Microsoft Word - Ghana2012-02-28.doc

FamilyTies,InheritanceRights,andSuccessfulPoverty

Alleviation:EvidencefromGhana



EdwardKutsoati

andRandallMorck

b



Thisdraft:February29,2012



Abstract



Ghanaian custom views children as members of either their mother’s or father’s lineage

(extendedfamily), butnotboth.Patrilineal  customchargesaman’slineagewith caringforhis

widow and children, while matrilineal custom places this burden on the widows’ lineage – her

father,brothers,anduncles.Deemingcustominadequate,andtopromotethenuclearfamily,

Ghanaenactedthe IntestateSuccession(PNDC)Law 111,1985and1998 Children’s Act560to

force men to provide for their widows and children, as in Western cultures. Our survey shows

that, although most people die intestate and many profess to know Law 111, it is rarely

implemented. Knowledge of the law correlates with couples accumulating assets jointly and

with inter‐vivos husband to wife transfers, controlling for education. These effects are  least

evident for widows ofmatrilineal lineage men, suggesting a persistence of traditional norms.

Widowswithclosertieswiththeirownortheirspouse’slineagereportgreaterfinancialsupport,

asdothoseveryfewwhobenefitfromlegalwillsoraccessLaw111and,importantly,widowsof

matrilineallineage.SomeevidencealsosupportsAct560benefitingnuclearfamilies,especially

if the decedent’s lineage is matrilineal. Overall, our study confirms African traditional

institutions’persistentimportance,andthelimitedeffectsofformallaw.

 We are deeply grateful to Kofi Awusabo‐Asare for coordinating our household survey and his numerous

suggestions along the way.We are also gratefulto the managementofSocialSecurity andNational Insurance

Trust (SSNIT) Fund of Ghana for providing access to individual records; and the team of research assistants:

Yvonne Adjakloe, Eugene Darteh and Kobina Esia‐Donkoh for their superb field work; William Angko and Alex

Larbie‐MensahforthepainfultaskofcollectingtheSSNITdata;andGreatjoyNdlovuandPanosRigopoulosfor

excellentassistancewiththedataanalysis.WethankparticipantsattheNationalBureauofEconomicResearch



(NBER)/Africa Project meetings in Cambridge and Zanzibar for their comments; and gratefully acknowledge

financialsupportfromtheNBERandtheSSHRC.Theviewsexpressedhere howeveraresolelyours anddonot

reflectthoseoftheNBERortheSSNIT.

 Department of Economics, Tufts University, Medford, MA USA 02155. Phone: +1 (617) 627‐2688.Email:

[email protected]

 Distinguished University Professor and Stephen A. Jarislowsky Distinguished  Chair in Finance, University of

Alberta School of  Business, Edmonton CANADA T6G2R6; Research Associate, National Bureau of Economic

Research.E‐mail:[email protected]

.Phone:+1(780)492‐5683.

1. Introduction

In much of sub‐Saharan Africa, the idea of a family extends beyond its conjugal members. A

lineage, or extended family, is a far larger web of relationships in which all members have a

commonancestor,eithermaleor female.One’srelationshipwithmembersofone’sextended

familymaybeasimportantas,andinsomecases,moreimportantthan,one’srelationshipwith

one’s spouses and children. Historically, lineages are bastions of emotional and financial

support (Plateau, 1991). Lineages can pay for education and training, and their social safety

netscansupport risk‐taking and entrepreneurship.However,expectationsofbeingsupported

by, and of having to support, members of one’s lineage can also deter human capital

accumulation, labor supply, entrepreneurship and risk taking.

 The actual importance of

extendedfamiliesinanygivencontextisthusanempiricalquestion.

This study explores how inheritance rules in two distinct Ghanaian systems of defining

extended family membership interact with formal legal inheritance rules to affect asset

accumulationduringmarriageandtheeconomicsituationsofwidowsandtheirchildren.Until

1985,intestateinheritancesweredeterminedbytraditionalcustom,andthisdependedonhow

one’s extended family was defined.People whose tribal customs are matrilineal define their

lineages through their female bloodlines only:their mothers and maternal cousins, aunts,

uncles, grandparents, and so on are their blood kin; but their fathers and paternal cousins,

aunts, uncles, grandparents, and so on are not.People whose tribal customs are patrilineal

analogously define blood kinship as flowing through their paternal, but not maternal

bloodlines.Undermatrilineal lineagenorms,aman’schildrenarethusnothisbloodkin,and

hisheirshouldhe

dieintestate(withoutalegalwill)ishissister’sson,hisnearestbloodrelative

in the next generation. Under patrilineal norms, his estate devolves to his children, who are

consideredhisbloodkin.Underbothtraditionalnorms, widowshavenoinheritancerights,and

Africanextendedfamiliesareattractingnewattentioninboththetheoreticalandempiricaleconomics

literatures.Bertrand,MullainathanandMiller(2003)findolderrelativesbecomingeligibleforpensionpayments

affectadultlaborsupplydecisionsinblackSouthAfricanhomes.ChitejiandHamilton(2002)findtransfersfrom

richertopoorermembersof

African‐Americanfamiliesdeterwealthaccumulationmorethaninwhitefamilies.

HoffandSen(2005)modelextendedfamiliesbecomingpovertytraps;andAlgerandWeibull(2008and2010)

shothatthetheexpectationoffinancialassistancefromfamilymemberscanpreventthedevelopmentof

insurancemarkets.

are oftenleftwithnoassetsbecause of a traditional presumption that assets acquired during

marriage belong to the husband.Rather than relying on their husbands’ estates, they must

dependontheirlineages’socialsafetynets.

The1985IntestateSuccession (PNDC)Law111 wasenactedto alterperceivedadverse

effects of these traditional norms,  especially on widows with husbands of matrilineal lineage.

Our surveys of widows living in villages selected for matrilineal, patrilineal or mixed lineage

norms,revealthataquarterofacenturylater,Law111islittleusedandtraditionalinheritance

normspersist.Thisconfirmspreviouswork(FIDA,2007;FenrichandHiggins,2005;andScholz

andGomez,2004).Welinkthistoadearthofinformationabouttheformallaw,lackofaccess

to the formal judicial system, and the continued social importance of overtly adhering to

traditionalnorms.Mostlow‐incomeGhanaians die intestate;andwhile someprofesstoknow

ofLaw111, remarkablyfewmakeuseofit.Thosewhoknowofthelawhave,however,built up

morefamilyassets jointly, evenaftercontrolling education level.Thiseffectis, however,least

evident for widows with husbands from matrilineal lineage traditions – the very people the

reformsfocusedonadvancing.

Wealsosurveywidowsabouttheextentofsupportfromtheirlineages,andtheiraccess

toeconomic(money,educationandhealthcare)andsocialsupport.Widowswhoacknowledge

closertieswithmembersofeithertheirownortheirspouse’slineagereportmoresupport,as

do those very few who made use of Law 111 or who inherited via a legal will.Intriguingly,

widows of matrilineal  lineage also report better economic support, consistent with maternal

lineagesocialsafetynetsbeingmoreeffective.

Oursurveytargetswidowslivinginvillages,andnotconnectedtotheformalsectorof

theeconomy.

Whilethesearemostrepresentative,theinheritancepracticesofpeopleinthe

formalsectorarealsoofinterest.Duetothelowsurveyresponseratesfrommiddleandhigh

incomefamilies,wetherefore complementoursurveyanalysiswithindividual‐leveldata from

theSocialSecurityandNationalInsurance Trust(SSNIT),

thesolepension annuityprogramfor

retirees.Thisprogramprovidesretireeswithfixedpensionannuitiesandshould theydiebefore

theannuityexpirationdates,survivorbenefitspayabletoselectedheir(s).Inasecondattempt

to improve the economicsituations of widows and their children, the 1998 Children Act 560

mandatesthat60%ofthissurvivorbenefitpasstothedecedent’schildrenagedunder18,with

thedecedent’schoiceofbeneficiariesgoverningonlythedispositionoftheremaining40%.

Due to confidentiality rules, the SSNIT allowed access only to older files.Despite this

working against finding significant effects, the data provide some evidence that Act 560

benefitsthenuclearfamiliesofdecedents,especiallythoseofmatrilineallineage.

Together these results indicate that these formal legal reforms have a very limited

impactonmost Ghanaians.Specifically,theyareefficaciousonlyforpeopleconnectedtothe

formal economy.For most Ghanaians, living in villages and dependent on the traditional

economy,thereformsareeitherirrelevantorofindirecthelponly.

Ourstudycomplementsa growing empiricalliteratureontheeconomicsofthefamily;

and on the importance of inheritance rights in developing countries. Quisumbing and Ostuka

(2001) link land inheritance rights to skills acquisition decisions in Sumatra; Quisumbing,

Panyongayong, Aidoo and Otsuka (2001) report that improved women’s land rights in Ghana 

incentivize the cultivation of tree crops, such as cocoa. Lastarria‐Cornhiel (1997) link

privatization to the land rights of marginalized Africans. Hacker (2010) provide a broad

literaturereview,anddiscussgender‐relatedinheritanceissuesindifferentpartsoftheworld.

Ellul,PaganoandPanuzi (2010), inasampleof10,004familyandnonfamilybusinesses across

38 countries, find that strict (traditional) inheritance laws interact with weak investor

protection laws to impede investment in family businesses,  but not in non‐family businesses.

Where inheritance norms allow (or require)  business owners to bequeath more substantial

proportions of their estates to non‐controlling heirs, investors are more reluctant to provide

externalcapital.

Thepaperisorganizedasfollows:Section2 provides a briefbackgroundontraditional

inheritancerulesinmatrilinealandpatrilineallineages.Section3

outlinestherelevantfeatures

of Law 111 and Act 560. Section 4 describes our data, and section 5 summarizes our

econometricmethodologyandempiricalresults.Section6concludes.



2. TraditionalInheritanceRules:ABackground

The inheritance  rights of spouses and children depend on the form of their marriage and on

their lineage traditions.Marriages in Ghana can be monogynous or polygynous, and can be

ordinance marriages (legally valid civil or Christian marriages) or customary marriages as

prescribedbycustomarytribaltraditions.

Thelastisbyfarthemostpopular,withuptoeighty

percent of marriages in contemporary Ghana entered solely under the customary system

(Awusabo‐Asare,1990).

Inpractice,almostallcouplesmarryinatraditionalceremonyrecognizingthenewbond

between the two families.Subsequently, some follow up with an ordinance marriage in a

church.Theseareusuallywealthiercouples.Notalltraditionalmarriagescanbe recapitulated

as ordinance marriages because traditional marriages can be polygynous, while ordinance

marriagescannot.While  theordinancemarriagemay specifyinheritancerules– for example,

thatequalthirdsmightgotothedecedent’sspouse,children,andextendedfamily–customary

rulestakeprecedenceinmarriagesthatarealsoenteredintraditionalceremonies.

Undercustomaryrules,thecorpseandallpropertyofapersonwhodieswithouthaving

writtenawill(anintestatedecedent)passestothefamily.One’sFamilyiscustomarilydefined

as one’s lineage: “the extended group of lineal descent of a common ancestor or ancestress”

(Kludze 1983, pp 60). The head of the lineage appoints a “successor” to assume the estate,

rights and obligations of the decedent on behalf of the lineage.Only a legal will overrides

customarylaw,andfewGhanaianshavelegalwills.

 Theapplicable customarylawvariesacrossethnicgroups,andeachtribaltraditionisan

intricatebodyofrules, obligations, andnorms.However,Ghana’scustomary legalregimesas

regards inheritances can be meaningfully divided into two broad categories:matrilineal and

patrilinealtraditions:matrilinealandpatrilineal.



 Islamicmarriagehasa'special'status,withtheQurandefiningmarriageandinheritancerules.Theseletaman

marryuptofourwomen,letonlymeninheritcertainassets,etc.Islamiclawshapesthecustomarytraditionsof

Muslimtribes,whichpredominateinthefarnorth.Consensualunions,withneitheran

ordinancemarriagenora

marriage under tribal custom, provide no inheritance rights whatsoever.A deceased common law spouses’

propertyrevertstotheirfamilies.Inter‐vivostransferstoacommonlawspousearesubjecttolegalchallenge.

Figure1:RegionalmapofGhanadepictingthe10regions.

Thematrilinealsocieties arefoundinthesouthwestregions(Ashanti,CentralandWestern),

and parts of Northern region. Patrilineal groups are in the southeastern (Greater Accra and

Volta)andtheUpperregions.



2.1 MatrilinealCustomaryInheritanceNorms

As figure 1 shows, the Akans (Ashanti, Central and Western regions; and the Lobi, Tampolese

and Baga (Northern Ghana)  all use variants of matrilineal customary law. The Akans,

constituting about 48% of Ghana’s population and the largest tribe

, are often considered an

archetypicalmatrilinealculture.Undermatrilinealtradition,afamily’scontrollingspiritpasses

fromgenerationto generationonlythroughfemalebloodlines,fromwhomAkanchildrenare

believedtoinherittheir“fleshandblood,”i.e.,theirsourceofexistence(Bleeker(1966).Family

ties,tracedonlythroughfemaleancestors,defineone’sextendedfamily,lineage,ormatriclan.



