)* ln(Loan)
α
12
−0.0601 −0.14
ln(Fund/PPC)* ln(oea)
α
13
0.7328 2.84***
ln(Loan)* ln(oea)
α
14
−0.0082 −1.24
ln(Labo /PPC)*
α
15
0.0135 3.34***
ln(Fund/PPC)*
α
16
0.2304 1.60
ln(Loan)*
α
17
−0.0058 −1.35
ln(oea)*
α
18
0.0007 0.20
½( )^2
α
19
−0.0018 −0.50
α
20
0.0904 1.70*
Cons an
α
0
0.4095 0.27
Wald Chi-squa e 2043.13 000***
sigma squa ed 0.4216 8.03***
gamma (γ) 0.7694 24.54***
e a (η) 0.0337 000***
Log-likelihood unc ion −456.7440
numbe o obse a ion 930
numbe o g oup 128
***p < 1%, **p < 5%, *p < 10%.
sou ce: au ho s’ eg ession model ou pu s.
16 . . eBissa e al.
Table 4. Fac o s a ec ing cos ine iciency o MFis in ssa.
Dependen a iable: Cos ine iciency Model 1 Model 2 Model 3
independen a iable Pa ame e Coe - alue Coe - alue Coe - alue
ln i msize
β
1
0.066 7.12*** 0.085 9.79***
lnRoa
β
2
0.030 0.81 0.040 1.06
lnCos pe bo
β
3
0.057 3.70*** 0.056 3.64***
Boa dg
β
4
−0.171 −2.73*** −0.223 −3.28***
Mgm g
β
5
0.079 1.16 0.168 2.32**
Femaleb
β
6
0.030 0.42 −0.329 −4.83***
lnFemales a
β
7
0.063 5.70*** 0.098 8.64***
Femaleo
β
8
−0.212 −2.81*** −0.252 −3.16***
uppe _middle_
income
β
9
−0.072 −0.57 0.006 0.05
Lowe _middle_
income
β
9
−0.179 −5.41*** −0.151 −4.60***
nBFi
β
10
−0.241 −5.19*** −0.233 −4.95***
Bank
β
10
0.002 0.04 0.003 0.06
ngo
β
10
−0.229 −4.83*** −0.220 −4.75***
Cons an
β
0
1.051 4.36*** 0.912 3.90*** 2.37 21.28***
numbe o obs. 930 930 930
Chi-squa e 313.66*** 253.64*** 121.89***
LR es o he one-sided e o 3.68** 11.64 *** 1.66 *
Log likelihood unc ion −601.44 −622.95 −683.05
*** p < 1%, ** p < 5%, * p < 10%.
sou ce: au ho s’ eg ession model ou pu s.
