A Loveless Economy? Intergenerational
Altruism and the Marriage Market in a
Tuscan Town, 1415-1436
MARISTELLA
BoTTmcim
This article examines the role of dowries andhighlights the variablesthat affected the
size of dowries in fifteenth-century Tuscany. The estimation, which matches the
households found in the marriagecontractswith the correspondinghouseholds in the
Florentine Catasto of 1427, offers support for the present net-value hypothesis and
for the altruism model. Results indicate a positive correlation between a bride's
dowry size and her age when used as proxy for her contributionto the maritalhousehold. Parents also provided their daughterswith largerdowries when they married
"down" into relatively less wealthy or socially prominenthouseholds.
medieval and Renaissance Italy, marriagesin which the bride's household provided the dowry were the norm. Resurfacing in the early Middle
Ages, the institution of the dowry survived in Italy for many centuries, disappearing only a few decades ago.' The dowry was administered by the
groom, or the groom's father, who could use and invest the dowry. However, if the groom died, the dowry returnedto the bride. The bride did not
control the daily management of her dowry, but retained legal ownership
over it.2Parentshadto provide dowries for all their daughters,even for those
In
The Journal of Economic History, Vol. 59, No. 1 (March 1999). C The Economic History
Association. All rights reserved.ISSN 0022-0507.
MaristellaBotticini is Assistant Professor,Departmentof Economics, Boston University, 270 Bay
State Road, Boston, MA 02215. E-mail: [email protected].
I wish to thankGaryLibecap(the editor),Kyle Brinkmanand SusanIsaac (the assistanteditors)and,
especially, two anonymousrefereesfor several commentsthat substantiallyimprovedthis article.I am
indebtedto Joel Mokyrfor his invaluableguidanceandinsights;EdwardMuirandWilliamRogersonfor
many helpful discussionsand suggestions;LorenBrandt,Saul Engelbourg,Joe Ferrie,GillianHamilton,
Anne McCants,DeirdreMcCloskey, CormacO'Grada,Aloysius Siow, ChrisUdry, as well as seminar
participantsat Boston University,NorthwesternUniversity,Universityof Iowa, Universityof Western
Ontario,thePopulationStudiesCenteratBrownUniversity,theWorldCliometricsConferencein Munich
(July 1997), the CanadianEconomic History Conferencein Niagara-on-the-Lake(May 1997), and the
Social ScienceHistoryAssociationmeetingin Washington,DC (October1997)forvery helpfil comments
andreferences;Nancy Bernhautforherexcellentjobwith copyeditingthisarticle;andFrancescoBonomi
for his simple yet powerfulsearchprogramon the 1427 FlorentineCatasto.Financialsupportfrom Ente
Einaudi(Rome) is gratefullyacknowledged.This articleis a revised version of chapter3 of the Ph.D.
dissertationI wrote for the Departnent of Economics at NorthwesternUniversity.
1 While an intriguing issue, this article does not aim to explain why medieval and Renaissance
Tuscanywas characterizedby a dowry system, andnot by abrideprice system. For excellent references
on the historical developments of dowries and bride prices, see Boserup, Women'sRole; Goody,
"Bridewealth";Herlihy, "Medieval Marriage Market";and Hughes, "Urban Growth" and "From
Bridepriceto Dowry."
2 If the dowry consisted of land holdings, the husbandhad to get his wife's permissionbefore selling
any land plots. In contractsdrawnup by notaries,in which husbandssold theirwives' land plots, there
was always a preamblein which the wife agreedto the sale. Historianshave strongly debatedwhether
104
A Loveless Economy?
105
who entered convents as nuns, although in the lattercase the dowries were
usually smaller.3
The dowry was a conditio sine qua non for the marriage. Even orphan
girls reared in charities received small dowries when they married.4Provisions for these dowries were made either by the charities themselves or by
people leaving sums of money in their wills explicitly for the purpose of
creating dowry funds.5 In these wills it was common for money to be bequeathed "to provide dowries for the poor girls of our town."6 With the
Monte delle Doti (Dowry Fund), Florence created a public funded debt in
which households could invest a share of their wealth, get a good returnin
terms of interest earnings, and upon maturity obtain a sum of money to
provide their daughterswith dowries.7
Dowries are a worthwhile topic of study for two reasons. First, in medieval and Renaissance Tuscany dowries representedvery large transfers of
wealth from bride households to groom households. In Cortona between
1415 and 1436 the averagedowry amountedto 125.5 florins, with 14 florins
being the annual average salary of an urban worker in Florence in 1427.
These transfersof wealth must have affected, and must have been affected
by, households' savings behaviorandwealth accumulation.Moreover,while
most studies of dowries, bride price, and marriagemarketsusually focus on
ruraland developing regions, Tuscantowns were vibrantcommercialcenters
during the Middle Ages and the Renaissance.8
Second,froma theoreticalviewpointthedowryis intriguingbecauseit is both
a marriagepaymentand an intergenerational
transfer.The formeris the aspect
studiedmoreextensivelyby economists.9Dowriesarepaidby bridehouseholds
to groom households when daughtersmarry;without a marriagethere is no
dowry.If marriagetransactionsoccurin a marketsetup,the dowryshouldbe the
price of the goods being purchasedand sold. It is not obvious, however, what
exactly is being purchasedandsold in the marriagemarket,norwhatthe dowry
is paymentfor.Whenviewed as marriagepayments,dowriesmay greatlyaffect
women in the Renaissance gained or lost control over wealth and assets-and over their dowries in
particular (see, for example, Brown and Goodman, "Women"; Chabot, "La reconnaissance";
Chojnacki, "Dowries," "MarriageLegislation," and " Most Serious Duty"; Cohn, Women;Gregory,
"Daughters";Hughes, "Urban Growth" and "From Brideprice";Kirshner, "Wives' Claims" and
"Materials";Kirshner and Molho, "Dowry Fund"; Klapisch-Zuber,"Cruel Mother" and Women;
Kuehn, "Cum Consensu Mundualdi"and Law; Molho, Marriage Alliance; and Riemer, "Women").