Inamatrilinealtribe,oneisthusrelatedbybloodtoone’smother,fullsiblings,andhalf‐

siblings by a common mother (uterine half siblings), but not to one’s father nor to any half‐

siblings by a common father.Thus, children belong to their mother’s lineage, but not the

father’s.AtraditionalAkanmalethusfeelsbloodkinshiptohis mother'sbrother(w

fa:pron.

wə‐fa),butatmostaweakconnectiontohisfather'sbrother.

An Akan male  does not consider his children to be his blood kin.His closest blood

relativeinthenextgenerationishissister’sson,andthismaternalnephew(w

fase:pron.wə‐

fa‐si)ishispresumedheirifhisbrotherspredeceasehimandhediesintestate.BecauseAkan

traditional rules revert a married couple’s acquired property to the decedent’s matrilineal

extended family (Awusabo, 1990), a widow  and her children can be left destitute by the

husband’sdeath.Shemustthuslooktoherbrothersforsupport;andherchildrenmustlookto

theirmaternalunclesforbequests.Theexpectationofinheritinga maternaluncle’swealthis

oftensaidtobluntanAkannephew’sincentivestoacquirehumancapitalorseekajob,andis

capturedneatlyinan

oldAkanadage“w

faw

hontimenyeegyuma”(Lit.“Ihavearichuncle;

Idon’tneedajob”).

Notethatamatrilinealdefinitionofwhois,andisnot,inone’sfamilydoesnotimplya

matriarchalpowerstructureoverthatfamily.Chiefsandtriballeadersinmatrilinealtribesare

almost

always male, and the leaders of matrilineally  defined extended families are almost

 TheAkantribecontainssub‐groups,definedbytheirmostlymutuallyintelligibledialects.Thelargestgroupsare

theAsante,AkuapemTwi,Akyem,Brong(intheBrong‐Ahaforegion),FanteandAgona.

One’smatriclanispreciselythosewithwhomonesharesidenticalmitochondrialDNA.

alwaystheirhigheststatusmales.



Figure 2 illustrates how a matrilineal controlling spirit flows from generation to

generation.Themembersofamatriclan all share acommonfemaleancestor,towhomtheir

mothersaretiedbyfemale‐to‐femalelinesofdescent–showninblack.



Figure2:MatrilinealDefinitionofBloodRelatives

A circle represents a female and square represents a male. One’s lineage consists of all descendants

(white)ofallcommonfemaleancestorsthroughfemalebloodlines.Childrenofbothgendersbelong

to their mother’s, but not their father’s, lineage. One is thus related to one’s mother, but not one’s

fathers,and

toallmembersofone’smother’slineagebutnottomembersofone’sfather’slineage.





2.2 PatrilinealCustomaryInheritanceNorms

Themainpatrilineal societiesinGhana aretheGatribe(intheGreaterAccraregion),theEwe

tribe(intheVoltaregion),andtheDagombaandNanumbatribesintheUpperEastregion.Ina

patrilinealtribe,afamily’scontrollingspiritpassesfromgenerationtogenerationonlythrough

malebloodlines,andtheseconnectionsdefineone’sextendedfamily,orpatriclan.



Underpatrilinealcustom, one’sextendedfamily thusincludesone’s childrenaswell as

one’s father, siblings, half siblings by a common father, aunts and uncles, and so on. One’s

MatrilinealdefinitionsofethnicityarenotunknownintheWest.Forexample,oneisJewishbybirthonlyifone’s

motherisJewish.A Jewishfatherdoes not count.As withthe Akan,amatrilineal definitionof family didnot

implymatriarchalcontrolofancientHebrewtribesor kingdoms.Many

Americanaboriginalculturesalsousea

matrilinealdefinitionofbloodkinship‐theCherokee,Gitksan,Haida,Hopi,Iroquois,Lenape,andNavajo,among

others.

 One’spatricianispreciselymalerelativeswithYchromosomesidenticaltothatofone’sfather,plustheir

immediatechildrenofbothgenders.

sistersandhalfsistersbyacommonfatheraremembersofone’slineage;buttheirchildrenare

not.Thisisbecause theybelongtothatsister’sorhalf‐sister’shusband’sfamily.Likewise,one’s

grandchildren through a son belong to one’s family, but grandchildren though a daughter

belongstotheirfather’sfamily,andarethusnotone’sbloodrelatives.



Figure3:PatrilinealDefinitionofBloodRelatives

A circlerepresents a femaleanda square represents a male.One’s lineageincludesall descendents

(white)ofcommonmaleancestorsthroughmalebloodlines.Childrenofboth gendersbelongtotheir

father’s,butnot their mother’s, lineage.Oneis thus related toallmembers of one’sfather’slineage

butnot

tomembersofone’smother’slineage.





In a patrilineal society, children inherit their father’s estate, and widows thus look to

theirchildrenforsupportseeOllennu(1966)fordetails.Figure3distinguishesmembersofa

commonpatriclan,inwhite,frompersonsnormallyconsideredrelativesinWesternsocieties,in

grey,whodonotcountasbloodrelativesin

apatrilinealculture.



2.3 CriticismsofTraditionalInheritanceNorms



AkeydifferencebetweenFigures2and3isthatmatrilinealculturesdonotnumberadeceased

man’s widow or children among his blood relatives.Western observers thus often see

patrilinealtraditionsasmoresupportiveofwidowsandchildren.

However, patrilineal norms also appears superficially more familiar to Western

observers, who may neither understand nor appreciate the support provided by brothers,

maternal half‐brothers, and maternal uncles in matrilineal societies. A widow with a wealthy

brotherinamatrilinealtribemaybemuchbetteroffthana widowinapatrilinealtribewhose

poorhusbandleftherchildrenameagerestate.Whichsystembetterprovidesforwidowsand

orphans on average is thus an empirical question, and may not even be subject to broad

generalization.Some communities might apply a given set of traditional norms with more

generositytowidowsandchildrenthanothers.

Publicized cases of impoverished widows and children in matrilineal tribes, buttressed

by survey evidence assembled by women’s advocacy groups and Christian organizations,

repeatedlymadethepovertyofwidows’andtheirchildrenapublicpolicyissueinthedecades

subsequenttoGhana’s1957independence.Widow‐headedhouseholdsthroughoutGhana,but

mostevidentlyinruralmatrilinealhomes,werehighlightedashavingextremelevelsofpoverty

– due, in part at least, to traditional inheritance norms. Intestacy law reform attracted

increasingdebate,butactualreformwasslowtocome.Onereasonforthisdeadlockwasthe

absence of a viable  reform proposal; another was doubtless the legislators’ fear of providing

fodderfortribalchauvinists.

The case  for reform grew to encompass several arguments.The most direct was the

caseforconjugal(nuclear)familiesretainingall ormostofadeceasedspouse’sassetstoshield

widowsandtheirchildrenfrompoverty.

But the case for reform went beyond such welfare considerations. Of at least equal

importance were the incentives inheritance customs

created for wealth accumulation by

individualsandconjugalfamilies.Especiallyinmatrilinealtribes,aplausiblecasewasmadethat

thetransferofaconjugalfamily’sassetstothedeceasedman’smaternalnephewundermines

the incentives of the husband and wife to acquire skills, exert effort, and accumulate assets;

andtoblunt

thesameincentivesinmaternalnephews.

Anotherproblemconcernsthealienabilityofassetspassed to a lineage.No individual

personownstheseassets,andtheconditionsunderwhichtheycanbeboughtandsoldarestill

murky.A lineage is a corporate entity, but often lacks necessary legal titles because

of the 

difficulties of deeding an asset to multiple owners. For example, throughout Ghana, lineages

ownlandandotherassetsthat havenovaluebeyondtheirprimaryuse.Theseassetscannot

serve as collateral for a loan; and improvementstothemare the property of the lineage, not

theindividualwhopaysfortheimprovements.Individualsthushavescantincentivetoaddto

thevalueofsuchassets.Traditionalinheritancesystemsmightthusexplain,inpartatleast,the

failure of many sub‐Saharan countries to formalize titles to land and capital assets (De Soto

2000).Reforms that keep assets from reverting to lineages might ultimately spread clearer

propertyrightsandthusimproveallocativeefficiency.

It is possible to contract around these problems.But lineages must solve a collective

action problem  to act in concert. Individuals can nullify traditional inheritance norms with a

legalwill.However,mostGhanaiansdieintestate.

7

Highilliteracyrates,alackofaccesstothe

formallegalsystem,andthe fearofretaliationbytheextendedfamilydoubtlessallplayarole.

Malesinmatrilinealhouseholdscanattempttoprotecttheirwivesandchildrenwithinter‐vivos

transfers;butthesecanbeundone–eitherlegallyorbysocialpressure.

In fact, actual monetary transfers may also go in the opposite direction: e.g., the child

pays money to the father. La Ferrara (2006) finds Akan (matrilineal) sons transferring more

money to their fathers than do otherwise similar sons in patrilineal cultures, especially if a

paternalaunt’ssonresideswith,orlivesinthesamevillageasthe,father.LaFerraraconcludes

thattheincreasedtransfersfrom Akansonsarepartiallyattemptstoinfluencetheirfathersto

directlandgiftstothem,ratherthantothefather’snephew.



3. LegalReforms

Bythemid‐1970s,acaseforcomprehensivereformwaswidelyacknowledged.Forexample,in

1979,theConstitutionoftheThirdRepublicofGhanaproclaimedinitsAarticle32(Woodman, 

1985):



§2 No spouse may be deprived of a reasonable provision out of the estate of a spouse,

whetherthe

estatebetestateorintestate.

§3 Parliamentshallenactsuchlawsasarenecessarytoensurethateverychild, whetheror

not born in wedlock, shall be entitled to reasonable provision out of the estate of its

Inoursurveyofwidows,only8%reportedthattheirspousehadalegalwill.

parents.



Parliament,ofcourse,didnosuchthing,andtheConstitutionwasabolishedinamilitarycoup

laterthatyear.Themilitaryjuntareiteratedthetwopledges;buttooknoimmediatejudicialor

legislativeaction.

On June 14 1985, the Provisional National Defence Council (PNDC), the ruling military

junta, proclaimed four interrelated reforms that, in theory at least, radically reformed the

ground rules for intestate inheritances.These were: the Intestate Succession Law (PNDC Law

111,1985),theCustomaryMarriageandDivorce(Registration)Law(PNDCLaw112,1985),the

Administration of Estates Law (PNDC Law 113, 1985), and the Head of Family Accountability

Law(PNDCLaw114,1985).AllfourinitiativeswerejustifiedinanaccompanyingMemorandum

asreflecting“thegrowingimportanceofthenuclearfamily”relativetotheextendedfamily.



3.1TheIntestateSuccessionLaw(PNDCLaw111,1985)

The most important of these for our purposes, and for rebalancing customary inheritance

norms against the needs of surviving members of the conjugal family, is the Intestate

SuccessionLaw(PNDC)Law111,1985–hereinafterLaw111.Indeed,ithasbeencharacterized

asthemostradicallegislativereformevermadeintheprivatelawofGhana(Woodman,1985).

Wethereforepausetoelaborate.

AlthoughLaw111isphrasedtobegender‐neutral,itwasseenasavictoryforwomen,

andsohailedbywomenadvocacygroups.Thelawallowsawidowandherchildren–hitherto

completely denied rights to

 the nuclear family’s assets under matrilineal norms – to be the

primarybeneficiariesofthedeceasedhusband.

ThewritofLaw111isrestrictedintwoways.First,thelawappliesonlytopropertynot

disposedofina legalwill.BecausemostGhanaiansdieintestate,thisrestrictionisnot

thought

ofparamountimportance.Moreimportantly,Law111 doesnotapply  tolineageproperty –a

conceptunfamiliartomostWesternobservers.Muchland,andofothersortsofpropertytoo,

belongstoalineage,nottoanyindividual.Suchpropertyassignedtothedeceasedhusbandfor

use during his life automatically reverts to the lineage upon his death; and, in a matrilineal

tribe,mostlikelypassestooneormoreofhismaternalnephews.

Law111doesapplyonlyto

self‐acquiredproperty–assetsthedeceased,orhisnuclearfamily,purchasedorcreatedduring

his life.Because the husband is typically considered its sole owner, a conjugal family’s self‐

acquiredpropertyvirtuallyalwaysrevertedtoadeceasedhusband’slineage.Awoman’srole,in

whatever form, was rarely recognized.



 The lawmakers explicitly referred to this issue in an

accompanying Memorandum, which explained the reforms thus: “It is the right that the

husbandwithwhomthewomanhaslivedandwhomshehasprobablyserved,isthepersonon

whosepropertyshemustdependafterhisdeath.”

