• No results found

Social capital of entrepreneurs in a developing country: the effect of gender on access to and requests for resources

N/A
N/A
Protected

Academic year: 2021

Share "Social capital of entrepreneurs in a developing country: the effect of gender on access to and requests for resources"

Copied!
13
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Social capital of entrepreneurs in a developing country

Citation for published version (APA):

Solano, G., & Rooks, G. (2018). Social capital of entrepreneurs in a developing country: the effect of gender on

access to and requests for resources. Social Networks, 54, 279-290.

https://doi.org/10.1016/j.socnet.2018.03.003

DOI:

10.1016/j.socnet.2018.03.003

Document status and date:

Published: 01/07/2018

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be

important differences between the submitted version and the official published version of record. People

interested in the research are advised to contact the author for the final version of the publication, or visit the

DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page

numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at:

openaccess@tue.nl

providing details and we will investigate your claim.

(2)

ContentslistsavailableatScienceDirect

Social

Networks

j ou rn a l h o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / s o c n e t

Social

capital

of

entrepreneurs

in

a

developing

country:

The

effect

of

gender

on

access

to

and

requests

for

resources

Giacomo

Solano

,

Gerrit

Rooks

SchoolofInnovationSciences,EindhovenUniversityofTechnology,P.O.Box513,5600MB,Eindhoven,TheNetherlands

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Availableonline4April2018 Keywords:

Developingcountries Gender

Personalnetworks Accesstoresources Requestsforresources Socialcapital Smallbusinesses

a

b

s

t

r

a

c

t

Thispaperaddressesgenderdifferencesinthesocialcapitalofentrepreneursinadevelopingcountry.

Socialnetworksareoftenanimportantassetforaccessingresources;however,theymayalsobea

lia-bilityindevelopingcountries,sinceentrepreneursareoftenexpectedtosupporttheircontacts.Usinga

recentsurveyamongurbanandruralUgandanentrepreneurs,wefocusonthefinancialresourcesthat

entrepreneurscanobtainfromtheircontactsontheonehand,andrequestsforfinancialsupportmade

totheentrepreneursfromthesecontactsontheotherhand.Ourresultsshowthattherearegender

differencesassociatedwithaccessto,andrequestsfor,financialresources.

©2018ElsevierB.V.Allrightsreserved.

1. Introduction

Female entrepreneurship in developing countriesis

increas-inglyreceivingattentionfromscholarsandpolicymakers(Minniti

and Naudé, 2010; Lindvert et al., 2017).Female entrepreneurs

canmake significantcontributions toinnovation and economic

growthindevelopingcountries(BrushandCooper,2012;deBruin

etal.,2007;Welteretal.,2007).However,femaleentrepreneurs

arestillan‘untappedsource’ofgrowth(Vossenberg,2013),given

thattheyfacemanybarriersrelatedtotheirgender,andthis

pre-ventsthemfromreachingtheirfullpotential(Jamali,2009;Lindvert

etal.,2017;Yetim,2008).Notwithstandingtheincreasing

atten-tionandpolicyinitiatives,anddespitetheimportanceoffemale

entrepreneurshipfordevelopingcountries,thereisstilla

signifi-cantgendergapwhenitcomestoentrepreneurshipindeveloping

countries(Vossenberg, 2013).Businessesowned bywomen are

generallymorelikelytounder-performorfail,duetoformaland

informalobstacles(RoomiandParrott,2008;Vossenberg,2016a,b).

Onepossiblereasonforthisgendergapisthedifferencebetween

thesocialnetworksofmaleandfemaleentrepreneursin

develop-ingcountries(Jamali,2009;Lindvertetal.,2017).Researchershave

longsinceacknowledgedthatentrepreneurialactivityis

embed-dedin network relationships(Dubiniand Aldrich,1991; Hoang

andAntoncic,2003).Thereisconsensusthatnetworksofpersonal

relationsareanimportantassetthatdeterminesthesuccessofa

∗ Correspondingauthor.

E-mailaddresses:giacomo.solano@gmail.com(G.Solano),g.rooks@tue.nl

(G.Rooks).

business,henceentrepreneurialnetworksareoftensaidto

consti-tuteaformofsocialcapital(Stametal.,2014).

Ingeneral,ithasbeennotedthatthetypeandamountofsupport

thatwomencanobtainfromtheirnetworksdiffersfromwhatmen

canobtain(vanEmmerik,2006).Thereisalackofsystematic

evi-denceonthedifferencesbetweenmaleandfemaleentrepreneurs

intheirsocialcapitalindevelopingcountries(Al-Dajanietal.,2015;

Mairetal.,2012;Lindvertetal.,2017;Myroniuk,2016).As

under-linedbyLindvertetal.(2017,759),“recentworkshaveincreasingly

questionedwhethertheoreticalframeworksonsocialcapitalfrom

matureeconomiccontextsapplytowomenentrepreneursin

devel-opingcountrycontexts,wherereligiousandculturalnormscould

beaprominenthindranceinleveragingsocialcapital”.

Moreover,mostresearchregardingnetworksofentrepreneurs

predominantlyfocusesonthepositive outcomesofnetworksin

developingcountries,namelytheresourcesthatanentrepreneur

can get from his/her contacts (Boso et al., 2013; Berrou and

Combarnous,2011,2012;Bruton etal.,2007;Fafchamps,2001;

FafchampsandMinten,1999,2002;FafchampsandQuinn,2016).

Thedownsidesofsocialnetworksbothindevelopingand

devel-oped countries have received less attention in the literature,

althoughvariousnegativeaspects derivedfromsocialnetworks

havebeenmentionedonoccasion(Barr,2002;Deguilhemetal.,

2017; Nordman, 2016; Nordman and Pasquier-Doumer, 2015;

O’Brien,2012;Portes,1998).Anecdotalevidence,early

anthropo-logicalresearch(Hunter,1962;KhalafandShwayri,1966)andafew

recentstudies(Albyetal.,2014;Grimmetal.,2013)suggestthat

excessiveclaimsonentrepreneursisanimportantissuein

develop-ingcountries,andthisislinkedtoascarcityofresources(Comola,

2016).Successfulentrepreneursfacedistributiveobligations.Once

abusinessbecomessuccessfulandgeneratesprofit,furthergrowth

https://doi.org/10.1016/j.socnet.2018.03.003

(3)

maybehinderedbecauseentrepreneursareexpectedtosupport

relatives,friends,andcommunitymembers.However,systematic

researchintothedownsidesofnetworksindevelopingcountriesis

scarce,especiallywithregardtotheSub-SaharanAfricancontext

(seeRooksetal.,2016).

Inthisarticlewecomparefemaleandmaleentrepreneursin

termsofboththebenefitstheycangetfromtheircontacts(access

toresources),andtheclaimstheymightreceivefromthesecontacts

(requestsforresources).Inotherwords,wefocusonaccessto,and

requestsfor,resourcesviatheentrepreneurs’contacts(i.e.,their

socialcapital).Wefocusonfinancialresourcesinparticular.

In developing countries, where the financial and legal

sys-tems arestill underdeveloped, financial resources are a critical

issuefor entrepreneurs(Beck and Demirguc-Kunt,2006; Cook,

2001).Siba’s(2016)elaborationonWorldDevelopmentIndicators

(WDI)showedthatthisissueisevenmorecriticalinsub-Saharan

Africa, where the formal borrowing rate is lower compared to

otherdevelopingareas(e.g.LatinAmericaandEastAsia).

Previ-ousstudiesunderlinedthattherearedifferencesbetweenmenand

womenintermsofaccessingfinancialcapital(AmineandStaub,

2009; Fletschner, 2009;Makena et al.,2014; Malmströmet al.,

2017;MarlowandPatton,2005;Mwobobia,2012;Lindvertetal.,

2017;Siba,2016).Byandlarge,women(entrepreneurs)havemore

difficultygettingformalfinancialsupport(e.g.,loans)thanmen

(entrepreneurs).FiguresfromtheInternationalFinance

Corpora-tionreport(IFC,2013)estimatedthat63–69%ofbusinessesowned

byawomanareunservedorunderservedbyfinancialinstitutions

indevelopingcountries.Womenarelesslikelythanmentohavea

bankaccountandtoborrowformally(Demirgüc-Kuntetal.,2015;

Siba,2016,ZinsandWeill,2016).Thisislinkedwith(formaland

non-formal)collateralrequirements.Forexample,studiesinKenya

(Makenaetal.,2014;Mwobobia,2012)highlightthat,although

for-mallymenandwomencanaccessloansequally,inpractice,women

facemoredifficultieswhentryingtoaccesscredit,astraditional

beliefsandgenderrolescontinuetoinfluenceresourceallocation.