I Trexler,"Le c6libat."
' enderson, Piety, pp. 316-20.
'Cohn, Women,pp. 76-97.
6 See, forexample,StateArchives
ofFlorence(hereafterASF),NotarileAntecosimiano
18912,fols. 9r,48r.
See Kirshnerand Molho, "Dowry Fund";and Molho, Marriage Alliance.
Forworks onIndiaandAfricasee Grossbard-Shechtman,
OntheEconomics;Jacoby,"Economics";
and Rao, "Rising Price."
I See, for example, the work of Becker, Treatise;Cigno, Economics; Grossbard-Shechtman,On the
Economics; and Rao, "Rising Price."
106
Botticini
the functioningof marriagemarkets.These markets,togetherwith the institutions and social norms that regulatethem, may influence economic growth,
consumptionsmoothing,savingsbehavior,and female laborsupply.10
On the other hand, the study of the dowry as an intergenerationaltransfer
from parents to their daughters has been neglected by economists.
Intergenerationaltransfersfrom parentsto childrencan be either inter vivos
transfers or bequests. Dowries fall into the first category. The timing of
intergenerationaltransfers can have economic implications in terms of the
savings behavior of the parents and the labor supplied by their children.
Children receiving inter vivos transfersmay behave differently than those
receiving bequests in terms of the attention and care they are willing and
able to provide to their parentsas they grow old.
This article analyzes the role of dowries and highlights the variables affecting the size of dowries in fifteenth-centuryTuscany.When considering
the dowry as a marriagepayment, the main issue is the amount the bride's
parentsare paying in the marriagemarket.To addressthis issue, this article
describes and tests the presentnet value hypothesis, according to which the
dowry depends upon the value of the bride's contribution to the marital
household. Ceteris paribus, the larger the contribution of the bride to the
marital household, the smaller the dowry her parents will have to pay to
convince the groom household to accept their daughter.
When considering the dowry as an intergenerationaltransfer, the key
concern is whether parentsare altruistictowardtheir daughters.In an altruism framework, if parents care about their daughter's well-being they will
provide a largerdowry when she marries"down"into a less wealthy or less
prominenthousehold than when marrying"up."
The estimation, based on a sample of 224 marriagecontractsfor the Tuscan town of Cortonabetween 1415 and 1436, supportsboth the present net
value hypothesis (dowries and the bride's age are positively correlated)and
the altruism model (parents provided their daughters with larger dowries
when the daughtersmarrieddown). " The latterfinding contradictsthe bleak
picturethatthe literaryevidence provides on parent-childrelationshipsin the
Middle Ages.12 Quite interestingly,this article sheds light on the marriage
strategies not only of wealthy and prominenthouseholds, but also of very
humble peasants, small artisans, and poor urban workers-those groups
1' See Cole, Mailath,andPostlewaite, "SocialNorms"for an intriguinganalysisofmarriagemarkets,
social norms, savings behavior,and economic growth.
"1Cortonawas no small village in the Florentineterritories.Sold in 1411 to Florence by Ladislao
D'Angio for 60,000 gold florins, at the time of the Catasto of 1427 Cortonawas the sixth most populous town belonging to the FlorentinedominionafterFlorence, Pisa, Arezzo, Pistoia, andVolterra,and
it had a large countryside(Herlihy and Klapisch-Zuber,Les Toscans;Mancni, Cortona; and Uccelli
Storia di Cortona). In other words, Cortonaoffers the perspective of a large town in this region for
which numerousnotarialcontractssurvived.
12See, for example, Klapisch-Zuber,Women.
A Loveless Economy?
107
whose social and economic behavioris challenging to study due to the paucity of memoirs or other recordsthat may document their history.
The article begins with a description of the main features of the Tuscan
marriagemarket.Next it discusses a conceptualframeworkfor understanding dowries and introduces the hypotheses to be tested. The article then
describesthe sample, presentsthe estimation,and ends with some directions
for futureresearch.
THE TUSCAN MARRIAGE MARKET IN THE FIFTEENTH CENTURY
To formulatehypotheses about the use of dowries, we need to know the
basic facts about the Tuscan marriage market. The reading of medieval
lettersanddiariesreveals threeaspects aboutthis market.First,parentscared
about their future son- or daughter-in-law'spersonal qualities. Second, the
qualities of futurein-laws were also of the utmost concern. Besides wealth,
parentscaredaboutthe prestige,the social andpolitical standingandconnections of their future in-laws. Third, providing the dowry was the critical
element of the marriagenegotiations.13
The most strikingfeatureof Tuscan society, as depicted in the Catasto of
1427 (a census and propertysurvey of the Florentineterritories),is the absence of the so-called Europeanmarriagepattern,which has been shown to
dominatein modernWesternEurope.14 In latemedievalandearlyRenaissance
Tuscany, women marriedin their late teens while men marriedmuch later.