Law 111 partitions a decedent’s assets into two categories:household chattels and

residue assets.Household chattels include all household belongings in regular use: clothes,

furniture, appliances, a family non‐commercial vehicle, farm equipment, and household

livestock. All household chattels automatically devolve to the conjugal family. Residue assets

includebusiness‐relatedandinvestmentassets:businessproperties,commercialvehicles,non‐

primary residential properties, bank accounts, savings, and investments.Residue assets are

distributed to members of the decedent’s conjugal family and extended family according to a

set of formulae set forth in sections 5 – 8 and 11 (Articles 1 and 2) of Law 111.Table 1

summarizesthese.

The first row of the table sets out a baseline case, where the decedent has surviving

relatives in all relevant classes.In this example, section 5 stipulates that 3/16 of the residual

goestothespouse,9/16tothesurvivingchildren,and1/4 tothelineage;thatlast to

besplit

equally between the parents and those entitled to inherit under the decedent’s traditional

norms.

 Lineage property, which encompasses the extended family‐owned assets, is distinct from tribal property. A

typical example of tribal property is the communal land “owned,” in principle, by the paramount chieftaincy

(called the stool) in trust. Individuals have rights to use the land for farming, or for some other commercial



activity,byvirtueofmembershipofthetribebutonlywiththeconsentfromthestoo1(i.e.,thechief).Tribaland

lineagelandareessentiallyinalienablebecauseofa“tragedyoftheanti‐commons”problem.Allmembersofthe

currentlylivinggenerationofthelineageortribeareconsideredcustodians

forpropertythatalsobelongstoall

pastandfuturegenerations,andthuscannotbesoldwithouttheexplicitconsentofalllivinglineagemembers

pluscountlessdeceasedandunborngenerationsofthelineage–aconditionprospectivebuyerscanbecertainis

neversatisfied.Sagasaboundofforeignersthinkingthey

havepurchasedsuchpropertywhentheyhavenot.

Table1.ResiduePropertyDistributionunderIntestateSuccessionLaw(PNDCLaw111,1985)

The decedent’s residue property (property not classified as household chattels or lineage property) is

apportionedtorelativesbyoneofthefollowingformulas.Residueassetsincludebusiness‐relatedand

investment assets: business properties, commercial vehicles, non‐primary residential properties, bank

accounts, savings, and investments.The applicable formula  depends on which of

 the decedent’s

relativessurvive.



Livingconjugal&

extendedfamilymembers

ShareofresidueLaw111assignsto:

Spouse Children Parents Lineage

a

State

b

Ifallsurvive 3/16 9/16 1/8 1/8 0

Nolivingspouse‐  3/4 1/8 1/8 0

Nolivingchildren 1/2‐1/4 1/4 0

Nolivingspouseorchildren‐‐3/4 1/4 0

Nolivingspouse,children,orparents‐‐‐1 0

Nosurvivingknownrelatives‐‐‐‐1

a. Tobedistributedinaccordancewiththetraditionsofthelineage.

b.Intrustforanypersonsubsequentlyidentifiedassufficientlyclosetodeceasedtobealegitimateheir. 



Theotherrowsmodifythebaselineformulaintheabsenceofsurvivingheirsofoneor

more sorts.For example, the second line shows that if the decedent has no living spouse –

becauseofeitheradivorce or thespouse’spriordeath–whatwouldhavebeenthe  spouse’s

share passes instead to the children.If the decedent has neither a living spouse nor living

children, what would have been their shares passes to the decedent’s lineage – with the

parentsreceiving3/4oftheresiduepropertyandtheremaining1/4distributedbythelineage

in accordance with its traditional norms.In the rare event of the decedent having no known

relativesofanykind,the  residue property goestothestateintrust,andcansubsequently be

disbursedtoone“whowasmaintainedbythe intestateorwithwhomtheintestatewasclosely

identified”, should such a person be found to exist.Thus, someone who lived with, or was

related in some sufficiently close way to, the decedent  can seek a court order to inherit a

portionoralloftheestate(seeWoodman,1985).

Over the more than a quarter of a century since Law 111 took effect, anecdotal

evidence and reports by women’s advocacy and religious groups concur that the law is not

widelyfollowed.Themost likelyreasons for this are a lack of information about the  law,the

inaccessibilityoftheformallegalsystemtomanypeople,andaveryrealfearofreprisalsfrom

thelineageforviolatingcustomarylaws.Manyfamilies, especiallyinruralareas,knowonlythe

customary laws of their tribes. Moreover, government officials in these  areas are often

reluctanttoenforcetheformallawandapplysanctionswhenitisviolatedbecausethesesame

officialsareoftenalsochargedbytheirtraditionalcommunitieswithupholdingcustomarylaws.

The formal law is a written body of knowledge, while  customary law is passed along

orally, and thus more accessible to illiterate people.If legislation from Accra conflicts with

tribalcustom,thelatterusuallywinsout.A2007studybytheGhanaofficeoftheInternational

Federation of Women Lawyers (FIDA, 2007) shows about 40% of survey respondents –

interviewed in Accra (the capital and mainly patrilineal), Kumasi (the second largest city and

predominantlymatrilineal),andKoforidua(amixtureofinheritancesystems)–hadeithernoor

an erroneous knowledge of Law 111 – with these responses much more frequent among

peoplewithlittleornoeducation.Amere3%hadacompleteknowledgeofthelaw.

TheFIDA

studypursuestheissuewiththefewrespondentswhoknewofthelaw.Eventhesefindactual

useofLaw111toberestrictedbymultiplebarriers.

Widows often lack the financial resources to mount a legal challenge, and are often

overwhelmedandfrustratedbyacumbersomelegalprocedure.Thewidowmustpetitionfora

LetterofAdministrationfromthecourtstogainstanding–andthisrequirestheapprovalofthe

head of the decedent’s lineage, typically a contending party. In addition, she must obtain 

competentlegal guidance to executethisdocumentpreciselyinaccordance with the letter

 of

Law111,foranyproceduralerrornullifieshercase.Addedtotheexpenseoflegaladviceisthe

cost of the decedents funeral and burial rituals, which the widow must pay in their entirety

shouldshecontestthecustomarylaw.Thesecostsareeasilyprohibitivegiventheimportance

of

elaborate funerals in Ghanaian cultures. The community typically expects a grand funeral,

andthisisonlyfinancialpossiblewiththesupportofthedecedent’slineage.

Perhaps even more daunting than all of these financial costs are the social costs a

 Recenteducationdrivesandsocialawarenessprogramsareactivelyworkingtoinformpeopleoftheirrights

undertheformallegalsystem.Prominentamongthemare:theMinistryofWomenandChildren,theFederacion

InternaciondeAbogadas(FIDA,knowninGhanaastheInternationalFederationofWomenLawyers),the

Women’sInitiative

forself‐Empowerment(WISE),andWomeninLawandDevelopmentinAfrica(WiLDAF).

widow risks by challenging traditional norms. The repercussions from overtly  disregarding

deeply rooted tribal custom can be devastating.This messy and divisive process, with all its

attendant costs, conflicts, and adverse consequences, can readily be avoided if the decedent

leftawill.ButtheFIDA(2007)studyreportsthatwomenareunlikely topresstheirhusbandsto

writeawill.Indeed,amajorityofinterviewedwidowsdidnotknowiftheirspouseshadawill,

and never discussed writing a will with him. This response from one respondent, when asked

why,capturesthegeneralsentiment:



“Icouldneverask[myhusband]ifhehadawillornot…..IfIasked,hemayeventhinkI

amplanningtokillhimsoIcantakehisassets;oraccusedmeofbeingawitchor

somethingbad.Hemayevenaskforadivorce.”



3.2 Survivors’(Pension)BenefitsundertheChildren’sAct560,1998

A second comprehensive reform to Ghanaian inheritance laws developed in stages, and

provides a wealth of government data pertaining to middle  and high income Ghanaian

households, whose  survey participation rates tend to be very low in any event. The Social

Security and National Insurance Trust (SSNIT) runs the sole government‐sponsored pension

annuity program for retirees. Should the contributor die before his accumulated benefits are

fullydisbursed,60%oftheremainingbenefitspasstothedecedent’schildrenunder18.Each

contributor apportioned the remaining 40% to one or more chosen heirs. This reflects a

sequence of reforms, but primarily the Social Security Act (PNDC Law 247, 1991) and the

Children’sAct(Act560,1998),hereinafterLaw247andAct560,respectively.

Law 247 designates the  permissible choices open to members of patrilineal versus

matrilineal cultures, and is summarized in Table 2.Thus, Law 247 prohibits a member of a

patrilineal culture from listing a sister’s son as an heir, and forbids a member of a matrilineal

culture from listing a father’s father or father’s brother as an heir. But in both cases, one has

theoptionofeitheradheringtothetraditionalnormsofone’slineageorbequeathingbenefits

toone’sconjugalfamily.

A pension contributor’s choice of heirs is confidential, buried in government files, and

available to interested parties only after the contributor’s death.This theoretically lets one

defy customary inheritance norms by bequeathing one’s accrued pension wealth to one’s

conjugalfamily,notone’slineage;withno‐onetoknowuntilwellafteroneissafelydead.

Thesocialsecuritysystemdatesback(atleast)to1946,whenChapter30ofthePension

Ordinance of 1946 provided government pensions for certain public sector employees, a

scheme that became known as CAP30. A mor e general social security system began with the

militarygovernmentofthetime,theNationalRedemptionCouncil,decreeing(NRCD127,1972)

the expansion of a previous Parliamentary Act 279 to establish a Provident Fund topayevery

formalsectorworkeralump‐sumupon retirement.In1991,another militarygovernment,the

ProvisionalNationalDefenceCouncilproclaimed theSocialSecurityAct(PNDCLaw247,1991),

hereinafter Law 247, under which the Social Security and National Insurance  Trust (SSNIT) is

made the sole government‐sponsored pension system.The system resembles standard

Westernpensionsystemsinsomeways,butdeviatesmarkedlyfromtheminothers.Ourdata

covermuchoftheperiodwhenLaw247wasineffect.



Table2.Schedule45oftheSocialSecurityAct(PNDCLaw247,1991)

UnderLaw247,onlycertainpersonsareeligibletobelistedasheirstoadeceasedbeneficiary’sSSNIT

pensionaccruals.Differentchoicesetsareoffered tomembersofmatrilinealversuspatrilinealtribes.



Decedent’sTraditionalNorm Patrilineal Matrilineal

Mother,father allowed Allowed

Husband allowed Allowed

Wife,son,daughter allowed Allowed

Brother,sister allowed Allowed

Father’sfather allowed Prohibited

Mother’smother prohibited Allowed

Father’sbrother allowed Prohibited

Mother’sbrother prohibited Allowed

Mother’ssister prohibited Allowed

Sister’sson prohibited Allowed

Sister’sdaughter prohibited Allowed



AnewNationalPensionsAct(Act766)wentintoeffectonJanuary2010andexpandstheschemeunderLaw247

toincludevoluntarycontributionsfromself‐employedpersonsandindividualsintheinformalsector.Social

SecurityfortheinformalsectorwillbeadministeredbytheSSNITInformalSectorPensionFund.

Likemanygovernmentpensionschemes,theSSNITis apay‐as‐you‐gosystem:formal‐

sector workers’ current contributions fund benefits paid to pensioners. Law 247 requires that

all employers contribute 17.5% of their base salaries to the fund.This appears as a 5%

deduction from the employee’s monthly pay check, and is matched by a 12.5% employer

contribution invisible to the employee.Act 766 now allows the self‐employed to join the

systembymakingthefull17.5%contribution.

Also like many other systems, the SSNIT is a  defined benefit system. The minimum

annualpensionbenefitis50%oftheaverageofthecontributor’shighestthreeannualsalaries

earned in the twenty years prior to retirement. The benefit rises by 1.5% of average of that

average for each additional year of work and contributions to the  fund. So, theoretically,

anyone who retires at age 60, having contributed for 40 years to the  fund, merits a pension 

equalto80%ofaverageearningsintheretiree’sthreemostprosperousyears.



The major difference between most other pension systems and the SSNIT it its fixed

term:paidupcontributorsareentitledtoexactly144consecutivemonthlybenefitpaymentsin

retirement, no more and no fewer.Under Law 247, an individual becomes entitled to full

pensionbenefitsaftercontributingtotheschemefor240monthsandattainingtheageof60, 

the mandatory retirement age.

 The retiree then receives a monthly pension for the next 12

years.Whentheretireeturns72,thebenefitterminatesandtheretireemustrelyonrelatives

orsavings.Individualsmayoptoutof thisdefaultscenarioand receive25%ofthepayments’

present value as a lump‐sum upon

retirement, and the other 75% as monthly payments over

thenext12years.

Ifthecontributordiesbeforeage72,thepresentvalueoftheremainingpaymentsthe

contributor would have received, computed at the treasury rate over the same period, was

thenpaidtotheheirsthecontributorchosefrom

theoptionsmadeavailableinTable2.