Asaconsequence,womendonothavetheassetsthatbanks

nor-mallyrequiretosecurecredit(Makenaetal.,2014).Inaddition,

womenoftenlackthepiecesofinformationrequiredtogetloans

(Vossenberg,2016a,b)Forexample,FletschnerandMesbah(2011)

foundthatParaguayanwiveswerelesslikelytohaveknowledgeof

financialmarketsandinstitutionsthantheirhusbands.Finally,as

notedbyVossenberg(2016,15),“womenentrepreneursoften(...)

mayfacediscriminatorypractices,suchasbankingclerks

question-ingthelegitimacyandabilityofwomenentrepreneurstogrowa

businesswhenaskingforaloan”.Giventhis,thedifferencebetween

maleandfemaleentrepreneursintheroleofsocialnetworkswhen

itcomestofinancialresourcesseemsparticularlyimportant.

Theaimofthisarticleistoempiricallyinvestigategender

dif-ferencesintheformationofnetworksofsupportandrequestsin

relationtosmallbusinessactivitiesinadevelopingcountry,namely

Uganda(EastAfrica).We conductedalarge-scalesurveyin two

regions:anurbanarea(thecountrycapital,Kampala)andNakaseke

(amoreruralareainCentralUganda).Thisallowedustocompare

genderdifferencesinnetworksbetweenamoretraditional,

collec-tivisticcontext(theruralarea)andamoremodern,individualistic

context(theurbanarea).Indeed,asnoticedbyVossenberg(2013),

thecontextinwhichtheentrepreneurisembeddedisparticularly

importantwhenitcomestogenderdynamics.

2. Theory

2.1. Socialcapitalandaccessto/requestsforresources

Socialcapitalisabroadconceptwithmanydifferent

interpreta-tions(forreviews,see:Lin,2001;AdlerandKwon,2002;Akc¸omak

andterWeel,2009).InthisarticlewedrawuponPortes(1998),

whodefinedsocialcapitalasthe“abilityofactorstosecure

ben-efitsbyvirtueofmembershipinsocialnetworks”(Portes,1998,

6).Entrepreneurs–indevelopedanddevelopingcountries–can

obtainvariousresourcesfromtheirsocial connections,suchas:

information, finances and emotional support(see for example;

GreveandSalaff,2003;HoangandAntoncic,2003).However,as

notedbyPortes(1998),thecreationof,andparticipationin,social

networksisnotcost-free.1Whileentrepreneursmaygainaccess

toresourcesfromcontacts,converselythose relationsmayalso

involvecosts,andtheentrepreneurs’contactsmayinturntryto

obtainresourcesfromthem.

InthispaperweadoptPortes’(1998,8)double-edgedview2that

socialcapitalentailsboth“network-mediatedbenefits”–namely,

theresourcesthatapersoncanobtainfromhis/hercontacts–and

“claimsongroupmembers”–namely,theresourcesthataperson

maybe‘forced’togivetohis/hercontacts–.We therefore

dis-tinguishbetweenaccesstoresourcesthroughsocialcontactsand

requestsforresourcesonthepartofthesecontacts.

Asalreadyillustratedintheintroduction,wefocusonfinancial

supportfortwomainreasons.First,financialissuesand,in

particu-laraccesstofinancialresources,iscriticalwhenitcomestorunning

abusiness,especiallyindevelopingcountrieswhereresourcesare

limited(BeckandDemirguc-Kunt,2006).Second,previous

litera-turehasunderlinedthatthereisagenderdimensioninaccessto

financialresources(MarlowandPatton2005;Lindvertetal.,2017;

Vossenberg,2016a,b).

2.2. Genderdifferencesinsocialcapital

Theexistingliteraturehasrepeatedlysuggestedthatmaleand

femaleentrepreneurialnetworksdifferintermsoftheir

compo-sitionand structure(Agneessenset al.,2006; Aidisetal.,2007;

Bastani,2007;Moore,1990;Myroniuk,2016;LieblerandSandefur,

2002;vanEmmerik,2006).Inherpioneeringarticleon

determi-nantsofmen’sandwomen’spersonalnetworksintheUS,Moore

(1990)founddissimilaritiesbetweenmenandwomen:whereas

mendiscusspersonalmatterswithawiderrangeofcontacts

(co-workers,friends,relatives,etc.),womenaremorelikelytodiscuss

themwithrelativesandneighbourhoodfriends,andhavecloser,

and more homogeneous contacts in their personal-advice

net-works.

Lessisknownaboutthedifferencesbetweenmaleandfemale

entrepreneurswhenitcomestotheresourcestheycanobtainfrom

theirnetworks(Ahl,2006;Foss,2010).Inoneofthefewarticles

addressingthistopic,vanEmmerik(2006)foundthatmenwere

moreabletoaccessjob-relatedresourcesthroughtheircontacts

thanwerewomen.

2.2.1. Gender(egolevel)

Genderrolesareinfluencedbyculturalcontext,whichshapes

expectationsandrelationsbetweenmenandwomen(Acker,1992;

Baughn etal.,2006).In more traditionalsocieties, genderroles

causemen–entrepreneursandnon-entrepreneurs–totakeon

theresponsibilityofprovidingfinancialsupportfortheir‘group’

(Farré,2013;Jamali,2009;RismanandDavis,2013).

1Portes(1998,5)providedaclearexampletoexplainhisview:“Saying,for

exam-ple,thatstudentAhassocialcapitalbecauseheobtainedaccesstoalargetuition loanfromhiskinandthatstudentBdoesnotbecauseshefailedtodosoneglects thepossibilitythatB’skinnetworkisequallyormoremotivatedtocometoher aidbutsimplylacksthemeanstodoso.Definingsocialcapitalasequivalenttothe resourcesthusobtainedistantamounttosayingthatthesuccessfulsucceed”.

2Wetooktheexpression‘double-edgeviewofsocialcapital’fromLindvertetal. (2017).

(4)

Byandlarge,inUgandamenareexpectedtohavethe

finan-cialpowertotakecareoftheir(extended)family,clan,etc.(Otiso,

2006).Duetotheirroleasprovidersofresources,weexpectmale

entrepreneurstohavemorepeoplewhoaredependentonthemfor

financialsupport,andfewerpeoplewhocanhelpthemfinancially.

Ontheonehand,maleentrepreneursmayfacegreater

redistribu-tivepressure;asmen,theseentrepreneursmightfeelobligatedto

helpcertainpeople.Ontheotherhand,theymightbeless

atten-tivetoincludeintheirnetworkspeoplewhocanprovideaccess

tofinancialcapital.Forthesereasons,weexpectthat,whenthe

entrepreneurismale,heislesslikelytohaveaccesstofinancial

resourcesfromhiscontacts,andthatthecontactsaremorelikely

torequirefinancialsupportfromhim.

Hypothesis1a. whentheentrepreneurismale,contactsareless

likelytoprovideaccesstofinancialresourcestotheentrepreneur;

Hypothesis1b. whentheentrepreneurismale,contactsaremore

likelytorequestfinancialresourcesfromtheentrepreneur.

2.2.2. Gender(alter)

Most studies on women entrepreneurship have focused on

genderasanattributeoftheentrepreneur.Lessisknownabout

whether male or female contacts provide similar resources to

entrepreneurs(Klyver,2011).The literature suggeststhat male

contactsaremorelikelytoprovideinstrumentalsupportsuchas

financialresources,whilefemalecontactsaremorelikelyto

pro-videemotionalsupport(Klyver,2011;LieblerandSandefur,2002;

Plickertetal.,2007;ReevyandMaslach,2001).Thispatternhas

beenobservedinadevelopingcountryaswell.Intheirresearch

intotheSidama,anagro-pastoralistpopulationinsouthwestern

Ethiopia,Caudelletal.(2015)foundthatmalecontactswerethree

timesmorelikelytobementionedaslendersthanfemalecontacts.

Thisseemsrelatedtothefactthatwomengenerallyhaveless

economicpowerandcontroloverfinancesthanmen.Firstly,they

have less economic power due to limited property ownership,

smallersavings,and greaterdifficultyinaccessingformalcredit

(AmineandStaub,2009).Secondly,whentheydohavefinancial

resources,womenfacegreaterdifficultiesinmaintainingcontrol

over these resources,especially due to issues of controlin the

households(Agarwal,1997;Ateridoetal.,2013;Minniti,2010;ILO,

2017;Jamali,2009;Vossenberg,2016a,b).Thehusband–oranother

malerelativeinthehousehold(e.g.,father,brother)ifthewoman

isnotmarried–normallycontrolsthehouseholdassets.Arecent

ILOstudyonUganda(ILO,2017)confirmsthisbyhighlightingthe

factthatfemaleentrepreneursaremorelikelytokeepcontrolof

theirsavingswhentheyarewillingtohidetheirmoneyfromtheir

husbands.