Celibacyamongwomen was virtuallyunknown:97 percentof Tuscanwomen
were marriedbefore the age of 25. Meanwhile, many men, especially those
living in towns, delayed getting married, or never married:just 47 percent of
Florentine men were married at the time ofthe Catasto of 1427. The mean age
for women at their first marriage was 18.8 with a mode at 16, while for men
the two values were 27.8 and 25, respectively.15 The age gap between men and
women when marrying for the first time was lower in the countryside, where
men married a bit earlier than their urban counterparts.
In Cortona, too, a ten-year gap existed between the age of the bride and
the age of the groom (Table 1).
On average, women married in their late teens and men in their late twenties. Bride households were, on average, slightly richer than groom households, and had more children. Assortative mating did not occur according to
wealth: the correlation coefficient between the wealth ofthe bride household
and the wealth of the groom household is just 0. 14. Meanwhile, there was
13 Moiho, Marriage Alliance, pp.
130-32.
14 Hajnal,"EuropeanMarriagePattems,"The Europeanmarriagepattern,which uniquelycharacter-
izes Westem Europe in modem times, presents two key elements: older age of women marryingfor
their first time and a higher percentageof people who never married.
lSHerlihy and Klapisch-Zuber,Les Toscans,p. 537, fig. 24.
108
Botticini
TABLE1
SUMMARY STATISTICS OF MARRIAGES IN CORTONA, 1415-1436
Dowry (in florins)
Groom's age
Bride's age
Groom household's wealth (in florins)
Bride household's wealth (in florins)
Number of children in grooms' households
Percentage of daughters in grooms' households
Number of children in brides' households
Percentage of daughters in brides' households
N
Mean
Median
Standard Deviation
125.5
28.1
18.8
609.7
700.7
2.25
0.08
3.14
0.65
70
27
18
164
196
2
0
3
0.6
224
105.9
8.3
4.7
1,692.84
1,997.66
1.87
0.17
2.33
0.27
Note: The marriages refer both to households living in the town of Cortona and in 44 villages in its
countryside.
Sources: Florence, State Archives (hereafter ASF), Catasto 213, 214, 215, 216, 252, 253, and 254;
Notarile Antecosimiano 1143, 1144, 1145, 1146, 5441, 10038, 18905, 18906, 18907, 18908, 18909,
18910, 18911, 18912, 18913, and 18914.
a strongassortativematingregardingtheplaceof residence,withhouseholds
living in the town of Cortonaor in the villages in the countrysidechoosing
sons- anddaughters-in-lawfromtheirrespectiveareas. Assortativemating
also occurredwith regardto the type of occupation.Outof 224 marriages,
54 percentwere betweena peasantbrideanda peasantgroom,and 14 percent betweena nonpeasantbrideanda nonpeasantgroom.
Anotherstrikingfeatureof themarriagemarketin Cortonais thelargenumberof womenwhomarried"down."Marryingdownis definedaccordingto two
Thismakessensein societiessuchasmedieval
variables:wealthandoccupation.
andRenaissanceTuscany,inwhichthesocialstandingofa householddepended
on both its wealth and prestige-the latteroften being determinedby the
groom'soccupation.A bridefrom a peasanthouseholdcould eithermarrya
peasantgroom,or marry"up"into a merchantor a notaryhousehold.Conversely,a bridefroma merchantornotaryhouseholdcouldeithermarrya merchantgroom,or marrydown into a peasanthousehold.Ceterisparibus,both
typesof bridescouldmarryupordownas faraswealthwasconcerned.Accordingto thesetwocriteria-wealthandoccupation-thereareeightpossiblecomof bridesfrom
binations,as indicatedin Tables2 and3. A significantproportion
bothpeasantandnonpeasanthouseholdsmarrieddown.
It may seem oddthatso manywomenmarrieddown.Accordingto David
Herlihyand ChristianeKlapisch-Zuber'sstudyof the FlorentineCatastoof
1427,womenwere scarcein Tuscany,andthis by itself shouldhave favored
womenmarryingup (seethe sex ratiosin Tuscantownsdisplayedin Appendix
Table1).Indeed,thetownof CortonawastheonlyTuscantownwitha surplus
of women.As Table2 shows,womenmarryingdownwerealsocommonin the
countrysideof Cortona,wherewomenwerescarceas in the restof Tuscany.
A Loveless Economy?
109
TABLE2
PROPORTION OF WOMEN MARRYING UP OR DOWN IN WEALTH AND OCCUPATION
(BRIDES FROM PEASANT HOUSEHOLDS)
Occupation
Wealth
Same
MarryUp
Total
Marrydown
Marryup
Total
0.49
0.31
0.8
0.08
0.12
0.2
0.57
0.43
Notes: N = 152. The figures representpercentages.
Sources: See Table 1.
The significant proportionof women marryingdown can be explained by
high wealth concentrationin late medieval Tuscany coupled with the necessity for women to marrywhen young. Wealthwas highly concentrated:in
Florence itself, the wealthiest 1 percent of households owned more than a
fourth of the overall wealth of the town.'6 Moreover, while Tuscan men
preferredto delay getting married,women had to marrywhen very young. 17
These three factors-wealth concentration,delayed marriagesfor men, and
early marriagesfor women-along with a less favorablesex ratiofor women
in the fourteen to twenty-four age cohort made it quite difficult for a bride
household to find a "suitable"groom, particularlyif the bridehousehold was
wealthy or prominent.