The

largestsubsequentchangetothissystem,theChildren’sAct(Act560,1998)of1998,mandates

Earlyretirement,startingfromage55,withreducedpensionpaymentsispermittedunderLaw127.Individual

whoretirebeforeage60lose7.5%oftheirage60retirementbenefitforeachyearsuntiltheir60

birthday.

Peopleinhigh‐riskoccupations,suchasmining,areexemptedandcanretireat55yearswithfullpension.

Ifacontributorfallsshortoftherequired240months,thetotalcontributionsplusinterestathalftheT‐Billrate

isrefundedasalumpsumatretirement.

BenefitsarediscountedattheloweroftheprevailingTreasurybillrateand10.

thattheSSNITpay60%ofsuchsurvivorbenefitstothecontributor’sminorchildren(age18or

younger).In theory, whenever the SSNIT receives a procedurally complete claim, with all

necessary supporting documents (e.g., death and burial certificates), it should investigate the

familytoascertainwhetherornotthedecedenthasminorchildrennotlistedasbeneficiaries.

In practice, the SSNIT lacks the resources to do this, and merely ascertains the validity of the

submitted claim. The remaining 40% of the survivor benefit is then disbursed  to the

beneficiaries the decedent selected from the appropriate column in Table 2, in the selected

proportions.

AnothermajordifferenceisthatGhana’sgovernmentsponsoredpensionprogramwas,

until very recently, restricted to formal‐sector employees.Workers in the informal sector –

subsistence agriculture, fishing, roadside stands, etc. – are not covered.

Some 80% of

Ghanaiansworkintheinformalsector‐subsistenceagriculture,fishing,roadsidestands,petty

trading,andthelike(Heintz,2005).SSNITdatathuspertainonlytomiddleandupperincome

Ghanaians.WethereforeusebothSSNITdataandsurveysofinheritancepatternsamonglow‐

incomeGhanaianstogainabroaderpictureofthecurrentsituation.



4. DataandEmpiricalStrategy

4.1 SurveyDataDescriptiveAnalysis

Wesurvey322widowslivinginfourvillagesinSouthernGhana:BortianorandIngleshiAmanfro

are both predominantly patrilineal, and in the Greater Accra Region; Abura Dunkwa and

NyankomaseAhenkroareintheCentralRegionandmostlineagestherearematrilineal.

Potential survey participants were identified with the help of a town or village council

member, a town leader or the traditional chief, whose approval were sought for our team of

researcherstoconductthesurvey.Householdswererandomlychosen,andthequestionnaires

were administered to an adult person in the house, in private. Because visits to randomly

selected seemingly more

affluent households in the urban areas generally yielded no

responses, our final data consist mostly of very low‐income respondents, though not the

Recentreforms(Act766,2008)mandatethattheSSNITorganizetheInformalSectorSocialSecurityFund,which

activelyencouragesinformalsectorworkerstosignupandsavefortheirretirement.

poorest of the poor living rough. About 50% of our respondents report no formal education,

and25%reportedthesamefortheirspouses.Theaverageis5yearsofformaleducation.Only

5.2%reportabankaccount(singleorjoint),though26%haveotherpersonalassets.Overhalf

nonethelessreportcontributingdirectlytotheconjugalfamily’swealth.

The survey data contain information to assemble a profile of each respondent’s age,

education level, inheritance system, years married, children (by spouse), minor children; and

marriage type (polygynous or monogynous). Our youngest respondent was 20 and our oldest

was93.ConsistentlywiththeethnicdistributioninGhana,about48%reporttheirtraditional

custom as matrilineal. Respondents also provided information to let us assemble  similar

profiles of their spouses.Spouses’ profiles appear similar, though they are slightly more

educated–about8yearsofformaleducation.

A key part of our analysis is to gauge how respondents’ information about the  law

shapedtheownershipstructure of thefamilyassets(i.e., individually ownedorjointly),which

law was applied in the distribution of the assets when a spouse died, and the welfare of the

family after the death. We therefore ask widows an additional set of questions.Nearly 47%

reportpriorknowledgeofthelaw,attheir marriage orbeforethedeathoftheirspouse.But

only3.2%ofwidowsreportitsuseintheircases.Rather,justover7.4%reporttheirdeceased

spousehavingawillandover75%reporttheestatehavingbeendistributedinaccordancewith

traditional norms. The remaining 14% did not know which rule was employed, or reported

usingareligious‐basedprincipletodivideuptheestate(e.g.,Muslim).Almost46%reportbeing

dissatisfiedwiththedistribution.

To assess widows’ economic and social status, we ask respondents to compare their



situations in the years immediately before versus after their spouse’s death. Because they

virtually all dwell entirely in the informal economy, questions about monetary income do not

capturetheir economicsituations. Wethereforeaskthemtoratetheireconomicsituationsor

opportunities, defined as access to financial services (formal or

informal), health care for

themselves and their children, and educational opportunities for their children. Access to

financial services in this context means an expectation of being able to borrow money in a

pinch from a  financial institution, or from the head of the lineage or its more prosperous

members. Health care and education typically entail small informal monetary outlays. Thus,

obtaining needed medical care and clothing and provisioning children for school require cash

outlaysthat,inasubsistenceeconomy,typicallyrequireeconomicsupportfromone’slineage.

We are also interested in whether or not widows, if they attain better financial

situations than traditional law prescribes, encounter tension with their lineages, or those of

theirhusbands.Wethereforealsoaskeachwidowtocompare thequalityofherrelationships

withherspouse’sextendedfamily,beforeversusafterhisdeath.

Finally, we are interested in how divergent formal and customary laws affect welfare.

Wethereforeaskedwidowsquestionsabouttheireconomicpositionstoassesshowwelltheir

traditional support networks performed.Here, quantitative assessment is tricky.Previous

work(e.g.,Awusabo‐Asare,1999)findsthatpoorvillagers,suchasthoseweinterview,tendto

report their economic situation as very bad, leaving little variation to study.It is therefore

necessary tointroduceanorm,towhichtheycan compare themselves.However,acommon

norm is inappropriate because different lineages have markedly different capacities to help

theirlessfortunatemembers.Somelineagesincludegovernmentofficials,police,formalsector

workers, émigrés, or others, whose formal or informal income can be redistributed to needy

relations.Somelineagesbelongtotribeswithmineral rights,whoserevenue streamscanalso

be redistributed from chiefs to lineage heads and then to needy widows. However, most

lineages have few or no such resources, and genuinely cannot provide more than minimal

subsistencesupport.

Toextricateinformationaboutwelltraditionallineagesupportsystemsfunction, weask

eachwidowtoconsiderthefinancialcapacityofthelineageinquestiontosupport peoplesuch

as herself. With this benchmark in mind, she was then asked if she received any lineage

support; and if so, whether this was less than, about in line with, or more than that capacity

allowed. A study of “mixed marriages,” in which one spouse has a patrilineal tradition and

anotherhasamatrilinealtradition,wouldhavebeenespeciallyinstructivehere.Unfortunately,

interlineage marriages within our sample are extremely rare,

as Table 4.Roughly nine in ten

Ghanaians marry within their lineage tradition.Studying interlineage marriages would be

interesting,butTable4revealsoursampleofthesetobeverysmall.

Table4:MarriagePatterns:Withinversuscross‐lineagemarriages

Numbers ineachcellarenumbersofrespondentsofhorizontal lineage classificationwhose spouses

are in vertical lineage classification. Percentages to the right of these numbers are percentages of

respondents,percentagesbeloweachnumberarepercentagesofspouses.

Spouses’traditionallineage

Matrilineal Patrilineal Total

Respondents’

traditionallineage

Matrilineal

145 88.4% 19 11.6% 164

93.5% 11.4 50.9%

Patrilineal

10 6.3% 148 93.7% 158

6.5% 88.6% 49.1%

Total 155 48.1% 167 51.9% 322



Table 5 highlights differences in means by respondents’ inheritance options. Panel A

compareswidowswhoknowofLaw111torespondentswhodonot;whilePanelcomparesby

thedeceasedspouse’slineagetype.

Unsurprisingly, Panel A shows that knowledge of Law 111 correlates positively with

education, for both the respondents and spouses; and the educated are more likely to be in

monogamous marriages.Confirmingtheimportanceof financial incentives,knowledgeofLaw

111isalsogreateramongrespondentswhoreport contributingmorewealth totheirconjugal

families.

Widows who had knowledge of the law were significantly more likely to have settled their

deceasedspouse’sestateunderit.In fact,no‐oneindicatednoknowledgeofthelawpriorto

their spouse’s death made use of the law  in dealing with the estate – suggesting that

proponents of the Law’s greater usage might consider more energetically distributing

information at the time

of a spouse’s death or serious illness.More surprisingly, though

consistently with previous studies such as FIDA (2007), Panel A shows that about 70% of

respondents with a prior knowledge of Law 111 nonetheless settled their deceased spouses’

estatesinaccordancewithtraditionallineageorreligiouscustoms,afigurethatis

onlyslightly

lessthanthatforrespondentswithnopriorknowledgeofLaw111.Presumably,muchorallof

theestatespassedtothelineageinbothcases.

Table5.MeansurveyresponsesbyknowledgeofLaw111

PanelA.Meansurveyresponses,bywidow’sknowledgeofLaw111,whichextendslimitedformal

legalinheritancerightstoconjugalfamiliesregardlessofcustomarylaw

 KnowledgeofLaw111

 Yes No Difference

RespondentProfile 

Age(years) 52.9 56.6‐3.7 **

Education(yrsofformaleduc) 5.1 2.8 2.3 ***

Monogamyatmarriage(%ofwidows) 82.4 65.5 16.9 ***

Monogamyatspouse’sdeath(%ofwidows) 79.1 66.8 12.2 **



InformationonSpouse 

Yearsmarried 23.8 23.9 0.1

Spouseeducation(yrsofformaleduc)

 8.0 5.5 2.5 ***

Customary/Islamicmarriage(%) 82.6 85.0‐2.4

Noofchildren(withspouse)  4.6 4.8‐0.2

‐aged<18(atspousedeath) 1.8 1.9‐0.1



OwnershipofAssets 

Jointaccountw/spouse(%ofwidows) 4.7 5.8‐1.1

Percentcontributedtofamilyassets(%) 27.6 18.4

9.2 ***

PersonalAssets(%ofwidows) 29.7 23.8 5.9

Inter‐vivosfromspouse:Asperabilityorbetter(%) 34.2 27.1 7.1 *



DistributionofAssets 

Will(%) 12.8 2.9 9.8 ***

PNDCLaw 6.1 0.6 5.5 ***

Customary/Islamic 67.1 87.3‐20.2 ***

%dissatisfiedwith

distribution 36.2 47.9‐11.7 **



WelfareafterSpouseDeath 



Economicsituation(%worseoff) 55.7 69.9

‐14.2 ***

Relationshipwithin‐law(%worse) 22.8 16.7

6.1 *

LineageFinancialSupport 

Own:Asperabilityorbetter 31.5 31.2 0.3

Spouse’s:Asperabilityorbetter

14.1 12.1 2.0

Difference(ownminusspouse’s) 17.4 ***19.1 ***

LineageEmotionalSupport 

Own:Asperabilityorbetter 49.7 52.6‐3.9

Spouse’s:Asperabilityorbetter 25.5 27.2‐1.7

Difference(ownminusspouse’s) 24.2 ***25.4 ***

PanelB.Meansurveyresponses,bycustomarylawapplicabletodeceasedspouse

Thecustomarylawofthedeceasedhusband’slineagedeterminesherinheritancerights.

Spouseinheritancecustom

 Matrilineal Patrilineal Difference

RespondentProfile 

Age(years) 54.6 55.1 0.5

Education(yrsofformaleduc) 5.1 2.8 2.3 ***

Monogamyatmarriage(%ofwidows) 79.1 68.1 11.0 **

Monogamyatspouse’sdeath(%ofwidows) 80.4 65.3 15.1 ***



InformationonSpouse  

Yearsmarried 24.5 23.2 1.3

Spouseeducation(yrsofformaleduc) 7.5

 5.8 1.7 ***

Customary/Islamicmarriage(%) 85.1 82.6 2.5

No.ofchildren(withspouse) 4.2 5.3‐1.1

‐aged<18(atspousedeath) 2.1 1.6 0.5 ***



OwnershipofAssets  

Jointaccountw/spouse(%ofwidows) 3.9 6.6‐2.7

Percentcontributedtofamilyassets(%) 26.1 19.4 6.6

***

PersonalAssets(%ofwidows) 34.4 19.3 15.1 **

Inter‐vivosfromspouse:Asperabilityorbetter(%) 28.4 32.3‐3.9



DistributionofAssets  

Will(%) 9.7 5.4 4.3 *

PNDCLaw 5.2 1.2 4.0 **

Customary/Islamic 70.7 85.4‐14.7 ***

%dissatisfiedwithdistribution 45.1 40.1‐5.0



WelfareafterSpouseDeath  



Economicsituation(%worseoff) 50.3 75.4

‐25.1 ***

Relationshipwithin‐laws(%worse) 27.7 12.0

15.7 ***



LineageFinancialSupport  

Own:Asperabilityorbetter 31.6 31.1 0.5

Spouse’s:Asperabilityorbetter 13.5 12.5 1.0

Difference(ownminus

spouse’s) 18.1 ***18.6 ***



LineageEmotionalSupport

Own:Asperabilityorbetter 58.1 44.9 13.2 **

Spouse’s:Asperabilityorbetter 31.0 22.2 8.8 *

Difference(ownminusspouse’s) 27.1 ***22.7 ***



Knowledgeofthe formallawcorrelateswithabettereconomicsituationinwidowhood

overall. However, it has little traction in explaining the relative economic and emotional

support widows receive from their lineages versus those of their husbands.Widows

knowledgeable of the law obtain insignificantly better financial support both from their own

and spouses’ lineages than do widows unfamiliar with the formal law. Indeed, widows

unfamiliar with the law actually  report insignificantly better financial and emotional support

from their spouse’s lineage, relative to their own. Across the board, widows’ own lineages

providemoresupport.