Therefore,asmenhavemorefinancialpower,theyaremore

likelytokeepcontroloverhouseholdassets,andsincewomenhave

difficultyaccessingotherlending options,we expectmale

con-tactstobebetterabletoprovidefinancialsupportthanarefemale

contacts.Thus,wehypothesisedthat:

Hypothesis2a. malecontactsaremorelikelytoprovideaccessto

financialresourcestotheentrepreneurs;

Hypothesis2b. malecontactsarelesslikelytorequestfinancial

resourcesfromtheentrepreneurs.

2.2.3. Gender(egoandalter)

Relationshipsaregender-orientedastheychangebasedonthe

genderofthepersonsinvolved (Klyver,2011).Providingaccess

to,orrequestingfinancialsupportmaydependonwhetherthe

entrepreneurandthecontactareofthesameoroppositegender.

Homophilyreferstothetendencyofpeoplewithsimilarattributes,

suchasgender,tointeract(McPhersonetal.,2001).Thereisaclear

lackofresearchonhowgenderhomophilyinfluencesaccessto,and

requestsfor,resources,especiallyindevelopingcountries(Caudell

etal.,2015).However,previousstudiessuggestthatlending

net-worksarecharacterisedbyalackofhomophily(Caudelletal.,2015;

Fafchamps,1992;Platteau,1997).Thisseemstoapplyevenmore

readilytogenderhomophily.Asnotedabove,womenareless

eco-nomicallypowerfulthanmen,andtheyareusuallydependenton

men(theirhusband,theirfather,etc.).Therefore,itseemslesslikely

thatpeopleofthesamesexwouldprovideaccessto/requestsfor

finances:

Hypothesis3a. whentheentrepreneurand thecontactareof

thesamegender(male-maleorfemale-female),thecontactisless

likelytoprovideaccesstofinancialresources

Hypothesis3b. whentheentrepreneurandthecontactareof

thesamegender(male-maleorfemale-female),thecontactisless

likelytorequestfinancialresources

Tofurtherdisentangletheinteractionbetweenegoand alter

gender,wenowfocusonsituationswheretheentrepreneurand

thecontactareofdifferentgenders(male-femaleorfemale-male).

Asillustratedbefore,womengenerallylackfinancialpower,access

to,andcontrolover,finances(AmineandStaub,2009;ILO,2017;

Vossenberg,2016a,b).Thus,wehypothesisedthat:

Hypothesis4a. whentheentrepreneurisfemaleandthecontact

ismale,thelikelihoodthattheentrepreneurmaybeprovidedwith

accesstofinancialresourcesishighercomparedtotheopposite

situation(namely,whentheentrepreneurismaleandthecontact

isfemale);

Hypothesis4b. whentheentrepreneurisfemaleandthe

con-tactismale,thelikelihoodthattheentrepreneurmaybeasked

forresourcesislowercomparedtotheoppositesituation(namely,

whentheentrepreneurismaleandthecontactisfemale).

2.2.4. Genderandurbanisation

Previousliteratureunderlinesthatthewiderculturalcontext

(urbanvs.ruralareas)–intermsofindividualistic/lesstraditional

culture(urbanareas)versusacollectivistic/moretraditional(rural

areas)–influencesrelationshipsandresourcesexchange(Rooks

etal.,2012,2016).Inacollectivistictraditionalculture,whichisstill

dominantinruralareas(Oysermanetal.,2002;Otiso,2006),

gen-derrelationshipsareparticularlypowerfulininfluencingeconomic

andsocial life(Lauras-Lecoh,1990;McKenzie,2011;Onjalaand

K’Akumu,2016;Stoeltje,2015;Stone,2013;Vossenberg,2016a,b).

Weexpectthatinurbanareas,wherethecultureislesstraditional

andcollectivistic(MaandSchoeneman,1997),theeffectofegoand

altergenderisweaker.Therefore,weformulatethefollowingfour

hypotheses:

Hypothesis5a. theeffectofegogenderonaccesstoresourcesis

weakerintheurbanareacomparedtotheruralarea;

Hypothesis5b. theeffectofaltergenderonaccesstoresourcesis

weakerintheurbanareacomparedtotheruralarea

Hypothesis5c. theeffectofegogenderonrequestsforresources

isweakerintheurbanareacomparedtotheruralarea;

Hypothesis5d. theeffectofaltergenderonrequestsforresources

isweakerintheurbanareacomparedtotheruralarea

3. Methods

To test our hypotheses, we conducted a survey amongst

Ugandan entrepreneurs. Uganda is a very interesting place to

studyentrepreneurship,sinceentrepreneurialactivityinUganda

is relatively high (Balunywa et al., 2012). Over one in three

(5)

Table1

Demographiccharacteristics:comparisonbetweentheurbanandtheruralarea,andbetweenmaleandfemaleentrepreneurs.

Totalsample Kampala(urban district) Nakaseke(rural district) T-test Male entrepreneurs Female entrepreneurs T-test Numberofobservations 608 294 314 – 313 281 – Individualcharacteristics Age(mean) 34.1 33.4 34.9 1.66 34.1 34.2 0.09 %ofmales 47 48 47 −0.29 – – –

Yearsofeducation(mean) 9 9.8 8.2 −4.38*** 9.6 8.5 −2.99***

Businesscharacteristics

Businessage(mean) 7.3 6.3 8.2 2.87** 8.2 6.4 −2.75**

Numberofemployees(mean) 1.3 1.5 1.2 −0.52 1.7 0.8 −5.84***

%ofbusinessesinthetradesector 50.2 50 50.3 0.08 41.4 57.4 3.92***

%ofbusinessesintheservicessector 30.8 32.7 29.2 −0.93 29.3 32.7 0.89

%ofbusinessesintheproductionsector 11.7 14.6 9 −2.17* 20.4 4.5 −6.11***

%ofbusinessesinagriculturesector 7.3 2.7 11.5 4.24*** 8.9 5.5 −1.65

* p<0.05. ** p<0.01. ***p<0.001.

entrepreneurshiprateishigheramongstwomen,intheyounger

partofthepopulation(18–34yearsold),aswellasinthe

better-educatedmembersofsociety(Balunywaetal.,2012).

3.1. Samplinganddatacollection

Asasamplingframe,we usedtheCensusof Businessesand

Establishments(COBE)2011provided bytheUganda Bureauof

Statistics (UBOS) – the most updated list available. The COBE

was conducted in 2010–2011 and covered all businesses with

fixedestablishments,irrespectiveof theirdegreeofformality –

UBOSworkedautonomouslyfromtheUgandaRevenueAuthority–

(UgandaBureauofStatistics,2011).Duringthefieldwork,theUBOS

teamphysicallymovedupanddownthestreetsandregisteredall

businesses.

We selected entrepreneurs from two sampling sites: an

urbandistrict and a rural district.Then, we randomlyselected

entrepreneursfromtwooftheCOBElists,onefortheurbandistrict

andoneforaruraldistrict.TheurbandistrictwasKampala,the

cap-italofthecountry,withapopulationofapproximately1,500,000.

TheruralsamplingsitewastheNakasekedistrict,whichislocated

intheCentralRegion(oneofthefouradministrativeregionsof

Uganda),atapproximately150kmfromthecapital.TheNakaseke

districthasapopulationofapproximately94,800.

Theresearchonwhichthearticleisbasedwaspartofan

overar-chingresearchprojectthatfocusedonurbanandruraldifferences.

Thechoiceofincludingbothurbanandruralentrepreneurswas

basedonpreviousliterature.Rooksetal.(2016)showedthatthere

weredifferencesbetweenurbanandruralareasinUganda

con-cerningsocialcapital.Theauthorsfoundthattheeffectofnetwork

densityonaccesstoresourceswasweakerintheruralareathanin

theurbanarea.

We decidedto focus ontheCentral Region since, according

toUBOSstatistics(UgandaBureauofStatistics,2011),thisisthe

regionwherethemajorityofbusinessesarelocated;indeed,30%

ofall businessesin Uganda are locatedinthat area(59%if we

alsoconsiderKampala).Kampalawasselectedgiventhat,asitis

thecountry’scapital,itisthemostimportantUgandancity,and

becauseitishometo29%ofallbusinessesinthecountry(Uganda

Bureauof Statistics, 2011). Tomaximise thevariation between

urbanandruralareas,weselectedNakasekebecauseitisoneof

themostruraldistrictsoftheCentralRegion.3

3 Forexample,theNakasekedistricthasoneofthelowerpopulationdensitiesof

theregion(source:elaborationfromUBOSdata,www.ubos.org).