THE DOWRY AS MARRIAGE PAYMENT AND INTERGENERATIONAL
TRANSFER
Study of the dowry as a marriagepaymentcenters aroundthe amountthat
the bride's parentspay. This article proposes the present net value hypothesis, which states thatthe dowry dependsupon the value of the bride's contribution to the marital household. The bride's contribution is based on the
value of the housework (and possibly farmwork) she will perform,the love
and pleasure she will bring to her husband,her ability to bear children and
to rear and supervise them, minus the value of her lifetime consumption.
While housework has positive value in both the paternaland the marital
households, it has greaterworth in the maritalhousehold because there the
bride's housework is regardedas essential, while in the paternalhousehold
the bride's mother can performthe same housework. The love and pleasure
16 Herlihy
and Klapisch-Zuber, Les Toscans, p. 340.
reasons have been offered to explain why men, especially those living in urban centers,
often married later in life. Young men, particularly merchants, notaries, artisans, and medical doctors,
had to accumulate enough wealth before starting a career on their own. Moreover, to prevent the
household's patrimony from becoming too fragmented, sons, especially cadets, remained bachelors,
or delayed their marriage until the parents' death.
17 Various
Botticini
110
TABLE3
PROPORTION OF WOMEN MARRYING UP OR DOWN IN WEALTH AND OCCUlPATION
(BRIDES FROM NONPEASANT HOUSEHOLDS)
Occupation
Wealth
Marry down
Marry up
Total
Marry Down
0.2
0.3
0.5
Same
Total
0.32
0.18
0.5
0.52
0.48
Note: N = 72. The figures represent percentages.
Sources: See Table 1.
a bridebringsto her husbandand the birthof childrenhave zero, or even
negative,value in the paternalhousehold,while they have positive value in
the maritalhousehold.In the paternalhousehold,the daughterrepresentsa
potential financial loss, which includes the monetaryvaluation of the
"shame"thatwould follow an out-of-wedlockpregnancy.
Consequently,the bride'sparentsalwayshave an incentiveto find a husbandfor theirdaughterin orderto minimizetheirfinancialloss. The older
theirdaughter,however,the greatertheirfinancialloss, andthe smallerthe
probabilitythatshe canfmda husband.An "old"bridewill be less attractive,
and,more importantly,will not be ableto give birthto as many childrenas
a youngerbride.A womanwaitingtoo long beforemarryingwill become
too expensiveto "sell"in the marriagemarket:herparentswill have to pay
a largerdowry to compensatethe groom's householdfor the smallernet
positive contributionthe bridewill providein the maritalhousehold.
The presentnet value hypothesispredictsthatthe youngerthe bride,the
largerher net positive contributionto the maritalhouseholdand,therefore,
the smallerthe dowryher parentshave to pay. While the "optimal"age at
which a womanshouldmarryis equalto, say, t1 for all women, it may happen thatsome womenareunableto marryatthatage. Parentswith unattractive daughtersmay encounterdifficulty finding husbandsfor them. Poor
householdsmay have difficultyprovidingthe dowrywhen theirdaughters
reachthe age of t, andmayhaveto delayanymarriage.The longera bride's
parentsdelay, however,the largerthe dowry they will have to provideto
convincethegroomhouseholdto accepta "lessvaluableproduct."Underthe
presentnetvaluehypothesis,largerdowriesareassociatedwitholderbrides.
The dowry,though,is not only a marketprice;it also representsa wealth
transferfromparentsto theirdaughters.The studyof the dowryas an intergenerationaltransferis primarilyconcernedwith the altruismshown by
parentswhenmakingtransfersto theirdaughters.In analtruismframework,
parentscaringfor theirchildren'swell-being compensatethe "less lucky"
childrenby makinglargertransfersto them.Inthecase of dowries,if parents
arealtruistictowardtheirdaughters,the dowriesprovidedwhenthey marry
A Loveless Economy?
111
down into less wealthy or less prominent households will be larger than
those given when marryingup.
Suppose that parentsare selfish and use their daughter's marriageto enhance their own social and economic standing.Whenparentsmake a daughter marry up, we expect them to pay a high dowry, in order to secure the
desirable match. However, what would be the purpose of paying a higher
dowry in the marryingdown case, unless parentsare concerned about their
daughter's well-being? If parents are altruistic and wish to ensure their
daughter,who is marryingdown, a standardof living comparableto the one
she had in the paternalhousehold, then they would provide a larger dowry
than in the marryingup case.
Alternatively, another test of the altruism model consists of checking
whetherthere is fair division of household wealth, with parentsbequeathing
their children equal shares. If parentsare altruistic,two daughtersmarrying
grooms of similar backgroundswill receive equal dowries.
THE SAMPLE
Data to test the presentnet value hypothesis and the altruismmodel have
been collected from two primarysources. First, informationon the size of
dowries is availablefrommarriagecontractsdrawnup by notaries.Theparties
anticipatingmarriageappearedbefore a notarymore commonly than before
a priest."8The notary was needed to recordthe instrumentumsponsalitii, or
formalbetrothal,and a confessio dotis, or dotal agreement.19These contracts,
written in Latin by notaries using many abbreviations(and often horrible
handwriting)have left valuable informationon the size of the dowry and the
names of the prospective spouses and theirrespective fathers.20
Yet, these names are boundto remain"pureshadows"if researchershave
no access to a source sheddinglight on who these people were. Fortunately,
scholarshave been blessed with the opportunityto studythe census andproperty survey of 60,000 households known as the FlorentineCatasto of 1427.