Panel B repeats the exercise, but partitions the data by respondents’ spouses’

inheritance tradition. Knowledge of the law is substantially greater among widows whose

husbands were from a matrilineal tradition – suggesting women’s and various organizations

have successfully reached more of those the law was specifically intended to empower.

However,PanelBalsorevealsbothmatrilinealdeceasedspousesandtheirwidowstobebetter

educated, perhaps also partially explaining their better familiarity with the formal law.

Matrilineal spouses’ widows are also more likely to have been in monogynous marriages; a

situationthatpresumablyimprovedtheirimplicitbargainingpowerwiththeirhusbandandhis

family. Matrilineal widows reported having fewer children with their deceased husband,

althoughaslightly largernumberoftheirchildrenwerelessthan18 yearsoldatthetimeshe

becameawidow.

Widows of matrilineal spouses, those least likely to inherit a deceased spouse’s assets

under customary law, nonetheless report having contributed more to

the conjugal family’s

wealth, and also having accumulated more personal assets. This might explain their small,

though insignificantly, greater dissatisfaction with the distribution of those assets after their

husbands’deaths.

It seems knowledge of Law 111 mitigates any adverse effects husbands’ matrilliny on

theirwidow.Matrilinealmen’swidowsknowledgeableaboutLaw

111reportlessdissatisfaction

with the distribution of the conjugal family’s assets (36% vs. 48% were dissatisfied). Perhaps

knowledge of the law strengthens their bargaining power within the traditional inheritance

process.

Finally,widowsofmatrilineal lineagemenreportinsignificantlybetterfinancialsupport

bothfromboththeirownandtheirdeceasedhusbands’lineagesthandowidowsofpatrilineal

men. Widows of matrilineal men also report better emotional support from both lineages.

Recall from Table 4 that 93% of matrilineal men’s widows are themselves from matrilineal 

cultures,sostrongsupportfromthewidow’slineageisunsurprising.Butthefindingthattheir

deceased husbands’ matrilineal lineages also support them, even though their children from

suchamarriagearenotconsideredtheirs,suggeststhatmatrilineallineagesingeneralprovide

unexpectedlystrongtraditionalsafetynetsforwidows.



4.2 PensionBequestsDataDescriptiveAnalysis



To complement our survey data, which cover very low‐income Ghanaians, we utilize official

dataonindividuals’bequestinstructionsregardingtheirSocialSecurityandNationalInsurance

Trust (SSNIT) benefits. These data pertain only to Ghanaians with employment in the formal 

sector–personsconsideredtobeofmiddletohighsocioeconomicstatus.

Hardcopy records of each beneficiary’s instructions are retained by the SSNIT, and are

consideredconfidentialuntilthecontributor’s death. Thereafter,therecordisopenedso that

interested parties can learn of their rights, if any, to the deceased contributors remaining

benefits.Because of this confidentiality requirement, we were only allowed access to the

bequeathalinstructionsofdeceasedcontributorswhoseresidualpensionshadbeendisbursed,

and whose files were closed. Names, addresses, and other information that might identify

contributorsortheirrelativeswerewithheld.

OurtotalsampleofSSNITdataconsistsofrecordsof860contributorswhopassedaway

between1992(whenLaw247cameintoeffect)and2006.Themedianageatdeathis54years

(mean52.5years),whichislowerthanthemandatoryretirementageof60years.About70%

of our contributors were married at death; and only 10% are women – reflecting the

overwhelming predominance of men in the formal sector, and a corresponding slighter

predominance of women in the far larger informal sector. The data the SSNIT made available

alsocontainmoreobservationsfromlateryears.Weweretoldthisisbecausemanyolderfiles

areincomplete,missingmuchcriticalinformation,andlessreadilyaccessible.

Table6:SummaryStatisticsofPensionBequestData

Means of key variables for all files, files of matrilineal decedents, and files of patrilineal decedents.

Numbers in square brackets are sample sizes.Final column contains difference between matrilineal

and patrilineal mean, with t‐statistic for the difference bein g  significantly  different from zero in

parentheses.One,two,andthreeasterisksdenote

significancelevelat10%,5%,and1%,respectively.

Sample All Matrilineal Patrilineal

Difference

(t‐stat)

 (i) (ii) (iii) (ii)minus(iii)

Ageatdeath(years)  

Alldecedents 52.5[860] 52.13[421] 52.75[439]‐0.62(0.82)

Maledecedents 53.12[761] 53.14[357] 53.1[404] 0.02(0.20 )

Femaledecedents 47.27[99] 46.56[64] 48.57[35]‐2.01(‐0.83)

Married(fraction)  

Alldecedents 69.7[600] 69.0[290] 71.0[310]‐2.0(0.55)

Maledecedents 71.9 72.5 71.3 1.3(‐0.39)

No.heirslisted  

Alldecedents 2.64 2.85 2.43 0.422(3.25)***

Marrieddecedents 2.94 3.21 2.68 0.53(3.19)***

Maledecedents 2.67 2.96 2.41 0.55(3.79)***

Marriedmaledecedents 2.96[547] 3.29[259] 2.65[288] 0.64(3.63)***

IfclaimsadjustedbyAct560

 4.91[250] 5.06[123] 4.76[127] 0.30(‐1.04)

Bequesttonuclearfamily(%)  

Alldecedents 58.4 59.8 57.1 2.7(0.85)

Maledecedents 58.2 60.5 56.2 4.3(1.30)

Marriedmaledecedents 73.3 74.1 72.7 1.4(0.43)

Pre‐Act560decedents

b

51.8[348] 55.7[168] 48.2180] 7.5(1.52)

Post‐Act560decedents 62.9[512] 62.5[253] 0.633[259]‐0.008(‐0.21)

Act560auditeddecedents 0.547 0.564 0.530 0.034(0.575)

  

Benefitspaid/claim(2006/7)

c

 4,500[85] 5,279[45] 3,622[40] 1,657(‐1.38)

Notes:

a. Lengthoftime(inyears)fromdeathofcontributortocompletionofdisbursementofsurvivorbenefits.

b. Act560,passedin1998,alteredthepermissibledistributionofsurvivorbenefits.

c. InGhanaiancedisperclaim.Theexchangeratein2006/2007wasapproximately¢1.00=US$1.00.





Summary statistics for the variables we construct from these records are reported in

Table6.Eachrecordsetsoutthecontributor’spensionbequeathaldecisionsandidentifiestheir

tribal background, from which we can infer their traditional inheritance custom.Each record

alsoliststhecontributor’smaritalstatus  and most, though not all, provide the average ofthe

contributor’sbestthreeannualincomesfromamongthetwentyyearspriortothecontributor’s

death.

The46.5%ofcontributorsreporting tribalaffiliationsthatimplymatrilineal inheritance

normsalignswellwithanestimated48%forthenationalaverage.Thus,46.5%ofcontributors

chose from the list of permissible heirs under the matrilineal heading in Table 2, and the

remaining53.5%selectedheirsfromthelistunderthepatrilinealheading.

Recallfromsection3thatthepurposeofthisrestrictionistorestrictthecontributorto

leavingresidualbenefitstotheconjugalfamilyorthetraditionallineage.Bequestsofpension

benefitstoothers–e.g.personsnotbelongingtotheconjugalfamilyortraditionallineage–are

proscribed. Thus, the SSNIT does not permit a contributor from a patrilineal tribe to list

maternal uncle as an heir. Should a contributor attempt this, the list would be rejected.In

private conversations with SSNIT staff, we were told that the SSNIT cannot enforce this rule

completely.In practice, a mislabeled maternal uncle might become a heir. SSNIT officials

informed us that they simply lack the resources to thoroughly investigate each list of

beneficiaries and, in the absence of a challenge from other  relatives and if the claims are

procedurally valid, simply distribute remaining funds to the pension recipients’ selected

beneficiaries without further investigation.The record of each contributor’s bequeathal

decision lists the chosen beneficiaries, the fraction of the total benefits bequeathed to each,

andtherelationshipofeachto

thecontributor.

Uponthedeathofacontributor,theSSNITtakesnoaction.Potentialheirsmustsubmit

claims for survivors’ benefits after a qualified contributor dies. SSNIT staff informed us in

private conversations that substantial benefits go unclaimed because heirs are unaware the

benefits exist, and because people who learn they are not listed beneficiaries often fail  to

informthosewhoare oftheirrights.Unsurprisingly,contributorswhowere marriedatdeath

list more beneficiaries: 2.94 versus 1.94 for single contributors.Male contributors list more

beneficiariesthandofemalecontributors,andcontributorsfrommatrilinealtribeslistaslightly

largeraveragenumberofbeneficiaries(2.85)thandothosefrompatrilinealtribes(2.43).

Finally,eachrecordprovidesthetotalvalueofsurvivorbenefitspaidout.Thesecanbe

substantial by Ghanaian standards:the median for the 319 records closed in the 2006‐2007

fiscal year, when the Ghanaian cedi was at ¢1.00 = US$1.00, was ¢2,142; the mean was

¢4,500;andthestandarddeviationwas¢5,539.

Thebequestswerethustypicallyfourtoover

seventimesmorethanGhana’sGDPpercapita,whichthenstoodatonlyaboutUS$500.

Also recall from Section 2 that the 1998 Children’s Act 560 altered the permitted

distribution of survivors’ benefits by the SSNIT.Prior to this, the contributor’s list of

beneficiariesdeterminedthedistributionofallremainingbenefits;butafterwards60%ofthe

totalbenefitsmustgotothecontributor’schild(ren)under18years,regardlessofwhetherthey

are listed as beneficiaries or not and the remaining 40% is distributed in accordance with the

contributor’s list of beneficiaries. The SSNIT theoretically investigates each claim to uncover

other minor children, including illegitimate children, though in practice resource constraints 

limitthis.Oneresultofthisisanincreaseinthenumberofbeneficiariesinrecordsclosedafter

1998becausetheSSNITaddsthenamesofminorchildrentothese.Thus,filesclosedunderAct

560 named an average of 4.91 beneficiaries each, while those closed prior to 1998 named an

average of 2.61 beneficiaries each. This Act thus substantially shifted SSNIT survivor benefits

away from what contributors initia lly intended and towards their own children.This proves

usefulineconometricanalysisbelow.



5. MethodologyandEconometricResults

Wenowexamineeconometricallytheimpactsofthesetworeforms.Ourgoalis toestimatethe

extenttowhichtribalinheritancenormsshapeeconomicoutcomesofthoseonthemarginsof

Ghanaian society; and for the case of retirees, how the reforms influence  private, end‐of‐life

bequest decisions. We begin

 the analysis with individuals’ bequest decisions about their

unexpiredpensions.Becauseweareinterestedin thestatusofwidows,we  first focusonthe

90percentofourSSNITrecordsthataremales.

Not all 319 passed in that year; some died earlier, but survivor-benefits not disbursed until 2006.

5.1 PensionBequestDecisionsandtheChildren’sAct(Act560,1998)

The1998ChildAct560soughttoimprovethestatusofwidowsandtheirchildrenbyinstructing

theSSNITtoadjustsurvivors’benefitsso thatatleast60%ofunexpiredbenefitspasses tohis

minorchildren,regardlessofthecontributor’sinstructions.

Ideally, we would also like to investigate how Act 560 altered contributors’ bequest

decisions. In 1998, the changes were widely publicized, and contributors were urged to alter

theirbequeststoaccordwiththenewrules.Nodoubt,manycontributorsignoredthisadvice,

andlefttheiroriginalinstructionsinplace.Unfortunately,ourdata onclosedSSNITfilesinclude

too few decedents whose initial instructions are dated after 1998 to allow statistically 

meaningfulanalysis.Wethereforecontrasttherecordedbequestdecisionsofcontributorswho

diedafter1999,someofwhichwerechangedtoreflectthenewrules,againstthedecisionsof

contributors who died earlier, and who thus felt no pressure  to alter their instructions to the

SSNIT.Wetake1999,ratherthan1998,asourtransitionyeartoensurethatcontributorshad

sufficientopportunitytoreacttotherulechange.