ThedatacollectiontookplaceinJanuary2016.Weinterviewed

608respondents,294entrepreneursintheurbanarea(Kampala)

and314intheruralarea(Nakaseke).Inalmostallcases,theselected

respondentswerewillingtoparticipateinthestudy.InKampala

therewere9refusals,whileinNakasekeonlyonepersondeclined

toparticipate,makingforaresponserateof98.3%.Whena

per-sonrefusedtobeinterviewed,orwhenthebusinesswasnolonger

present,4wereplaceditwiththenearestavailableequivalent.5

Face-to-faceinterviewswerecarriedoutbyeightexperienced

interviewers. Theywere given a three-daytraining program to

helpthemunderstandthequestionnaireandfamiliarizewiththe

datacollectionsoftware(QuestionPro).Afterthetrainingsessions,

apilotcollectionwasundertakeninwhich16respondentswere

interviewed.

Respondentswereinterviewedontheirbusinesspremises.The

interviewswere sometimes interrupted,for instancewhen the

entrepreneurhad toattendtobusiness.Theylastedan average

of25–35min.Aftereach interview therespondentwasgiven a

notebookasatokenofappreciation.

3.2. Sample

Table1presentsthedemographiccharacteristicsofour

respon-dents.ConsistentwithtrendsintheUgandanpopulationpresented

intheGEM(GlobalEntrepreneurshipMonitor)report(Balunywa

etal., 2012), thesampleconsistsof a slightmajority offemale

entrepreneurs(53%);therespondentsare34yearsoldonaverage,

with9yearsofeducation.Respondentsintheurbansampleare

bet-tereducatedthantheircounterpartsintheruralsample(number

ofyearsofeducation:Murban=9.8;Mrural=8.2;t=−4.38,p<0.01).

Similarly, male entrepreneurs are better educated than their

femalecounterparts(numberofyears ofeducation:Mmale=9.6;

Mfemale=8.5;t=−2.99,p<0.01). We alsofoundthat 65% ofthe

urbansampleconsistedofentrepreneurswhowereborninrural

areasandwho,atsomepoint,decidedtomovetoKampala.

Theentrepreneursinoursampleownsmall,butrather

well-established, businesses. On average, they started about seven

years ago(four, ifwe considerthemedian) andtheyhave one

employee(Table1).TheseresultsareinlinewiththeUBOSfigures

4Thelistdisplayedthelocationofthebusiness,andnotthedescriptionofthe

businessinitself.Onlyafewtimes–inabout1%ofthecases–didwegotothe indicatedlocationandfindoutthattherewasnobusinesspresent.

5Wereplacedthebusinesswiththeclosestone,firstontheoppositesideofthe

street,and,ifnotpossible,onthesameside.Inthisway,webelievethatwereplaced theoldbusinesswithasimilarone,sincefrequentlybusinessesinthesameareahave similarcharacteristics(e.g.,size,macro-sector).

(6)

Table2

Businesssectors.ComparisonbetweensurveysampleandUBOSstatistics(percentages).

Totalsample Kampala(urbandistrict) Nakasekea(ruraldistrict)

SurveySample(2016) UBOS(2011) SurveySample(2016) UBOS(2011) SurveySample(2016) UBOS(2011)

Trade 50.2 61.5 50 60.6 50.3 59.7 Service 30.8 29.4 32.7 30.9 29.2 31.5 Production 11.7 7.3 14.6 8.2 9 6.4 Agriculture 7.3 1.8 2.7 0.3 11.5 2.3 Total 100 100 100 100 100 100 N 608 454,766 294 133,663 314 137,541 *p<0.05;**p<0.01;***p<0.001.

aThisdatareferstoCentralRegion(excludingKampala),giventhatUBOSdataonlyfromNakasekedistrictisnotavailable.

Table3

Networkcharacteristicsofthesampledividedintotwosamplingsites(mean).

Totalsample Kampala(urban district) Nakaseke(rural district) T-test Male entrepreneurs Female entrepreneurs T-test Wholenetwork(0–15) 4.9 4.8 5 0.75 5.3 4.6 −3.37***

Personal-advicenetworksize(0–5) 1.1 1.1 1.1 −0.30 1.3 0.9 −4.48***

Business-advicenetworksize(0–5) 2.5 2.5 2.5 0.78 2.7 2.4 −2.62**

Requestnetworksize(0–5) 1.3 1.2 1.4 1.15 1.3 1.3 0.03

Density(0–1) 0.7 0.6 0.7 3.76*** 0.6 0.6 1.52

%ofmales 56 54 57 1.05 69 42 −12.99***

%ofkin 50 47 53 1.49 43 59 6.01***

Homophily 55 55 55 0.15 64 49 −8.56***

Numberofpeoplementioned 2983a 1414 1569 1495 1426

aThenumberofcontactswhencomparingmaleandfemaleentrepreneursis2921. **p<0.01.

***p<0.001

(UgandaBureauofStatistics,2011),whichshowthatUgandan

busi-nessesarerathersmall(twoemployeesonaverage),andbetween

twoand fiveyears old.Thesizeof thebusinessdoesnotdiffer

betweenKampalaandNakaseke;however,businessesinKampala

are,onaverage,morerecentthanthoseinNakaseke(numberof

yearssincestart-up:Murban=6.3;Mrural=8.2;t=2.87,p<0.01).As

forgenderdifferences,businessesownedbymaleentrepreneurs

areonaverageolderthanthoseownedbyfemaleentrepreneurs

(numberofyearssincestart-up:Mmale=8.2;Mfemale=6.4;t=2.75,

p<0.01).

Mostofthebusinessesinoursampleareeitherintrade-related

industriesorservices.Together,theyrepresentmorethan80%of

thebusinessesinoursample.ThisisinlinewithUBOSdata(Uganda

BureauofStatistics,2011)showingthatthemajorityofbusinesses

inUgandaareinthetradesector.IfwecomparetheKampalaand

Nakasekedistricts,businessesinKampalaarelesslikelytobeinthe

productionandagriculturalsectorscomparedtothoseinNakaseke

(Table 1).Our sampleis consistentwithUBOSfigures

concern-ingbusinesssectors,includingwhensortingthesamplebetween

KampalaandNakasekedistricts(Table2presentsthecomparison

betweensurveysampleandUBOSfigures).Finally,comparedto

femaleentrepreneurs,maleentrepreneursarelesslikelyto

oper-ateinthetradesector(Mmale=41.4;Mfemale=57.4;t=3.9,p<0.001)

andmorelikelytooperateintheproductionsector(Mmale=20.4;

Mfemale=4.5;t=−0.6.1,p<0.001).

3.3. Questionnaire

Weusedthreenamegeneratorstomeasuretheentrepreneurs’

socialnetworks:twotoassesspersonal-adviceandbusiness-advice

networkties(together,theadvicenetwork),andonetoassessthe

number ofpeoplerequesting resourcesfromthe entrepreneurs

(therequestnetwork).Tomeasurethepersonal-advicenetwork,

weaskedthefollowingquestion:“Fromtimetotime,most

peo-plediscussimportantpersonalmatterswithotherpeople.Looking

backoverthelastsixmonths,whoarethepeoplewithwhomyou

havediscussedanimportantpersonalmatter?”Forthe

business-advicenetwork namegeneratorwe asked,“From time totime,

entrepreneursseekadviceonimportantbusinessmatters.

Look-ingbackoverthelastsixmonths,whoarethepeoplewithwhom

youhavediscussedanimportantbusinessmatter?”Tomeasurethe

requestnetworkweasked,“Lookingbackoverthelastsixmonths,

couldyoumentionthenamesofpeoplewhoaskedyouforfinancial

support,freegoods,servicesorajob?”

Foreveryoneofthethreenamegenerators,respondentswere

asked to list a maximum of five names (Burt, 1984).By

com-biningthethreenamegenerators,wecollected2983names,i.e.

approximately5peopleperrespondent.6Thepersonsmentioned

constitutedtheentrepreneur’s socialnetworks.Thesearerather

close-knit,especiallyinruralareas.

Foreachpersonidentified(contacts),weaskedabouttheir

gen-derandtheirrelationshipwiththeentrepreneur(relative,7friend

orjobcontact).8Mostofthecontactsweremaleand,largely,most

wererelatives.Thecomposition ofthenetworkswassimilarin

urbanandruralareas(Table3).Wealsomappedtherelationship

betweenaltersbyaskingtherespondent(ego),“Dothesetwo

per-sonsknoweachotherquitewell?9(Responsecategories:yes/no).

3.4. Dependentvariables:accesstofinancialresourcesand

requestsforfinancialresources

Foreverycontact,entrepreneurswereaskedtoindicatewhether

theycouldobtainfinancialresourcesfromthiscontactoriftheyhad

receivedrequestsforfinancialsupportfromthiscontact.

6Somecontacts(26%)werementionedmorethanonce.Wecountedthemasone. 7Weincludedbothcloseandextendedfamilyasrelatives.Wedefinedarelative

asapersonbelongingtothesamefamilyastherespondent.Wedefinedfamilyasa groupofpeoplerelatedbybloodormarriage.