In thatyearthe governmentof Florence,pressedby urgentfinancialneeds due
to continuous warfarewith other Italiancities, tried to increase its tax revenues. To ascertainthe wealth of the citizens living in all its domains,Florence
undertookthe Catasto-a comprehensivecensus and propertysurvey. Each
household head hadto declarethe houses, lands, and draftanimalshe owned;
the crops grown, andtypes of agrariancontractsused; the averagecrop yields
of the previous three years; his debts and credits;his shares of commercial
18 Herlihy, "Medieval MarriageMarkets,"p.
26.
19Herlihy and Klapisch-Zuber,Les Toscans.
20 In the fifteenth century,the use of surnameswas not yet widespread,being limited mainly to rich
and powerful families. Each individualwas identifiedby first name, father'sname, and grandfather's
name. Thus, for example, one finds Cristofano,son of Niccolaio of Maso.
112
Botticini
he hadto reportthe composiandhis occupation.Furthermore,
partnerships;
tion of his familyby name,age, sex, andrelationshipto himself.
To createthe sampleused in this article,I havejointly consultednotarial
deeds and the FlorentineCatastoof 1427 housed at the StateArchives of
Florence.Marriagecontractswrittenby notariesprovidedinformationon the
size of the dowry (whetherin cash or in land holdings), the names of the
bride,the groom,andtheirrespectivefathers,andtheplaceof residence.The
deedsrecordmarriagesin Cortonaandforty-fourvillages in its countryside
between 1415 and 1436. The FlorentineCatastoof 1427 suppliedinformation on thewealth,occupation,numberof children,percentageof daughters,
and age of the spouses for the householdsfound in the notarialacts. Not
includedin this sampleare dowriesgiven by charitiesto orphangirls, because these young women did not have a correspondinghouseholdin the
Catasto,anddowriesthatconsistedof landholdingswhose monetaryvalue
was not providedin the marriagecontracts.
Searchingthe Catastoforthe namesof groom'sandbride'sfathersgiven
in notarialcontractswas not always an easy task. For marriagecontracts
priorto 1427the fathermighthave diedbeforethatdate,so one wouldhave
to look for the householdunderthe nameof the survivingwife or children.
Moreover,householdswere often formedby two brothersliving with their
familiesunderone roof, andthe householdin the Catastomightturnout to
be recordedunder the name of the other brother.Common names like
Agnolo, Antonio, or Giovanni-like theirmodem Americancounterparts
John SmithandMaryBrown-also turnedout to be a formidableobstacle
to the exact identificationof the headof each household.
Increatingthe sample,a conservativestrategyhasbeenadopted:onlythose
householdswereincludedforwhichtherewas no doubtaboutidentity.Thus,
after checking in the Catastoto find the correspondinggroom and bride
households,224 outof 328 marriagecontractswerefoundforwhichall informationon age, wealth,occupation,landholding,andhouseholdstructureis
the size of dowTable4 showsthatthe sampleis representative:
complete.21
ries in the sampleandin all 328 contractsis fairlyclose. If anything,the bias
indicatesthatthe sampletendsto representwealthierhouseholds.
THEESTIMATION
In the ordinaryleast squaresestimationpresentedin Table5, the dependent variableis given by the value of the dowry (in florins).22This number
21 These 328 notarial deeds are a complete enumeration of all surviving marriage contracts for
Cortonaandits countrysidebetween 1415 and 1436 for which the value of the dowrywas writtenin
the contract.
22 Onemightcorrectlypointoutthatsincethe dowrycannotbe negative,a moreappropriate
estimation would be a Tobitestimation.I have also runa Tobitestimationandthe resultswere identical.
113
A Loveless Economy?
TABLE4
SIZE OF DOWRIES IN CORTONA, 1415-1436
Sample
All contracts
Number of
Contracts
Mean
(florins)
224
328
125.5
110.6
Median
(florins)
70
60
Standard
Deviation
105.9
99.3
Sources: ASF, Notarile Antecosimiano 1143, 1144, 1145, 1146, 5441, 10038, 18905, 18906, 18907,
18908, 18909, 18910, 18911, 18912, 18913, and 18914.
representsthe sum of cash plus the monetaryvaluation of houses and land
holdings, if the dowry consisted of these assets.23The regression takes into
account the ages of the couple and whether the bride, the groom, or both
were marryingfor the second time.24
To capture occurrences of women marrying down, the regression uses
dummyvariables.Thereareeight dummyvariablescorrespondingto the eight
possible marital combinations described earlier and in Tables 2 and 3.
AgrAgrDownandAgrAgrUprefer,respectively,to those marriagesin which
an agriculturalbride(Agr)marriesan agriculturalgroom (Agr) with a smaller
(Down), or larger(Up), amountof wealth. In both cases, the bride is matched
with an "equal"groom from the viewpoint of occupation,but marriesdown
(or up) according to wealth. The dummy vanrablesNonagrNonagrDown
and NonagrNonagrUp are similarly defined for nonagriculturalbrides.
NonagrDownDownconsidersthose marriagesinwhich a nonagriculturalbride
marries an agriculturalgroom (Down) with a smaller amount of wealth
(Down). This bride is marryingdown both from the viewpoint of occupation
and wealth. NonagrDownUp refers to marriagesin which a nonagricultural
bridemarriesan agriculturalgroom (Down) of greaterwealth (Up). The variablesAgrUpDownandAgr UpUp are siniilarlydefinedfor agriculturalbrides.