Todeterminewhetherthenewrulealteredthenumberofheirscontributorslistintheir

SSNITfiles,weestimateregressionsoftheform





(1)



where isanindicatorvariablesettooneifthecontributor’sdeathoccurredafter

1999, and zero otherwise;  is the logarithm of the age at which he died, and the 

are individual characteristics including marital status and an indicator variable for the

contributorbeingamongthetop25%intotalunexpiredpension(Top25Pension)inthecohort 

whodiedinthesameyear.Weinterpretthisindicatorvariableasaproxyforthecontributor’s

totalwealth,whichisunavailable.Finally,wecontrolforage,thecontributor’sageatdeath.

Table 7 reports estimated parameters for (1).The 2.67 grand mean summarizes a

statistically significant increase from a bit over 2.5 heirs per contributor prior to 1999 to just

below three thereafter.The typical contributor dying after 1999 thus names more heirs as

beneficiaries, regardless of customary law. The positive significant coefficient on ln(age)

indicatesthatoldermenalsonamemoreheirsinthebequestsdecisions.Whilethismightbe

explainedbyoldermenhaving had timetosirelarger families,butcouldalsobe duetoolder

menbeingmoreconscientiousabouttheirlegacies.

We include all male contributor files in the above analysis because many males who

declarethemselvesunmarriednonethelessreportchildren.Thesemaybewidowers,menwho

interpret the question as pertaining only to formally  recorded marriages, or men with

illegitimatechildren;ortheSSNITmaynotupdatethisfieldwhenupdatingdataaboutchildren.

Re‐estimatingtheregressionsusingmaleswhoreportthemselvesasmarriedintheirSSNITfiles

generatesbroadlysimilarresults,thoughthepostAct560indicatorvariablenowfailstoattain

significance in the smaller sample.For completeness, we also estimate the regression for

femalecontributors(notshown)andformarriedfemalecontributors.Likemales,olderfemale

contributors have more heirs; but unlike  their male coworkers, females list markedly fewer

heirsafterAct560thanbeforeit.Theverysmallsamplemakesthisresultsomewhatuncertain,

despiteitsstatisticalsignificance.

Concedingtheconsiderablelimitationsofourdataandmethodology,weinferthatthe

dataarenot inconsistentwitha discernable differencebetweenbequestdecisions beforeand

aftertheAct.ThepurposeoftheActwastoinducecontributorstoprovidemorefullyfortheir

conjugalfamiliesandtodivertbequestsawayfromtheirlineages.Toquantifyitseffectiveness,

wecalculatethetotalfractionsofunexpiredpensioncontributors’bequestspassedtodifferent

categoriesofrelatives,implicitlyassigningzerotounmentionedrelations.

 Again, data limitations necessitate caveats.We need not have complete information

about each contributor’s family. For example, no children listed as beneficiaries means the

contributormadenoprovisionsforchildren,notthathehadnone.Hemayhaveneglectedto

updatehisrecordashisfamilygrew;orhemighthavedeliberatelyomittedhischildren.

 We partition each contributor’s heirs into

two groups: conjugal (or nuclear) family –

sons, daughters and surviving spouse(s); and other lineage members. This partition  highlights

the difficulties  we confront in drawing inferences from these data:only 72% of males are

classifiedasmarried.Thisisfar lowerthatthemarriedfractionofthemalepopulation,known

fromcensusrecords;andthereforesuggeststhatmanycontributorslikelydonotupdatetheir

SSNITrecords.BecauseGhanaiancultureexertshugesocialpressureonmentofatherchildren,

mostplacemarriageandraisingafamilyamongtheirhighestlifepriorities.Thisisparticularly

soformenintheformalsector,whoseeconomicpositionsmakethemhighlymarriageable.







Table7.NumberHeirsListedinSSNITRecords

Dependentvariableisthelogarithmofthenumberofheirslistedbythedeceasedmalecontributorto

the SSNIT pension system. Reported values are OLS estimates, and numbers in parentheses are t‐

statistics. One,two,andthreeasterisksdenotesignificancelevelat10%,5%,and1%,respectively.



Sample Allmales

 Marriedmales Marriedfemales

Matrilineal 0.086 (0.90) 0.120 (1.26)

‐0.332 (‐1.46)

Post‐Act560 0.100 (1.66*) 0.098 (1.25)

‐0.577 (‐2.75***)

Married 0.223 (3.40***)‐

‐ 

Top25 0.127 (1.68*) 0.158 (1.54)

‐0.967 (‐0.47)

Post‐Act560*matrilineal‐0.038 (‐0.42) 0.087 (0.74)

0.240 (0.78)

Married*matrilineal 0.122 (1.25)‐

‐ 

Top25*matrilineal 0.017 (0.16) 0.433 (0.31)

‐0.009 (‐0.03)

log(ageatdeath) 0.572 (5.85***) 0.593 (4.28***)

1.034 (3.04***)

Intercept‐1.842

 (‐5.03***)‐1.707 (‐3.16***)

‐2.686 (‐1.95*)

 0.13 0.08 0.26

Noofobs 761 547 53

Finally, every individual, no matter how isolated, belongs to a lineage.The only

conceivable exceptions would be orphaned foreigners from outside Sub‐Sharan Africa who

adoptGhanaiancitizenship.

Our regressions explaining the fraction of residual benefits each contributor in our

sample bequeathed to members of his nuclear family, as opposed to his lineage, which we

denote ,areoftheform



(2)



where the right‐hand side variables are defined as in regression (1).Because the dependent

variable is bounded by the unit interval, with mass at both endpoints, Table 8 reports tobit

regressions.

Unsurprisingly,marriedmenbequeathmoretoconjugalfamilies thandomenlistedas

unmarried.Because of the problem, mentioned above, of stale records, we re‐estimate the

tobitsrestrictingthesampletomendesignatedasmarried.Thethirdcolumnpresentsresults

forfemalecontributors.Again,ageissignificant:oldermenleavemorepensionwealthtotheir

nuclearfamilies.Asabove,thismaybebecauseoldermenhavelongertobuildlargerfamilies,

orbecausetheygrowmoreattachedtotheirconjugalfamilies.

Participants whose deaths occur after the 1999 implementation of Act 560 bequeath

12.7%moreoftheirpensionwealthtotheirnuclearfamilies.ThistooindicatesthattheActhad

an effect:When the law mandated that

 contributors provide more to their nuclear families,

theycomplied.



Table8:Fractionofunexpiredpensionbequeathedtonuclearfamilymembers

Marginal effects estimated from Tobit regressions explaining fraction  of benefits bequeathed  to

surviving spouse(s) and children, as opposed to lineage. Right‐hand side variables are as in Table 5.

Numbersinparenthesesarerobustz‐statistics,adjustedforclusteringbyageofdeath.One,two,and

three asterisks denote significance level

 at 10%, 5%, and 1%, respectively. Note: Pseudo R‐squared

maynotrepresentvariationexplainedbydependentvariables.



Sample Allmales Marriedmales Allfemales

matrilineal 0.17 (1.97)** 0.001 (0.02)‐0.209 (‐0.88)

post560 0.127 (2.45)** 0.098 (1.97)**‐0.168 (‐0.91)

married 0.543 (7.71)***‐‐ 0.353 (1.97)**

Top25‐0.069 (‐1.15)‐0.055 (‐0.94)‐0.215 (‐1.14)

Post560*matrilineal‐0.136 (‐1.91)*‐0.074 (‐1.09) 0.267 (1.18)

Married*matrilineal‐0.146 (‐1.77)*‐‐ ‐0.120 (‐0.58)

Top25*matrilineal 0.034 (0.43)  0.018 (0.23) 0.296 (1.21)

Log(ageatdeath) 0.508 (5.47)*** 0.415 (4.59)*** 0.534 (2.20)**

 0.18 0.04 0.11

Observations 761 547 99

In addition, the data show a secular time trend towards increasingpension allocations

tonuclearfamilies–perhapsbecauseofanongoingerosionoftraditionalvalues.Atimetrend

also accords with a growing social advocacy role for the SSNIT. SSNIT records office staff (in

particular, female staff) shared with us stories about how, over time, they increasingly

assertivelyremindedmenoftheir“responsibilitiestotheir nuclear family”, andtoprovidefor

theirspousesandchildrenwhenlistingheirsintheirSSNITrecords.Atimetrendaddedtothe

regressionsinTable8issignificant,butajumpisnonethelessdiscernibleat1999.Weconclude

thattheActhadaneffect.

Intriguingly, the tobits reveal that the Act’s major effect was not its intended one:

altering bequests by men from matrilineal tribes.These bequeathed more to their conjugal

families before the reform, and did not substantially increase these bequests to conjugal

familiesafter1999.Onepossibleexplanationisthatmenfrommatrilinealtribesusedpension

bequests to circumvent their tribal inheritance customs all along.These results appear

independentofthemagnitudeofunexpiredbenefits.TheSSNITbequestdecisionwas,afterall,

deliberatelyheldconfidentialuntilthecontributor’sdeath,andcould thusprovideprivacyfrom

pressure to adhere to the traditional inheritance system. Indeed, the SSNIT was intended, in

part at least, to provide a defense to men from matrilineal tribes who wished to provide for

theirchildren;butwhofearedthewrathoftraditionally‐mindedrelatives.

To further test if values might be genuinely changing, we gauge for married men’s

“generosity”towardstheirnuclearfamilies.Wedefine“generosity”asbequeathingmorethan 

themandatoryminimumof60%tohisnuclearfamily.Thatis,generosity=1ifpercentNUC >

60%andzerootherwise.

Table 9 presents logistic regressions, similar in form

to (2) but with generosity on the

left‐hand side. Matrilineal and married males are more “generous” than patrilineal and

unmarriedmalestotheirchildren,andallmalesgrowmoregenerousafterAct560.However,

again, no discernible difference is evident in the Act’s effect on men from matrilineal versus

patrilinealbackgrounds.



Table9:GenerosityofPensionBequeststoNuclearFamily

Marginaleffectsare estimated usinglogitregressions explaininganindicatorvariablesettooneif

bequest to surviving spouse(s) and children, as opposed to lineage, exceeds the mandatory

minimumrequiredbylawforminors(60%).Right‐handsidevariablesareasinTable5.Numbersin

parentheses are robust z‐statistics,

 adjusted for clustering by age of death.One, two, and three

asterisks denote significance level at 10%, 5%,and 1%, respectively. Note: Pseudo R‐squared may

notrepresentvariationexplainedbydependentvariables.



Sample Allmales Marriedmales Allfemales

matrilineal 0.235 (2.30)**  0.048 (0.76)‐0.361 (‐2.04)**

post560 0.118 (1.95)** 0.081 (1.47)‐0.346 (‐2.39)**

married 0.550 (10.92)*** ‐‐ 0.373 (2.09)**

Top25 0.035 (0.48) 0.024 (0.37)‐0.278 (‐1.29)

Post560*matrilineal‐0.105 (‐1.20)‐0.025 (‐0.31) 0.414 (2.70)***

Married*matrilineal‐0.133 (‐1.30)‐‐ ‐0.219 (‐0.88)

Top25*matrilineal‐0.118 (‐.018)‐0.02 (‐0.02) 0.218 (1.16)

Log(ageatdeath) 0.558 (5.61) 0.427 (4.71)***

0.523 (2.41)**

PseudoR

 0.22 0.05 0.14

Observations 761 547 99

The generosity of female contributors’ pension bequeathals is remarkably different.

Althoughboth maleandfemalecontributorsaremore“generous”iftheyreportmorechildren, 

femaleformalsectorworkersaredecidedlylessgeneroustotheirconjugalfamiliesiftheirtribal

traditionismatrilineal,andgrowevenlessgenerous afterAct560takeseffect.Policy makers

may wish to consider education programs directed at matrilineal females if further steps to

forcecontributorstoprovidefortheirownchildrenaredeemeddesirable.

Next,weestimatethelikelihoodthattheSSNITascertainsadecedent’sinstructionsto

beinviolationoftheAct560.Also,in caseswherethe SSNITdiscoversaviolationoftheAct,

weexplorethesizesoftheadjustmentsitimposes.



We estimate the probability of the SSNIT finding a violation using the following probit

regression:



(3)



where Act560violated is an indicator variable set to one if the SSNIT decides that the

contributor’sbequestsdecisionviolatesAct560,andtozerootherwise.Table10reportsthese

resultsinitsfirstcolumn.Theinstructionsofmarrieddecedentsareabout19%morelikelyto

violateAct560.Butolder menareactuallyless likely toleaveinstructionsthat violate the Act.

Thismaywellreflecthigheroddsthatoldermenhaveonlyadultchildren.Finally,high‐income

matrilinealmalesaremorelikelytoleavebequestsinstructionstheSSNITmustoverride.