8Wedecidedtofocusonlyongenderandrelationshipwithego,sincewewanted

toexplorein-depththeexchangeofresources(seealsobelow),forwhichweasked7 questionsforeachalter.Toavoidobtainingaquestionnairethatwouldbetoolengthy –sinceourrespondentswereclearlylosingattentionandwillingnesstoanswerafter 20min–,weonlyincludedthesepiecesofinformationasnameinterpreters.

9Duringtheinterviews,wemadeitclearthat‘knowquitewell’meantthatthey

(7)

Table4

Numberandpercentageofcontactswhoaskedforsupportfromtheentrepreneurs.

Totalsample (N=2983) Kampala(urban district,N=1414) Nakaseke(rural district,N=1569) T-test Male entrepreneurs (N=1492) Female entrepreneurs (N=1428) T-test Financialsupport 1707(57%) 798(56%) 909(58%) 0.8 887(59.5%) 799(56%) −1.94*

Freegoodsorservices 972(32%) 441(31%) 531(34%) 1.55 475(31.8%) 475(33.3%) 0.82

Job 143(5%) 94(7%) 49(3%) −4.51*** 98(6.6%) 45(3.2%) −4.29***

Atleastonekindofsupport 2125(71.3%) 1040(73.6%) 1085(69.2%) −2.68** 1120(75.1%) 973(68.1%) −4.20*** * p<0.05.

** p<0.01. ***p<0.001.

Table5

Numberandpercentageofcontactsthatareabletosupporttheentrepreneur.

Totalsample(N=2983) Kampala(urban district,N=1414) Nakaseke(rural district,N=1569) T-test Male entrepreneurs (N=1495) Female entrepreneurs (N=1428) T-test Financialsupport 1208(40.5%) 574(40.5%) 634(40.4%) −0.06 592(39.6%) 591(41.4%) 0.98 Information 1669(55.9%) 806(56.9%) 863(55%) −1.03 821(54.9%) 813(56.9%) 1.10

Introductiontootherpeople 665(22.3%) 290(20.5%) 375(23.9%) 2.25* 339(22.7%) 322(22.6%) −0.08

Freelabour 647(21.7%) 383(27%) 264(16.8%) −6.81*** 322(21.5%) 303(21.2%) −0.21

Atleastonekindofsupport 2633(88.2%) 1309(92.3%) 1324(84.4%) −6.80*** 1342(89.8%) 1236(86.6%) −2.69** * p<0.05.

** p<0.01.

First,theentrepreneurwasaskedwhattypeofresourcescould

beobtainedfromaspecificcontactviathequestion,‘What

sup-portfor the businesscan you getfrom this person?’Then, the

entrepreneurwasaskedaboutrequestsforresourcesonthepartof

thesespecificcontactsthroughthequestion,‘Whatkindofsupport

dideachcontactaskofyou?’

Following the definition of social capital, we did not ask

the entrepreneurs about actual resources received, but rather

aboutpotentialaccesstoresources(namely,resourcesthatcould

begained). However, note that, when it came to investigating

claimsfrom contacts,we asked whetherthecontact had

actu-allyrequestedresourcesinthepast.Earlierfieldworkconducted

bytheauthorsshowedthatentrepreneurswerepronetogiving

socially desirable answers when asked what kind of resources

thecontactmight obtainfromthem.Thus, wedecided torefer

to actual requests for resources. On the contrary, Rooks et al.

(2016)foundthatsocialdesirabilityissueswerelessproblematic

whenaskingwhattypeofresourcescouldbeobtainedfroma

con-tact.

Wecreatedtwovariables:(1)accesstoresources:this

dichoto-mousvariableindicatesthatacontactisabletoprovidefinances;

(2)requestsforresources:thisdichotomousvariableindicatesthat

theentrepreneurhasbeenaskedbythecontactforfinancial

sup-port.

3.5. Independentvariables

3.5.1. Gender(alterlevel)

Toaccountforgenderdifferences,weincludedthegenderofthe

contacts(‘0’forfemaleand‘1’formale).

3.5.2. Gender(egolevel)

To account for the different network composition between

maleand femaleentrepreneurs, we includedthegender ofthe

entrepreneur(‘0’forfemaleand‘1’formale).

3.5.3. Homophily(ego-alterlevel)

Toaccountforsituationswheretheentrepreneurandthe

con-tactarethesamegenderweincludedthevariablehomophily(‘0’

fordifferentgenderand‘1’forsamegender).

3.6. Controlvariables

3.6.1. Kinship(alterlevel)

Eachcontactwasclassifiedbasedonwhetherornotheorshe

wasa relative(includingpartner/spouse).We createdadummy

variablelabelled‘relative,’usingallothercategories(friendshipand

business-onlyrelationship)asareference.

3.6.2. Urbanisation(egolevel)

To account for differences between urban and rural areas

we inserted an urbanisation variable, indicating whether the

entrepreneurwaslivingintheurbanorruralarea(reference

cate-gory).

3.6.3. Networksize(egolevel)

Toaccountfordifferencesinnetworksize(WellmanandFrank,

2001),weaddedtwoadditionalvariablestocontrolfortheimpact

ofnetworksize:[1]advicenetworksize,whichisthetotalnumberof

contactsmentionedinthepersonalandbusinessadvicenetworks,

and[2]requestnetworksize,namelythenumberofuniquecontacts

mentionedbytherespondentinthe‘requests’namegenerator.

3.6.4. Density(egolevel)

Previousresearchunderlinesthefactthatnetworkdensitymay

influencetheexchangeofresources(Burt,2001;ShaneandCable,

2002).Densityshowshowcloselyanetworkofrelationshipsisknit

and,morespecifically,howwellanentrepreneur’scontactsmight

knoweachother.Wecalculatedthevariabledensityasthenumber

ofactualties,outofthenumberofpossibletiesinthenetwork

(namely,ifeverycontactmentionedhadarelationshipwithevery

othercontactmentioned).

3.6.5. Yearsofeducation(egolevel)

Weincludedthenumberofyearsofeducationasavariableto

controlfortheconfoundingeffectsofhumancapital,sincehigher

levelsofhumancapitalaregenerallyassociatedwithgreatersocial

resources(vanTilburg,1998).

3.6.6. Age(egolevel)

Agemayaffectnetworkcompositionandsocialsupport(Moore,

(8)

Table6

CorrelationsbetweenVariables.

1 2 3 4 5 6 7 8 9 10 11

1 Accesstoresources 1

2 Requestforresources −0.53*** 1

3 Urbanisation 0.00 0 1

4 Kinship 0.01 0.12*** −0.06* 1

5 Gender(alter) 0.08*** 0.07*** −0.03* −0.18*** 1

6 Homophily −0.06** 0.01 −0.01 −0.24*** 0.00 1

7 Request-networkcontact −0.39*** 0.43*** −0.02 0.18*** −0.05** 0.02 1

8 Advicenetworksize 0.12*** −0.04* −0.01 −0.18*** 0.05** 0.09*** −0.13*** 1

9 Requestnetworksize −0.17*** 0.15*** −0.05 0.06** 0.00 −0.02 0.38*** 0.19*** 1

10 Density 0.01 0.01 −0.18*** 0.16*** −0.03 0.01 0.03 0.27*** 0.21*** 1 11 Gender(ego) −0.02 0.03 0.01 −0.05** 0.27*** 0.15*** −0.02 0.18*** 0.00 −0.06 1 12 Age −0.02 −0.01 −0.07 0.08*** 0.01 0.00 0.04* 0.00 0.06 0.01 0.00 13 Yearsofeducation 0.07** 0.01 0.18*** −0.09*** 0.01 0.00 −0.06** 0.12* −0.06 −0.03 0.12* 14 Maritalstatus 0.03 0.03 −0.16*** 0.01 0.02 −0.02 0.01 −0.07 0.08* 0.01 −0.20*** 15 Numberofchildren 0.02 −0.01 −0.20*** 0.07*** −0.01 −0.01 0.03 0.04 −0.09 0.05 0.03 16 Businesssize −0.04* 0.00 0.02 −0.07*** 0.08*** 0.05** −0.03 0.21*** 0.05 0.09* 0.25*** 17 Businessage −0.03 −0.01 −0.12* 0.02 0.05* 0.01 0.01 0.08 0.02 0.08* 0.11** 18 Sector:Production 0.03 0.02 0.08* 0.04* 0.07*** 0.05* 0.05* 0.05 0.12** 0.02 0.24*** 19 Sector:Service −0.02 −0.01 0.04 0.07 0.06 −0.01 −0.02 0.00 −0.03 0.02 −0.04 20 Sector:Trade 0 0.0 0.00 −0.05* −0.06** −0.03 −0.01 −0.06 −0.02 −0.05 −0.16*** 21 Sector:Agriculture 0.01 −0.04** −0.17*** 0.02 0.01 −0.01 −0.03 0.04 −0.05 0.05 0.07 12 13 14 15 16 17 18 19 20 21 12 Age 1 13 Yearsofeducation −0.35*** 1 14 Maritalstatus 0.46*** −022*** 1 15 Numberofchildren 0.60*** −0.32*** 0.33*** 1 16 Businesssize −0.01 0.08 −0.07 0.06 1 17 Businessage −0.64*** −0.34*** 0.25*** 0.51 0.12** 1 18 Sector:Production 0.02 −0.03 −0.02 0.01 0.17*** 0.14*** 1 19 Sector:Service −0.03 0.04 −0.04 0.03 0.05 −0.10** −0.24** 1 20 Sector:Trade −0.05 0.07 0.04 −0.13** −0.27*** 0.14 0.37*** −0.37*** 1 21 Sector:Agriculture 0.12*** −0.15*** 0.02 0.20*** 0.23*** 0.27*** −0.11** −0.19** −0.28*** 1