Similarto common methods of testing the altruismhypothesis, the regression controls for households' wealth. The wealth assessments aretakenfrom
the census declarations. One may raise two objections to this data. While
occupationsremainedfairly stable over years, the wealth of a household was
more subjectto change. The question is whetherthe wealth a household had
in 1427 was the same as, or at least close to, its wealth ten years earlier,or
five years later for marriagesnot contractedin 1427. Moreover, for households whose children marriedbefore 1427, the 1427 wealth assessment of
the groom household would be overestimated and the wealth of the bride
household would be underreported,given that the amount of the dowry had
been transferredfrom the bride household to the groom household before
23 Ifthe dowrywas partlyprovidedby the parentsandpartlyby a charity,I have disregardedthe latter
component,since for the hypothesesto be tested here, the dowry providedby the parentsis the relevant
variable.
24 Out of 224 marriages,in 11marriagesthe groom was marryingfor the second time, in 4 marriages
the bride was marryingfor the second time, and in 5 marriagesboth spouses were remarrying.
114
Botticini
thatyear.The latterproblemhas been correctedby addingthe amountof the
dowry to the wealth of the bride household and subtractingit from the
wealthof the groomhousehold.The formerproblemprobablyhas no solution for this sample.However,to reducethepossiblebias of havingonly the
1427 wealth assessment,a year dummyhas been introducedin the regression:for marriagesafter1427this dummyvariabletakesthe value of one.25
The total numberof childrenincludesall boys and girls, independentof
theirrelationshipwith the head of the household.In otherwords, a household headhaving,say, two daughters,one son, one unmarriedbrother,two
young unmarriedsisters, and one unmarriedniece is consideredto have
seven children-of whichfive areconsidereddaughtersandtwo areconsideredsons. The rationalefor this computationis thatthe householdheadhad
to providedowriesto his femaledependentsregardlessof whethertheywere
his daughters,sisters,or nieces.26
TESTINGTHE PRESENTNET VALUEHYPOTHESIS
Thepresentnet valuehypothesispredictsa positiverelationshipbetween
the size of the dowryandthe age of the bride.The basic assumptionis that
the dowryis dependentuponthe value of the bride'snet contributionto the
maritalhousehold,andtheyoungerthebride,thegreaterherexpectedcontribution.Among otherthings, a younger brideis more attractiveand more
fertile.It follows thatthe youngerthe bride,the smallerthe dowryher parents have to pay in the marriagemarketbecausethe groomhouseholdwill
be morewilling to accepta young bride.
The estimationseems to supportthis hypothesis.The coefficient of the
bride's age has the predictedpositive sign and is statisticallysignificant
when the regression also considers the age differencebetween the two
spouses (Table5, first column). This coefficient is no longer statistically
significantif the regressionallows for the groom's age (Table 5, second
column). Because the bride's and groom's ages are correlated(0.47), the
formerspecification,which considersthe bride'sage andthe age difference
of the two spouses, seems moreappropriate.27
Moreover,Tuscanparentsseem to have providedlargerdowries when
their daughtersmarriedfor the second time. Again, this finding seems to
supportthepresentnetvaluehypothesis:becausebrideswho remarriedwere
25 The correlation coefficient betweenthe year dummy andthe wealth ofthe bride (groom) household
is 0.04 (-0.01).
26
The number of children comes from the census, which in this case may lead to a bias: daughtersthat
have alreadymarriedand left theirpatemnal
householdwould beundercounted. To account for this possibility, I have included the age ofthe household head as proxy for the numberof childrenin a given household.
None of these variables turned out to be significant nor have they affected the results of the regression.
27 The correlationcoefficient between the bride's age and the age difference ofthe two spouses is - 0.10.
A Loveless Economy?
115
TABLE 5
ESTIMATES OF THE DOWRY FUNCTION (OLS)
DEPENDENT VARIABLE: DOWRY (IN GOLD FLORINS)
1
Coefficient
CONSTANT
YEAR DUMMvY
GROOMREMARRIED DUMMY
BRIDE REMARRIED DUMMY
18.51
15.46
-14.33
86.75
2
t-Statistic
(0.32)
(1.46)
(-0.71)
(3.28)***
Coefficient
5.87
15.67
-15.14
86.52
GROOM'SAGE
4.01
(GROOM'SAGE)2
-0.03
-0.69
5.57
-0.16
-0.05
AGE OF THE GROOM'SFATHER
BRIDE'SAGE
(BRIDE'SAGE)2
AGE OF THE BRIDE 'S FATHER
A GE DIFFERENCE
(AGE DIFFERENCE)2
GROOMHOUSEHOLD'S WEALTH
BRIDE HOUSEHOLDS WEEALTH
AGRAGRUP DUMMY
AGRUPDOWN DUMMY
AGRUPUP DUMMY
NONAGRNONAGRDOWNDUMAMY
NONAGRNONAGRUP DUMMY
NONAGRDOWNDOWNDUMMY
NONAGRDOWNUP DUMaY
NUMBER OF CHILDRENIN GROOM
HOUSEHOLD
PERCENTAGE OF DAUGHTERS IN
-0.68
8.59
-0.19
-0.09
3.29
-0.06
0.01
0.01
-0.53
33.3
98.25
109.43
75.89
98.3
63.21
-4.5
45.32
(-1.25)
(2.33)**
(-2.45)**
(-0.2)
(2.02)**
(-1.46)
(4.72)***
(6.14)***
(-0.04)
(1.51)
(5.26)***
(6.12)***
(3.26)***
(4.9)***
(3.69)***
(-1.27)
(1.4)
t-Statistic
(0.09)
(1.48)
(-0.73)
(3.26)***
(1.3)
(-0.87)
(-1.27)
(1.39)
(- 1.93)*
(-0.12)
-
0.01
0.02
-0.14
33.1
95.7
109.59
76.25
98.51
63.78
-4.35
(4.71)***
(6.19)***
(-0.01)
(1.5)
(5.14)***
(6.1)***
(3.25)***
(4.89)***
(3.71)***
(-1.23)
44.48
(1.37)
-8.43
(-2.95)***
GROOMHO USEHOLD
NUMBER OF CHILDREN IN BRIDE
-8.28
(-2.91)***
HOUSEHOLD
PERCENTAGE OF DA UGHTERS IN
-11.05
(-0.59)
-11.72
(-0.62)
BRIDE HOUSEHOLD
RESIDENCE DUMMY
R2
ADJUSTED R2
F-STATISTIC
N
26.36
0.68
0.65
19.18
(2.19)**
25.55
0.68
0.65
18.99
224
(2.11)**
* = Significant at the 10 percent level.