The second  column of Table 10 reports regressions explaining

the magnitudes of the

adjustmentsSSNITstaffmaketobequestsfoundtoviolateAct560.Theseregressionshavethe

form



(4)



TheActgivesSSNITstaffadegreeofdiscretionwhereaviolationisdiscovered.Insuchcases,thebequeststo

minorchildrencanberaisedtomorethantheminimumallowedbytheAct.

where isthepercentageof unexpiredpensionultimatelypaid tothe

decedent’s conjugal family  minus the percentage bequeathed.The left‐hand side variable is

alwaysnon‐negative,butcanbezero; soweemploy tobitregressions.Largeradjustmentsare

madetobequestsbyunmarried,youngermatrilinealmen.

Ourdatado notfullyclarify theircircumstances.Oneinterpretationisthat,despiteall

oftheabovegovernmentpolicyinitiativesandtheSSNIT’sexhortations,many youngmenfrom

matrilinealtribesstillfeelalesserdutytowardstheirwives’children,orperhapstotheirout‐of‐

wedlockchildren.Alternatively,youngmatrilinealmenmightbelessattentivetoupdatingtheir

SSNIT files to record marriages and new children. However, we have no a priori reason them

lessattentivethantheirpatrilinealpeers.Andiftheyarelessattentive,thisalsosuggestsless

concernabouttheirconjugalfamiliescomparedtoyoungmenfrompatrilinealtribes.



Table10.Act560AuditsandBequestAdjustments

The first column is estimated coefficients and z‐statistics for a probit regression explaining which

bequests the SSNIT files ultimately judges in violation of Act 560, which requires 60% of residual

pensionbenefitsbepaidto adecedent’sconjugalfamilyifhehasoneormoreminorchildren atthe

time

ofhisdeath.Thesampleof454filesincludesthoseofallmales whodiedafter1999,whereafter

theSSNITgainedauthoritytoalterbequests.Thesecondcolumnpresentsestimatedcoefficientsand

z‐statisticsforatobitregressionexplainingtheadjustmentstobequeststonuclearfamiliesinthe232

files

theSSNITdeemedinviolationofAct560.One,two,andthreeasterisksdenotesignificancelevel

at10%,5%,and1%,respectively.



Dependentvariable Pr(Act560violated)

Percentpaidminuspercent

bequeathedtonuclearfamily

Regressiontype Probit Tobit



Marginal

effect z‐stat

Marginal

effect z‐stat

matrilineal‐0.035 (‐0.04) 1.324 (1.71)*

married 0.189 (2.34)**‐0.303 (‐4.46)***

Ln(ageatdeath)‐0.487 (‐2.76)***‐0.104 (‐0.71)

Top25‐0.006 (‐0.08) 0.0001 (0.00)

married*matrilineal 0.010 (0.08) 0.066 (0.69)

Ln(Age)*matrilineal 0.002 (0.01)‐0.393 (‐1.96)**

Top25*matrilineal 0.182 (1.69)* 0.016 (0.18)

McFadden’sR

 0.04 0.38

Sample



(No.ofobs)

Allmalecontributorswhodied

after1999

454

Allmalecontributorswho

violatedAct560rules

232

5.2 KnowledgeofLaw111andAssetAcquisitionDecisions

Wenowturnfromrelativelywell‐offformal sector Ghanaian men backtooursurveysof very

poorinformalsectorGhanaianwomen.Here,weexaminethecorrelationbetweenknowledge

of the  Intestate Succession Law 111 and decisions that affect asset accumulation within the

marriage,inheritanceoutcomes,andtheeconomicconditionsofwidows.

AswiththeSSNITfiles,datalimitationspreventcutanddriedassessments.Wecannot

compare responses before and after the Law 111 – the law was enacted many decades ago.

Moreover, knowledge of the law is almost surely not randomly distributed across women.

SocialnetworksamongGhanaiansaregenerallyverystrong,and awomanmaywellbelongto

many social groups – religious organizations, trade groups, and so on. Information about the

lawislikelydisseminatedthroughthesegroups,somoresociallyactivewomenaremorelikely

tolearnofit.Accessto the media –newspapers,radio,andTV–dependson living nearer an

urban center, literacy, literate friends or relatives, access to electricity,and other factors that

quiteplausiblyalsocorrelatewithwidows’economicsituations.Wecancontrolforeducation

andliteracy,butconcedethepotentialimportanceofleft‐outvariables.

Toexploretheseissues,weestimateregressionsofwomen’sknowledgeofLaw111and

other factors on her fractional contribution to her nuclear family’s assets, whether or not she

buildsupownassets,andthelikelihoodofinter‐vivosduringmarriage.Theseregressionsare

oftheform



(5)



whereKnowLaw111isoneiftherespondent indicates

sheknewofLaw111atthetimeofher

husband’sdeathandzero otherwise,educisheryearsofformaleducation,

matri_spouseis

oneifherdeceasedhusbandwasfromamatrilineal tribeandzerootherwise,andXisavector

of control variables:the husband’s education level, the log of their years married, and

dummiesforthewidow’slineagesystembeingmatrilineal,theirmarriagebeingmonogamous,

andthewidowhavingminorchildrenatthetimeofherhusband’sdeath.

This variable is zero in many cases, generally low, so the efficiency gain from using the log of one plus years of

formal education is small. Using the latter approach yields qualitatively similar results throughout.

eKnowLawspousematrispousematriKnowLawY







 111__111

221



Thedependentvariable,Y

in(5),isoneof:themagnitudeofintervivostransfers–that

is, assets the widow received from her husband before his death; the fraction of the conjugal 

family’sassetssheestimatestohavebeenhercontribution;andanindicatorvariablesettoone

ifshehadpersonalassetswhenherhusbanddied,andtozerootherwise.

To gauge the magnitude of inter‐vivos transfers, widows were asked the following

question:Didyoureceivefinancialsupportfromyourspouse whenhewasalive?Theresponses

were (i) not all; (ii) very little; (iii) as expected; (iv) very much. To elicit widows’ estimates of

theirproportionalcontributionto  their conjugalfamilies’assets,we asked themthequestion:

Inyourestimation,whatpercentageoftotalfamilyassets(sincemarriage)isyourcontribution?

Respondentswerepresentedwithsixchoices:(i)zero,(ii)lessthan20%,(iii)20to39%;(iv)40

to59%,(v)60to79%,(vi)80%ormore.Whetherornotthewidowhadpersonalassets,that

wereexplicitlyherproperty,asopposedtothepropertyofherhusbandoralineage,weasked

asimpleyes/noquestion.

Because of the categorical nature of these responses, we employ ordered logit

regressions to estimate (5) regressions on the magnitude of inter vivos transfers and on the

fraction of the conjugal family’s assets the widow estimates as her contribution. A probit

regressionsufficesinregressionsonherpossessionofpersonalassets.

Table 11 summarizes the results of these regressions.Respondents who report

knowledge of Law 111 also report making significantly higher contributions to their nuclear

family’s assets, and are significantly more likely to receive inter‐vivo transfers from their

spouses.Notethatboththewidow’seducationandthatofherspouseareincludedascontrols.

However, having a matrilineal spouse statistically counteracts the inter

 vivo effect entirely,

perhaps indicating persistent tendency of matrilineal men to pass wealth to their lineages.

Widows with matrilineal spouses contributed insignificantly smaller fractions of their conjugal

families’assets–thoughnotsignificantlylessthanwithapatrilinealspouse.



Table11.Economic/FinancialDecisionswithinMarriage

Intervivo

Husbandsgiftsto

wifewhilealive



Contribution

Widow’sreported

contributionto

conjugalfamily

assets

Personalassets

Widowhadpersonal

assetsathusband’s

death

Knowlaw111

1.228*** 0.861** ‐0.043

(3.76) (2.44) (‐0.56)

Matrilineal spouse

0.776* ‐0.031 0.021

(1.92) (‐0.06) (0.21)

KnowLaw x Matrilineal spouse

‐1.675*** ‐0.204 0.081

(‐3.62) (‐0.44) (0.75)

Controls Yes Yes Yes

Regression orderedlogit orderedlogit probit

Sample 297 292 295

Pseudo‐Rsquared 0.042 0.045 0.041

Tablereportsthecoefficientsoforderedlogitregressionsexplainingthewidow’sfinancialsituationat

the time of her husband’s death.The first column summarizes ordered logit regressions explaining

approximate fractionsof theconjugalfamily’sassetsthatwereinter‐vivotransfers; thatis,gifts from

husband to wife during his life. The

 second column summarizes ordered logit regressions explaining

thewidow’sestimateofherapproximatefractionalcontributiontotheconjugalfamily’swealth.The

third column summarizes marginal effects in probit regression explaining the widow having acquired

personal assets.Robust z‐statistics are in parentheses. Controlvariables are: the log of the widow’s

and spouse’s education levels, log of years married, widow’s lineage system, whether monogamous

marriage,andwhetherwidowhadchildrenunder18atthetimeofspouse’sdeath. (*),(**)and(***)

denotesignificancelevelat10%,5%,and1%,respectively.



5.3 TheEconomicStatusofWidows

Ultimately, the purpose of the reforms we study above is to improve the lot of widows and

their children.Our final set of regressions therefore examines the economic outcomes and

opportunities for widows in the informal economy.Based on survey responses in the four

villageswestudy,weconstructthe

followingdependentvariables.

Toassesswidows’economicstatus,weaskedthemtocomparetheiraccesstofinancial

services (formal or informal), healthcare, and education for their children in the years

immediately after losing their spouse to the years immediately before. For each comparison,

respondents were asked to choose between: things got much worse than before the death

(quantifiedas1),slightlyworse(2),thesame(2.5,slightlybetter(3),andmuchbetter(4).Each

widow’s responses to the three questions were averaged, with equal weights, to construct a

changeinoveralleconomicstatus.

Toassesstheefficacyofwidows’supportfromtraditionalsafetynets,wealsoinquired

abouttheleveloffinancialsupporttheyreceivedfromtheirownlineage,andfromthatoftheir

deceased husband.Their responses to each of these two queries is quantified as follows: no

support at all (1); worse support than the lineage’s economic situation could readily have

allowed(2),alevelofsupportroughlyinaccordancewithwhatthe lineage’seconomic situation

allowed(3),andmoresupportthanthelineage’seconomicsituationreadilyallowed(4).

Wecontrolforthe widow’s“closeness”toeachextendedfamilyduringthemarriageby

askinghowoftenthey“visitedorwerevisited

by”membersofeachlineage.Respondentswere

asked: how often they saw own or spouse’s family members during the marriage? They were

give a choice of four responses:  never (quantified as 1), rarely – less than once per year (2),

occasionally–aboutonceevery6months(3),andveryoften,atleastonceamonth(4).

Becauseallthreevariablesarediscrete,werunorderedlogitstoexplaintheminterms

ofotherresponsesfromoursurvey.TheseresultsaredisplayedinTable12.



Table12.Widows’financialSupportandChangesinEconomicStatus

Ordered logit regressions explain affirmative survey response to questions about the generosity of

financial support from the widow’s own lineage (column 1) and her deceased spouse’s lineage

(column2)afterherbereavement,takingintoaccount theoverallwealth ofthelineageinquestion;

aswellasthewidow’sestimateofhow

heweconomicstatuschanged,comparinghersituationinthe

yearsbeforeherhusband died tohersituationin the2to 5yearsafterwards.Controlvariablesnot

shownare:spouse’seducationlevel,logofyearsmarried,andindicatorvariablesforamonogamous

marriage,andwhetherthewidowhavinglivedwith

thespouse.Asterisks(*),(**)and(***)denote

significanceat10%,5%,and1%,respectively.



Financial

support

from

ownlineage

Financial

support

from

spouse’s

lineage

Changein

economic

status

afterwidowed

Changein

relationship

within‐laws

(worsening)

InheritedviaLaw111orawill 0.016‐0.0700 1.049*** 1.053***

(0.05) (‐0.16) (2.64) (2.62)

“Close”toownfamily 0.358** 0.115 0.091‐0.120

(2.30) (0.72) (0.70) (‐0.92)

Matrilineallineage‐0.497‐0.078 0.556* 0.514

(‐1.16) (‐0.18) (1.76) (1.35)

“Close”tospousefamily 0.110 0.323** ‐0.064 ‐0.219**

(0.79) (2.24) (‐0.62) (‐1.96)

Matrilinealspouse 0.196‐0.013

0.329 0.565

(1.92) (‐0.03) (0.98) (1.51)

Childunder18yrs 0.634** 0.269 ‐0.913***‐0.654***

(2.03) (0.87) (‐3.85) (‐2.71)

Educationlevel 0.004‐0.047 0.091*** ‐0.0001

(0.15) (‐1.43) (2.96) (0.01)

Othercontrols Yes Yes Yes

Yes

Sample 296 297 295 300

R‐squared 0.044 0.046 0.038 0.062



Unsurprisingly,awidows’“closeness”toherownextendedfamilycorrelateswiththeir

providinghergreaterfinancialsupportforherduringwidowhood.Asimilareffectisevidentfor

thewidows’“closeness”totheirdeceasedhusbands’extendedfamily,withanalmostidentical

coefficient.Minorchildrenelicitstrongerfinancialsupportfromherownextendedfamily,but

notfromherhusband’slineage.