Note:whenthecorrelationisbetweentwovariablesattheegolevel(e.g.densityandage),thecorrelationswerecalculatedatanegolevel.Otherwise,correlationswere calculatedatanalterlevel.

*p<0.05. **p<0.01. ***p<0.001.

theageoftherespondent–specificallytherespondent’sexactage

atthetimeoftheinterview–asacontrolvariable.

3.6.7. Maritalstatus(egolevel)

Giventheimportanceofintra-householdrelations(ILO,2017;

Vossenberg, 2013), we included the marital status of the

entrepreneur(‘0’forunmarriedand‘1’formarried).

3.6.8. Numberofchildren(egolevel)

Toaccountforfamilycomposition,wealsoincludednumberof

children.

3.6.9. Businesssize(egolevel)

Asaproxyforbusinesssuccess(Freseetal.,2007),weincluded

businesssize(i.e.numberofemployees).Weinsertedthe

logarith-micversioninthemodelbecausetheoriginalvariableresultedin

askeweddistribution.

3.6.10. Businessage(egolevel)

Tocontrolforthedatethatthebusinesswasstarted,weincluded

theageofthebusiness(innumberofyears).Weinsertedthe

loga-rithmicversioninthemodelbecausetheoriginalvariableresulted

inaskeweddistribution.

3.6.11. Sector(egolevel)

Tocheckforanyeffectstemmingfromsectordifferences,we

createdthreevariables:[1]production(whetherornotabusinessis

inthemanufacturingorconstructionsector);[2]services(whether

ornotabusinessisintheservicessector)[3]trade(whethera

busi-nessisretailorwholesale).Thereferencecategoryisagriculture

(whetherornotabusinessisintheagriculturalsector).

3.6.12. Request-networkcontact(alterlevel)

Tocontrolfortheeffectofwhichnamegeneratorthecontact

hadbeenmentionedin,weincludedavariableindicatingwhether

thecontactwasmentionedintheadvicenetwork(‘0’)orinthe

requestnetwork(‘1’).

4. Results

Inthissection,wefirstillustratethedescriptiveresultsofour

research,focusingonthedifferentformsofsupportaskedforand

accessed.Then,inordertotesttheabove-formulatedhypotheses,

wepresenttheresultsoftwoseparatemultilevellogistic

regres-sions.

4.1. Descriptiveanalyses

Table 3 displays the results of the descriptive analyses.

Entrepreneurs have a limited number of people in their

net-works (fewer than five people, on average). On average,

maleentrepreneursmentioned significantlymorecontactsthan

femaleentrepreneurs,(Mmale=5.3Mfemale=4.6;t=−3.4,p<0.001).

The same holds for the personal-advice network (Mmale=1.3

Mfemale=0.9;t=−4.5,p<0.001)andthebusiness-advicenetwork (Mmale=2.7Mfemale=2.4;t=−2.6,p<0.01).Bycontrast,maleand

(9)

Table7

MultilevelLogisticRegressionAnalysis.

Accesstoresources Requestsforresources

ModelI ModelII(ModelI+

interactionego gender*altergender)

ModelI ModelII(ModelI+

interactionego gender*altergender)

␤ SE ␤ SE ␤ SE

Context(entrepreneur)

Urbanisation 0.00 0.21 0.00 0.21 0.03 0.19 0.03 0.19

Relationallevel(alter)

Kinship 0.72*** 0.13 0.72*** 0.13 0.59*** 0.13 0.59*** 0.13

Gender(alter) 0.52*** 0.13 −0.41*** 0.13

Request-networkcontact −3.23*** 0.20 −3.23*** 0.20 2.94*** 0.16 2.94*** 0.16

Homophily −0.44*** 0.13 0.16 0.13

Networkcharacteristics(entrepreneur)

Advicenetworksize 0.14** 0.05 0.14* 0.05 −0.04 0.05 −0.04 0.05

Requestnetworksize −0.06 0.07 −0.06 0.07 0.08 0.70 0.08 0.70

Density 0.20 0.29 0.20 0.29 0.14 0.28 0.14 0.28

Individuallevel(entrepreneur)

Gender(ego) −0.45* 0.22 0.23 0.3 Age −0.01 0.01 −0.01 0.01 −0.01 0.01 −0.01 0.01 Yearsofeducation 0.03 0.03 0.03 0.03 0.05 0.04 0.05 0.04 Maritalstatus −0.01 0.24 −0.01 0.24 0.34 0.23 0.34 0.23 Numberofchildren 0.35 0.23 0.35 0.23 −0.18 0.21 −0.18 0.21 Firmlevel Businesssize −0.44** 0.16 −0.44** 0.16 0.15 0.15 0.15 0.15 Businessage −0.17 0.15 −0.17 0.15 0.04 0.14 0.04 0.14 Sector:Production 0.38 0.49 0.38 0.49 −0.43 0.45 −0.43 0.45 Sector:Service −0.42 0.44 −0.42 0.44 −0.19 0.40 −0.19 0.40 Sector:Trade −0.46 0.44 −0.46 0.44 −0.14 0.40 −0.14 0.40 Interactions Female(alter)*Female(ego) – – 0.01 0.26 – – −0.08 0.24 Male(alter)*Female(ego) – – 0.97*** 0.26 −0.65** 0.26 Male(alter)*Male(ego) – – 0.07 0.18 – – −0.25 0.18 Constant −1.18 0.69 −0.63 0.69 −2.74*** 0.65 −2.50*** 0.66 Nobservations 2548 2548 2548 2548 Nentrepreneurs 491 491 491 491 SD(u) 1.65 1.65 1.39 1.39 Loglikelihood −1327.66 −1327.66 −1200.52 −1200.52 Wald(df) 317.25***(18) 317.25***(18) 401.65***(18) 401.65***(18) * p<0.05. ** p<0.01. ***p<0.001.

Themajorityofcontactsaremales(onaverage,56%ofnetwork

contacts).Homophilyisadrivingfactorwhenitcomestoincluding

apersoninthenetwork,giventhat55%ofthecontactsinthe

net-workareofthesamegender.Comparedtofemaleentrepreneurs,

maleentrepreneurshaveahigherpercentageofmales(Mmale=69

Mfemale=42;t=−13.0,p<0.001),andalowerpercentageof

rela-tives(Mmale=43Mfemale=59;t=−6.0,p<0.001),intheirnetworks.

Entrepreneurs’networksareratherclose-knit–densityisequal

to0.6–,meaningthatthepeoplewithinthenetworkarelikelyto

knoweachother.Thereisnodifferencebetweenmaleandfemale

entrepreneurs.Networksofentrepreneursintheruralareaandin

theurbanaresimilar,exceptfornetworkdensity.Entrepreneursin

theruralareahavedensernetworksthanthoseintheurbanarea

(Murban=0.6Mrural=0.7;t=−3.8,p<0.001).

Tables4and5presentthenumberofrespondentswhoeither

provided access toresources, or requested resources from the

entrepreneurs.Inapproximately 70%of cases,thecontacts had

askedforsupportfromtheentrepreneur.Financialsupportisthe

mostoftenrequested(57%).88%ofthecontactswereableto

pro-videtheentrepreneurwithsomeformofsupport,althoughless

thanhalfwereabletoprovidefinancing.

Maleentrepreneurshave(onaverage)ahigherpercentageof

contactsintheirnetworkwhohadaskedforatleastoneformof

support(Mmale=75.1Mfemale=68.1;t=−4.2,p<0.001)compared

tofemaleentrepreneurs.Thisdifferenceholdsforfinancialsupport

andrequestsforajob,butnotrequestsforfreegoodsandservices.