** = Significant at the 5 percent level.
Significant at the 1 percent level.
Notes: YEARDUMMYequals one for marriages after 1427. AGEDIFFERENCE is groom's age minus
bride's age. The omitted dummy is AGRA GRDOWN. AGRUPDO WN equals one when a bride from
a peasant household (AGR) marries a nonpeasant groom (UP) with lower wealth (DOWN).
Analogously, AGRUPUP equals one when a bride from a peasant household (AGR) marries a
nonpeasant groom (UP) with higher wealth (UP). NONAGRDOWNDOWN (NONAGRDOWNUP) are
similarly defined for nonpeasant brides marrying peasant grooms. Lastly, NONAGRNONAGRDOWN
(NONAGRNONAGRUP) are similarly defimed for nonpeasant brides marrying nonpeasant grooms.
RESIDENCE equals one if the bride and/or the groom households lived in the town of Cortona, zero
if both households lived in the countryside.
Source: See Table 1.
116
Botticini
likely to be olderthanthose marryingfor the first time, parentshad to pay
largerdowriesto find groomswilling to marrythem.
TESTINGTHE ALTRUISMMODEL
Tables2 and3 haveindicatedthatwomenwhomarrieddownwerecommon
bothin the agricultural
andin the nonagricultural
group.The questionis, did
manyingdown influencethe size of dowries?Theregressionin Table5 indicatesthatthiswas indeedthecase.Householdswhosedaughtersmarrieddown
seemto haveprovidedlargerdowriesto theirdaughters:
the coefficientsof the
"down"dummiesaregreaterthanthe coefficientsof the "up"dummies.28
Given thatdaughterswere potentiallya financialdrainon theirfamilies,
that any marriagewas betterthan their continuingto stay in the parents'
house, andthata dowrywas given with eachmarriage,even if it was down,
one mightobjectthatthis by itself would be evidenceof self-interestrather
than of altruism.True, a dowry went with each marriage,regardlessof
whetherthe marriagewas up or down. However,parents'altruismtoward
their daughtersis not determinedby the existenceof the dowry,but by its
size. In medievalandRenaissanceTuscany,parentsdid not have a choice of
whether or not to provide a dowry for their daughters. They could, and did,
decide on the size of the dowry. The data for Cortona in the 1420s reveal
that parents cared about their daughters' well-being by providing them with
larger dowries when marrying down.
Of course, the provision of the dowry was also affected by factors besides
paternallove. As mentioned earlier,the fascinating feature of the dowry is its
double nature.On one hand,the dowry was a marketprice, and from this point
of view, brides' parents had to face the reality of a marriagemarket in which
older brides were not highly valued. Therefore,if their daughtermarriedlate,
parentshad to provide a largerdowry to convince the groom to accept an older
bride. This outcome is what the present net value hypothesis predicts. On the
other hand, the dowry was also a transfer of wealth from parents to their
daughters, and it is in this context that parents could show their altruism, if
any, toward their daughters.In the Tuscan case, parents seem to have acted
altruisticallyby providing larger dowries for daughtersmarrying down.
Anotherway oftesting the altruismmodel consists of checkingwhetherthere
was a fairdivision of the households' wealth among children,and, in particular,
among daughters.Table 5 indicatesthatparentswere concernedabout all their
children:the size of the dowry provided for one daughterdecreases with an
28The only exception is with the AgrUpUp (AgrUpDown) dummies referred to agricultural brides
marrying a wealthier (poorer) merchant groom (though the Agr UpDown coefficient is not statistically
significant). It might have been the case that, when an agricultural bride married a wealthier merchant
groom, the significant improvement both in wealth and status required a very large dowry. A rich
merchant groom would have agreed to marry a poorer peasant bride only if the dowry was very large.
117
A Loveless Economy?
TABLE6
DOWRIES TO SISTERS OF THE SAME HOUSEHOLD MARRYING BROTHERS OF
ANOTHER HOUSEHOLD
Groom Household's
Wealth(florins)
Bride
Household
Sister
Dowry
(florins)
Bride Household's
Wealth(florins)
1
Bride 2
55
243
8
~~Bride
Bride 2
275
275
550
454
Bride 1
45
208
67
Bride 1
225
674
1,167
2
3
4
Bride 1
225
....
.......
.... ...
Source: See Table 1.
increasein the totalnumberof children.More convincingly,when considering
24 marriagesof bridesbelongingto 12 householdsin which two sistersmarry
two brothersbelongingto the samehousehold,the dowriesprovidedforthe two
brideswere identical.Table 6 shows eight of these 24 marriages.