The third column of Table 12 shows that a widow’s economic status deteriorates less

uponthedeathof her husbandifherlineage  ismatrilineal,butmore sharply ifshehasminor 

children to support.The first result is somewhat at odds with the presumption of many

advocacy groups and government officials that matrilineal widows are less financially secure;

though the latter suggests that more could be done to support widows with minor children.

Weinquiredcasuallyaboutwhatsortsofsupportmatrilineallineagesprovided,andweretold

of brothers’ foster parenting their nieces and nephews, and of the lineage overall providing

economic situations widows deemed more appropriate. The last column in Table 12 sheds

somelightonhowlegalreforms,well‐intendedastheymaybe,canleadtotensionsbetweena

spouseorwidowandherin‐laws.Whiletheresultsshowthatwidowswithclosertiestothein‐

laws during marriage and those raising younger children tend to experience improved

relationship with the spouse’s lineage members after his death, the use of legal structures to

inheritassets(e.g.,willorLaw111),leadstoaworseningrelationshipwiththein‐laws.

Advocacygroupscan,however,clearlytakeheartfromthesignificantcoefficientinthe

first row of the third column.The very few widows – about 11% of the sample for this

regression – who either made use of Law 111 or inherited under a will report a significantly



better shiftineconomicstatusaround their husbands’ deaths. The coefficient is economically

large:almosttwicethatassociatedawidowbeingofmatrilineallineage.Finally,all elseequal,

widowsaresubstantiallybetteroffiftheyareeducated.



6.Conclusions

Traditional Ghanaian cultural norms, especially as regards customary inheritance rules, differ

starklyacrosstribesandfromfamiliarWesternandAsiannorms.Reformerssawsomeofthese

traditions deterring wealth accumulation among the country’s poorest citizens, and

impoverishingwidowsandtheirchildren.

These concerns were most voluble regarding the traditions of matrilineal tribes, in

which men considered their sisters’ children, not their wives’ children, to be their nearest

relatives in the next generation.These traditions were thought to leave widows and their

children economically vulnerable.This latter result seems consistent with that traditionalist

argument that brothers and other members of matrilineal lineages support widows, including

foster‐parenting the widow’s children. Matrilineal traditionalists countered that their customs

heldbrothers,maternaluncles,andthe likeaccountableforsupportingtheirwidowed female

relatives,andthatthissystemworkswell.

Ghana’s governments has enacted two major legal rules designed to nudge men of all

tribal traditions towards making decisions that would better protect their own children, thus

bringing Ghanaians into closer accord with global cultural norms. First, the 1986 Intestate

SuccessionLaw111changedtherulesfordividinguptheestateofindividualwhodieintestate.

Thisreformletswidowsfileformallegalprocedurestosecureownershipofcertainclasses’the

conjugal family’s assets, and was a major departure from customary rules that assigned most

assets to the decedent’s lineage (extended family) which, if the deceased belonged to a

matrilinealtribe,doesnotincludehiswidowandchildren.

Our extensive survey of widows living in matrilineal and

patrilineal traditional village

societies shows that Law 111 is little used, even by women familiar with it.Law 111 is

procedurallytouchy,necessitatingexpensive legalrepresentationthatmakesitsuseexpensive

beyond the means of people in the traditional economy. Its use also overtly challenges

traditional law, a sensitive course  of

 action for anyone who must reside in the traditional

culture for the rest of her life.Those few widows who do use Law 111 are financially more

secure,andevidencethattheybearsocialcostsisunpersuasive.

Oursurveyevidencefindmatrilinealwidowstobemorefinancialsecurethanpatrilineal

widows.Thissupportsthepositionofmatrilinealtraditionaliststhatbrothers,uncles,andother

matrilineal lineage members, support widows financially, acting as de facto foster‐parents for

herchildren.

However, our survey also supports the efforts of reformers.Widows who used the

formal legal system to settle their husbands’ estates – either by executing a formal will or by

applying Law 111, gain almost twice the financial advantage associated with a matrilineal

custom.Further, we find evidence consistent with mere knowledge of the law providing an

incentiveforcouplestobuildupfamilyassetsjointly,andtomotivateinter‐vivotransfersfrom

husband to wife, irrespective of education level. But surprisingly, this is least evident among

widows whose spouses have a matrilineal lineage tradition – the very people the reforms

focusedonadvancing.

These findings are consistent with knowledge of the  law encouraging people to

accumulate wealth.However, we cannot preclude the possibility that people who have

accumulated assets are more concerned about losing them and therefore seek information

aboutthelawregardinginheritances.

The survey data also show that lineages remain an important safety net. Widows who

hadclosetieswith ownrelativesobtainmorefinancialsupportfromtheirlineages;andwidows

whohad closertieswiththeir in‐lawslikewisereportmorefinancial support from his lineage.

Whileformallegalrightsmightgivewidowsmorebargainingpower,theirscantusagesuggests

majorobstaclesimpedeaccesstothelawforthevillagerswesurvey.

The second major reform we  examine

is the 1998 Children’s Act 560, mandating that

the Social Security and National Insurance Trust (SSNIT), the sole pension administrator,

allocate 60% of deceased contributors’ unexpired pensions to their minor children, regardless

of their written instructions.The SSNIT reform affects only the estates of Ghanaians who

workedintheformalsector,andthuscontributedtothepensionsystem;whileLaw111sought

toaffectwidowslivingatthemarginsoftheGhanaianeconomy.

Regression analysis of SSNIT records is somewhat problematic. First, we could only

accessclosedfiles–thoseinwhichallbequestshadbeenfullypaidout.Asmany

filesremain 

open because the beneficiaries are unaware of their rights, our data suffer from unknown

sample selection problems. Second, conversations with SSNIT staff and management indicate

that many contributors’ bequest instructions were made when they first filled out a SSNIT

membership from and never updated.This “stale data” problem weighs against our finding

significanteffectsofthelegalreformsaffectingSSNITsurvivors’pensionrights.

Nonetheless, we find some limited evidence consistent with the reforms benefiting

decedents’nuclearfamilies.MaleswhodiedafterAct560 wasimplementedallocatemoreof

their pensions to the nuclear families; indeed, they bequeath significantly more than the

mandatory 60% minimum prescribed by the law.Also, when the SSNIT judges a bequest

decisiontoviolateAct560,theadjustmentsitimposesareovertlyfavorabletonuclearfamilies,

especiallythoseofdecedentswithmatrilineallineagetraditions.

While our analysis finds some evidence of successful reform, we feel their deeper

message is a confirmation of the tenacity of African cultural traditions.Formal legislation

adopted in Accra does not change the social, economic, and spiritual forces that constrain

people’s decisions.Genuinely effective  reforms appear to require intense efforts to promote

social awareness and provide legal aid for those who would exercise their rights under the

reforms,andeventhencanachieveonlyqualifiedsuccess.

OurfindingscomplementrecentworkbyScholzandGomez(2004).Theirexamination

of formal‐sector inheritance rights in Botswana, Ethiopia, Ghana, Nigeria, Rwanda, Senegal,

South Africa, Swaziland, Zambia and Zimbabwe leads them to conclude that traditional

inheritancenormsprevailregardlessofthepaperlaw.Ourfindingssuggestthattheabsence

of

resourcestoexerciseformallegalrightsmaywellbemoreimportantthanalackofinformation

aboutthoserights.Absentformalsocialassistanceorlegalaidprograms–whicharetenuous

to non‐existent in all of these countries, poor Africans rationally conclude that the economic

security, however minimal, offered

 by one’s lineage is too valuable to sacrifice for uncertain,

inaccessible,andofteneffectivelyunenforceableformallegalrights.

IfAfricangovernmentswishtoeffectreformstoinheritancecustoms,theirformallegal

reforms need bolstering by awareness campaigns, meaningful rule of law, and a sensitivity to

existingtraditions.However,challengestolongstandingtraditionsarelikelytomeetsustained

resistance in Africa, as in Asia and in the Western world. Our findings also suggest that

governmentsseekingto counterculturalnormsthoughttodeterdevelopment mightconsider

designingformallegalrightstoempowerpeoplewithinthecontextoftheirtraditionalcultures.

Matrilinealwidows’economicwelfaremaywellhavebeenimprovedbythefactthattheycould

challengetraditionalinheritancenorms,eveniffewactuallydid.Thelawsoundshollowwithout

anaccompanying“highlife”drumbeat.



References



Alger,IngelaandJorgenWeibull(2008),"Family  ties, incentivesanddevelopment:amodelof

coerced altruism,"in Basu, K., and R. Kanbur, eds., Arguments for a Better World:

Essays in Honor of Amartya Sen, Volume II: Society, Institutions  and Development,

OxfordUniversityPress:Oxford

----- (2010), "Kinship, incentives and evolution," American Economic Review Vol. 100, 2010,

pp. 1725-1758

Awusabo‐Asare,K(1990)“MatrilinyandtheNewIntestateSuccessionLawofGhana,”Canadian

JournalofAfricanStudies,24(11),1–16

Becker,GaryS.(1981);ATreatiseontheFamily,HarvardUniversityPress,Cambridge

Bertrand, Marianne;Sendhil Mullainathan and Douglas Miller (2003), “Public policy and

extendedfamilies:EvidencefromSouthAfrica;WorldBankEconomicReview,vol17(1),

pp.27–50

Chiteji,N.S.andDarrickHamilton(2002),“FamilyconnectionsandtheBlack‐Whitewealthgap

amongmiddle‐classfamilies”ReviewofBlackPoliticalEconomy,vol30(1),pp.9‐27

DeSoto,Hernando(2000), The MysteryofCapital:WhyCapitalismTriumphsin theWestand

FailsEverywhereElse.BasicBooks

EllulAndrew,MarcoPaganoandFaustoPanunzi(2010),“InheritanceandInvestmentinFamily

Firms,”AmericanEconomicReview,100(December),pp2414–2450.

Fenrich, Jeanmarie and Tracy  E. Higgins (2001), “Promise Unfulfilled: Law, Culture, and

Women'sInheritanceRightsinGhana,”

FordhamInternationalLawJournal,25pp.259‐

341:http://ir.lawnet.fordham.edu/ilj/vol25/iss2/1 

FIDA (2007), “The Impact of Intestate Succession Law on Women’s Inheritance:Accra,Kumasi

andKoforidua,” InternationalFederationofWomen’sLawyers,Ghana,ResearchReport,

Accra.Ghana.

Hacker, Daphna (2010), “The Gendered Dimensions of Inhe ritance: Empirical Food for Legal

Thought,”TelAvivUniversity

LawSchoolPaper#116.

Heintz,James(2005),“Women,Work,andPovertyinGhana”.PresentedatConferenceonthe

Informal Economy in Ghana: A Comparative Perspective, co‐hosted by the Ghana

StatisticalServiceandWomeninInformalEmployment:GlobalizingandOrganizingand

heldatLaPalmRoyalBeachHotel,Accra,

Ghana,October24,2005.

Hoff,KarlaandArijitSen(2006),“Thekinsystemasapoverty trap,”inSamuelBowles,Steven

N. Durlauf and Karla Hoff (eds), Poverty traps, Princeton University Press, Princeton,

NewJersey

La Ferrara, Eliana (2006), “Descent Rules and Strategic Transfers: Evidence from Matrilineal

GroupsinGhana,”JournalofDevelopmentEconomics,83pp.280‐301

OllennuN.A. 1996; The Law ofTestateandIntestateSuccessioninGhana.London:Sweetand

Maxwell

Plateau,Jean‐Phillipe(1991),“TraditionalSystemsofSocialSecurityandHungerInsurance,’’in

E. Ahmad, J. Dreze, J. Hills and A. Sen (eds), Social Security in  Developing Countries,

Oxford:ClarendonPress;112–170

Quisumbing, A., Otsuka K. (2001), “Land Inheritance and Schooling in Matrilineal Societies:

EvidencefromSumatra,”WorldDevelopment,vol29(12),2093‐2010.

Quisumbing,AgnesR.,EllenPayongayong,J.B.Aidoo,andKeijiroOtsuka.(2001),“Women’sLandRights

in the Transition to Individualized Ownership: Implications for theManagement of Tree

ResourcesinWesternGhana,”EconomicDevelopmentandCultural

Change50(1):157‐182.

Scholz, Birte and Mayra Gomez (2004), Bring Equality Home: Promoting and Protecting the

Inheritance Rights of Women; The Centre on Housing Rights and Evictions (COHRE)

SeriesonLawandPracticinginsub‐SaharanAfrica

Woodman, Gordon R. (1985), “Ghana Reforms the Law of Intestate Succession,” Journal of

AfricanLaw,vol29(2),pp.118–128.