Similarly,male entrepreneurshave a higherpercentage of

con-tactsintheirnetworkwhoprovidethemwithaccesstoresources

(Mmale=89.8Mfemale=86.6;t=−2.7,p<0.001)incomparisonwith

femaleentrepreneurs.However,nosignificantdifferencesemerge

foreachformofsupport.

Asforurbanandruralentrepreneurs,thoseintheurbanarea

haveahigherpercentageofbothcontactswhohadaskedforatleast

oneformofsupport(Murban=92.3Mrural=84.3;t=−6.8,p<0.001)

and contactswho had provided them withaccess toresources

(Murban=73.6Mrural=69.2;t=−2.7,p<0.01).

4.2. Mainanalysis

Inordertotestourhypotheses,werantwoseparatemultilevel

logisticregressions.Ourdataconsistsofmultipletiesper

respond-ingentrepreneur,andsoitischaracterizedbyanestedstructure.To

dealwiththenestedstructureofthedata,weappliedamultilevel

logisticregressionmodel(SnijdersandBosker,1999).Table6shows

thecorrelationsofthevariablesincludedinthemodels.The

cor-relationsbetweenindependentvariablesaregenerallylow,apart

fromthosebetweenageandnumberofchildren,ageandbusiness

age,businessageandnumberofchildren.However,sincethese

highvaluesofcorrelationswerenotaboutthethreemain

inde-pendentvariables(egogender,altergenderandhomophily),we

decidedtokeepthemascontrolvariables.Furtheranalysis(not

reportedhere)showsthattheeffectsdidnotchangewhenweran

themodelswithoutnumberofchildrenandbusinessage(i.e.the

(10)

Table8

MultilevelLogisticRegressionAnalysis–Interactionsbetweenurbanisation,andegoandaltergender.

Accesstoresources Requestsforresources

ModelIII(ModelI+ interactionego gender*urbanisation)

ModelIV(ModelI+ interactionalter gender*urbanisation)

ModelIII(ModelI+ interactionego gender*urbanisation) ModelIV(ModelI+ interactionalter gender*urbanisation) ␤ SE ␤ SE ␤ SE Context(entrepreneur) Urbanisation 0.20 0.28 0.13 0.25 −0.33 0.23 −0.26 0.23

Relationallevel(alter)

Kinship 0.71*** 0.13 0.72*** 0.13 0.59*** 0.13 0.59*** 0.13

Gender(alter) 0.52*** 0.13 0.63*** 0.17 −0.42*** 0.13 −0.68*** 0.17

Request-networkcontact −3.23*** 0.2 −3.23*** 0.20 2.94*** 0.16 2.94*** 0.16

Homophily −0.44*** 0.13 −0.43*** 0.13 0.17 0.13 0.15 0.13

Networkcharacteristics(entrepreneur)

Advicenetworksize 0.14** 0.05 0.14** 0.05 −0.03 0.05 −0.04 0.05

Requestnetworksize −0.06 0.07 −0.06 0.07 0.07 0.70 0.08 0.70

Density 0.20 0.29 0.20 0.29 0.14 0.28 0.14 0.28

Individuallevel(entrepreneur)

Gender(ego) −0.27 0.27 −0.45* 0.22 −0.09 0.26 0.23 0.20 Age −0.01 0.01 −0.01 0.01 −0.01 0.01 −0.01 0.01 Yearsofeducation 0.03 0.03 0.03 0.03 0.03 0.02 0.05 0.04 Maritalstatus −0.02 0.23 −0.01 0.24 0.36 0.26 0.34 0.23 Numberofchildren 0.36 0.23 0.35 0.23 −0.19 0.21 −0.19 0.21 Firmlevel Businesssize −0.44** 0.17 −0.44** 0.16 0.13 0.15 0.14 0.15 Businessage −0.17 0.15 −0.17 0.15 0.04 0.14 0.05 0.14 Sector:Production 0.45 0.49 0.39 0.49 −0.46 0.45 −0.46 0.45 Sector:Service −0.37 0.44 −0.42 0.44 −0.29 0.40 −0.22 0.40 Sector:Trade −0.39 0.44 −0.46 0.44 −0.25 0.40 −0.16 0.40 Interactions egogender*urbanisation −0.41 0.39 – – 0.75* 0.36 altergender*urbanisation – – −0.24 0.24 – – 0.56* 0.25 Constant −0.34 0.70 −0.26 0.69 −2.46*** 0.65 −2.56*** 0.65 Nobservations 2548 2548 2548 2548 Nentrepreneurs 491 491 491 491 SD(u) 1.65 1.65 1.38 1.39 Loglikelihood −1327.12 −1327.66 −1198.42 −1197.86 Wald(df) 317.63***(19) 317.25***(19) 402.80***(19) 403.37***(19) *p<0.05. **p<0.01. ***p<0.00.

Theresultsofthemultilevellogisticregressionsarepresented

inTable7.InModel1ofTable7,wetestedHypotheses1–3,

formu-latedinthetheorysection.

Hypotheses1aand1baddressegogender.Hypothesis1ais

sup-ported bythe results, which show thatcontacts are less likely

toprovideaccess tofinancialresourceswhen theentrepreneur

ismale(␤=−0.48;p<0.01).Thisimpliesthattheoddsofamale

entrepreneurhavingaccesstofinancialresourcesareabout36%

lowerthan thatof afemale entrepreneur.Hypothesis1b states

thatcontactsaremorelikelytoaskforfinancialresourceswhen

theentrepreneurismale.Thishypothesisisnotsupportedbyour

results.

OurresultssupportHypotheses2aand2b,whichrefertoalter

gender.Amalecontactismorelikelytoprovideaccesstofinancial

resources(␤=0.65;p<0.001),asstatedbyHypothesis2a,andless

likelytoaskforfinancialsupport(␤=−0.28;p<0.01),asstatedby

Hypothesis2b.Thisimpliesthattheoddsthatamalecontactwill

providefinancialsupportare68%higherthanafemalecontact.The

oddsofamalecontactrequestingfinancialsupportare34%lower

comparedtoafemalecontactaskingforsupport.

Hypotheses3aand3baddressgenderhomophily.The

hypothe-ses stated that, when the entrepreneur and the contact are

of the same gender, the contact is less likely to both provide

accesstofinancialresourcesandalsorequestresourcesfromthe

entrepreneur.Our resultsonly supportHypothesis3athat

gen-derhomophilynegativelyaffects accesstoresources(␤=−0.43;

p<0.001).Thisimpliesthatwhenentrepreneurandcontacthave

thesamegender,theoddsthatacontactprovidesfinancialsupport

are35%lower.Hypothesis3bisnotsupportedbyourresults.

In Model 2 (Table7), we takeinto account theinteractions

betweencontactgenderandentrepreneurgender.Inordertotest

Hypotheses4aand4b,weinsertedallpossibleinteractions,10

tak-ingthecombination offemale(contact)-male(entrepreneur)as

a referencecategory.Asfor Hypothesis4a,wefoundthatmale

contacts provided a female entrepreneur withaccess to

finan-cialresourcesmoreoftenthanafemalecontactprovidesamale

entrepreneurwithaccess(␤=0.97;p<0.001).Asforrequestsfor

resources(Hypothesis4b),incomparisonwiththereference

situa-tion(i.e.,femalecontactandmaleentrepreneur),amalecontact

is less likely to ask for resources from a female entrepreneur

(␤=−0.65;p<0.01).Therefore,bothHypotheses4aand4bare

sup-ported.

Asforcontrolvariables,themostinterestingresultsregard

kin-ship.Relativesaremorelikely toprovidefinancialsocialcapital

(␤=0.84;p<0.001).However,ifthecontactisarelative,heorsheis

morelikelytoaskforfinancialresourcesaswell(␤=1.13;p<0.001).

Theresultsalsoshowthaturbanisationdoesnotplayarolein

eitheraccesstofinancialresourcesorrequestsforresources.

Fur-thermore,inpartialcontrastwithotherliteratureonthetopic(e.g.

Bhagavatulaetal.,2010;Rooksetal.,2016;ShaneandCable,2002),

10Homophilywasnot insertedinthemodelforobviousproblems of

(11)

densityhasnoeffectonaccesstofinancialresources.Similarly,

despitethefactthatclose-knitnetworkshavealsobeen

associ-atedwithnegativeoutcomes(seeforexample,Shinnaretal.,2011),

densitydoesnotaffectrequestsforfinancialresources.

To test whether theeffect of ego and alter gender differed

betweentheurbanandtheruralarea(Hypotheses5a–d),weran

twomodels(Table8).Oneincludedtheinteractioneffectsbetween

urbanisation and ego gender(Table8, Model 3), and theother

includedtheinteractionbetweenurbanisation andalter gender

(Table 8, Model 4).11 Running the modelswith theinteraction

effects,wefoundthatthevaluesoftheinteractionswere

signif-icantonlyfor‘requestsforresources’.Therefore,bothHypothesis

5aandHypothesis5b,whichrefertoaccesstoresources(inthe

urbanincomparisonwiththeruralarea),arenotconfirmed.

Asforrequestsforresources,inourresults(Table8,ModelIII),

whentheentrepreneurismaleandlivesintheurbanarea,theeffect

onrequestsforresourcesispositive(combinedeffect:0.33)and

strongerthantheeffectoftheentrepreneurbeingmaleintherural

area(␤=−0.09).ThisresultcontradictsHypothesis5c,whichstated

thattheeffectofegogenderonrequestsforresourceswasweaker

intheurbanareathanintheruralarea.

Thesamereasoningappliestotheinteractionbetweenalter

gen-derandurbanisation(Table8,ModelIV).Whenthecontactismale

andlivesintheurbanarea, theeffectonrequestsforresources

ispositive (combinedeffect: 0.53) andstronger than theeffect

ofthecontactbeingmaleintheruralarea(␤=0.23).Therefore,

Hypothesis5discontradicted.

Overall,ourhypothesesthatgenderislesscriticalintheurban

areathanintheruralareaarenotsupported.Actually,theresults

showtheoppositewhenitcomesto‘requestsforresources’;the

effectofegoandaltergenderisstrongerintheurbanareathanin

theruralarea.

5. Discussionandconclusions

Inthisarticle,weaddressedgenderdifferencesinsocialcapital

amongUgandanentrepreneurs.Wefoundthatgenderinfluences

thelikelihoodofaccessingfinancialcapitalthroughcontactsand

receivingrequestsoffinancialsupportfromthesecontacts.

Specif-ically,whentheentrepreneurisaman,hiscontactsarelesslikelyto

provideaccesstofinancialresources.Asforthegenderofthe

con-tact,femalecontactsaremorelikelytoaskforfinancialsupport,

andtheyarealsolessabletoprovideentrepreneurswithfinancial

support.Furthermore,malecontactsaregenerallymorelikelyto

provideaccesstofinancialresourcestofemaleentrepreneursthan

theotherwayaround(femalecontactstomale entrepreneurs).

Asituationinwhichamalecontactrequestsfinancialresources

fromafemaleentrepreneurislesslikelytohappenthanasituation

whereafemalecontactasksamanforresources.

Ourresultssuggestthatfemaleentrepreneursgenerallyhave

lessfinancialpowerthanmen,duetodifficultiesinaccessing

finan-cialresourcesvialoansbyformalinstitutions,andalackofcontrol

overfinancesinthehousehold(Aidisetal.,2007;Demirgüc-Kunt

etal., 2015;Fletschner, 2009; Minniti, 2010; ILO, 2017; Jamali,

2009;Vossenberg,2016b;WorldBank,2015).Asaconsequence,

mostofthetime,requestingmoneyfromtheirhusbandsortheir

malerelativesseemstheonly optionfor womenentrepreneurs

whowishtoobtainfinancialsupport(Jamali,2009;Vossenberg,

2016a,b).

Wealsocomparedgenderdifferencesinurbanandruralareas.

Our resultssuggest that urbanisation doesnot matterwhen it

comestoexchangingresources.Urbanisationisnotassociatedwith

11 Fortheinterpretationofinteractioneffects,see:GrotenhuisandThijs(2015).

eitheraccesstoresourcesorrequestsforresources.Ourhypothesis

thatgenderislesscriticalinurbanareascomparedtoruralareas

isnotsupported.Apossibleexplanationissuggestedbythe

liter-atureonrural-urbanmigration(Mukwayaetal.,2011;Belletal.,

2015).Giventhehighnumberofpeoplemovingfromruraltourban

areasinUganda,culturaldifferencesbetweenurbanandruralareas

aremoreorlessblurred.Therelevanceofthisphenomenonisalso

confirmedinourstudy,since65%oftheentrepreneursinoururban

samplewerebornoutsideKampala.

Ourresearchconfirmsthattheroleofthefamilyiscriticalfor

entrepreneurshipindevelopingcountries.Inlinewithsome

pre-viousstudies(Arregleetal.,2015;BerrouandCombarnous,2012;

Egbert,2009;HoffandSen,2016),thisstudymakesitclearthatthe

roleofthefamilyisnotalwayspositive.Ontheonehand,relatives

aremorelikelytoprovideaccesstofinancialresources,butonthe

otherhand,theyarealsomorelikelytoaskforfinancialsupport.

Lastbutnotleast,ourfindingsshowthatasignificantpartof

aUgandanentrepreneurs’networkscanprovetobealiability.For

entrepreneursinUganda,beingembeddedinanetworkofrelations

oftenimpliesthattheyareexpectedtosupporttheircontacts.

Thisarticlecontributestotheliteratureonsocial capitaland

entrepreneurshipbyconnectingthesewithgenderandfocusingon

requestsforresources.Firstly,ourstudycontributestofillingthe

gapregardingdifferencesbetweenmaleandfemaleentrepreneurs’

socialcapitalindevelopingcountries,atopicthatcertainlydeserves

moreattention(Lindvertetal.,2017;Myroniuk,2016).Secondly,

thearticleconstitutesoneofthefirstattemptstoidentifythe

deter-minantsofrequestsforresources.Inparticular,ourresultsshow

thatthere isa seriousissuerelatedtothepressureenduredby

entrepreneurs,whoareexpectedtoprovidesupporttosomeof

theircontacts.Ourresearchsuggestspossiblelimitationspresentin

previousliterature,whichhasoftenadoptedanoveremphaticview

ofsocialnetworksintermsofsocialcapital(see:O’Brien,2012;

Portes,1998).

Theresultsofthisarticlearelimited,whichleavesroomfor

fur-therresearch.First,sincewestudiedUgandanentrepreneurs,we

arenotsurewhetherornot,andtowhatextent,ourresultsmight

begeneralisedtoothercontexts.Inthisregard,apossiblefuture

researchavenuemightbetoconductsimilarresearchstudiesin

otherdeveloping countries, or in developedWestern countries.

Combinedwithourresearch,thiscouldshedfurtherlightonthe

determinantsofrequestsforresourcesfromthecontactsthatmake

upanentrepreneurs’socialcapital.

Second,weemployedadichotomousdefinitionoftheexchange

ofresourcesbetweentheentrepreneur(ego)andeachhisorher

contacts(alter):contactsareabletoprovidefinancialsupportor

theydonot;contactsaskforfinancialsupportortheydonot.Inthis

way,wedidnottakeintoaccountpossibledifferencesconcerning

theamountoffinancialsupportextendedorrequested.

Third,inordertolimitthelengthoftheinterviews,wecollected

limitedinformationconcerningcontactcharacteristics.However,

moredetailedinformationcanbegathered,suchaswhetherornot

thepersonwastherespondent’spartner;thecontact’sage;the

contact’sjobposition.Otherpossibleinformationtocollectmight

bethesocial statusofthecontacts,assuggestedbyBerrouand

Combarnous(2011).Indeed,thecontact’ssocialstatusmight

influ-encethelikelihoodofhimorherbeingabletoprovidefinancial

supportoraskforit.

Lastly,wefocusononeruralcontextonly,butruralareasmay

welldifferfromoneanother.TheNakasekeareais(relatively)close

tothecountry’scapital.Therefore,itisnotclearhowwellthisarea

representsotherruralareasthataremoredistantfromthecapital.

Furtherstudiescouldthereforeexplorevariationsbetweenrural

areaswhenitcomestotheentrepreneur’ssocialcapital.

Inconclusion,thisarticlehasillustratedthattheeffectofsocial

Referenties

GERELATEERDE DOCUMENTEN

De klas zag in dat dit niet het geval was, en zo viel het kwartje dat het parkeren van de auto’s blijkbaar op meer manieren kan dan het uitkiezen van de leerlingen: bij het

Because our framework is based around ratings, the framework can- not be instantiated with recommender systems that are not based on ratings (i. demographic and

TAB L E 2 (Continued) Approach Experimental paradigm Minimum ligand set necessary Minimum number of experimental conditions Results confirming the presence of high ‐affinity state

South Africa is becoming a classroom of the world’s future managers if the increas- ing number of overseas students who attend programmes at the University of Stellenbosch

While through the social identity theory and the human capital theory, the types of ties (bonding or bridging) that ultimately affect firm performance can be explained, through

There is an econometric model developed to test which factors have an influence on the capital structure of firms. In this econometric model, one dependent variable should be

While overall appearance and communicative behavior of these robots was carefully designed to study human-robot interaction, visitors still liked to block the path of the robots,

After controlling for commonly known variables influencing the dependent variable cost of capital, the results show that there is a significant negative correlation