As for sistersmarying groomsbelonging to differenthouseholds,the case
of Giovanni,son of Tommaso,the wealthiestcitizen in Cortona,may be cited.
In 1421, Giovanni gave a dowry of 420 florinsto his daughterCaterina,who
was going to marry the legum doctor Mariotto, son of another Giovanni,
whose wealth was valued at 270 florins.29In 1425, the fathergave a dowry of
300 florins to his other daughterIsabetta,who was going to marryLazzaro,
son of Francesco,a seller of spices whose wealth was 1,547 florins.30
Regardless of whether parentswere altruistic or not, one might wonder
whether a large dowry would in fact have increasedthe welfare of the bride
duringmarriage.To addressthis issue, three observations are offered. First,
for marriages before 1427, the fertility rate of the bride can be measured
from data detailing the number of children the bride and the groom produced.3"The surprisingfinding is thattherewas a positive correlation(0.24)
between the size of the dowry and the numberof childrena bride had. Even
when controlling for wealth, the fertility of the bride seems to have been
affected by the size of the dowry she broughtto the maritalhousehold at the
time of her marriage.Of course, fertility is affected by many factors. Brides
receiving large dowries probablytended to be better fed and better treated
in the maritalhome thanthose bringing smallerdowries, and thus were able
to producemore children.Literaryandanecdotalevidence also suggests that
29 ASF, Notarile Antecosimiano
30 ASF, Notarile Antecosimiano
18910, c. 172r.
18911, c. 187v.
31 Of course, for marriages after 1427 the number of these children is zero simply because the bride
and groom were still single at the time of the Catasto of 1427.
118
Botticini
a bridehavinga largedowrywas morelikelyto be treatedbetterin the marital household.32
A second reason for thinkingthat largerdowries did improve brides'
welfareis thatbrides'parentssometimespaiddowriesin installments.This
had two advantages.It ensuredthat,hadthe groompredeceasedthe bride,
her parentswould easily recoverthe dowry from the groom's household.
Also, by payingthe dowryin installments,the bride'sparentscould control
the groom'sbehaviorto a certainextent.Separationanddivorce,thoughnot
very common,were indeeda possibilityin medievalandRenaissanceTuscany (the Catasto of 1427 reportsseparatedand divorcedwomen in Florence).Althoughthehusband,orhis father,controlledthe dailymanagement
of the dowry,the brideandherfamilyownedthe dowryandhada say in the
way the dowrywas managed.
A final factorconnectinglargerdowriesto bridalwelfare is that if the
groom died, the dowry returnedto the bride.In that case, it was she who
could decidewhetherto go backandlive in herpaternalhousehold(or with
herbrothers,hadthe parentsdied)or on herown. A largedowryguaranteed
thata widow wouldmaintaina decentstandardof living afterherhusband's
death, without having to rely on her brothers'hospitality.Thus, parents
providinglargedowrieshelpedtheirdaughtersto sustaina good standardof
living even in widowhood.
CONCLUDINGREMARKS
By using a sampleof 224 marriagecontractsfor a Tuscantown between
1415 and 1436, this articleexaminesthe dowrysystemfromthe standpoint
of the dowryas a marriagepaymentandas anintergenerational
transfer.The
presentnet value hypothesisis proposedto addressthe former.The model
predictsthat,ceterisparibus, the largerthe contributionof the brideto the
maritalhousehold,the smallerthe dowryherparentshaveto pay.The altruism model is presentedto study the dowry in terms of intergenerational
transfers.The altruisticbehaviorof parentsis testedby examiningwhether
largerdowrieswereprovidedwhendaughtersmarrieddown.Bothhypotheses find supportin the Tuscandata.
Thenext stepwill be to takeadvantageof notarialrecordsheldatthe State
Archives of Florenceto test the models of the dowry in Florenceitself as
outlinedabove. The Florentinemarriagemarketconstitutesa particularly
attractivecase studybecauseof the huge samplesize availableandthe great
varietyof occupationsandpoliticalstandings,which characterizedFlorentine citizens with respectto people living in the Florentineterritories.Anthony Molho shed much light on the marriagestrategiesof the rulingclass
32 Klapisch-Zuber, Women.
A Loveless Economy?
119
of Florence in the second half of the fifteenth centuryand the first half of the
sixteenthcenturythroughhis impressivestudyofthe FlorentineDowry Fund.
Yet, we still lack importantinformationon the marriagestrategiesand dowry
formationof middle-andlower-classhouseholds.How largewere the dowries
of carters,smiths,wine dealers,innkeepers,tanners,beltmakers,key makers,
carpenters,wool carders,dyers, peddlers, coppersmiths,weavers, shearers,
doublet makers, cutlers, spinners,and other urbanworkers?What were the
marriagestrategiesandwhatwas the degreeof altruismof householdsthatdid
not invest in the Dowry Fund but formed the vast majorityof the Florentine
population?In this area,much more work remainsto be done.
APPENDJxTABLE 1
SEX RATIOS (MALES/FEMALES) IN TUSCAN TOWNS, 1427
Florence
Pisa
Pistoia
Arezzo
Prato
Volterra
Cortona
Montepulciano
Colle
San Gimignano
Castiglion Fiorentino
Pescia
Town
Countryside
117.6
112.3
102.7
108.1
105.7
108.3
97.3
110.7
117.8
113.3
104.2
112.5
108.9
106.9
112.7
106.2
115.5
118.3
108.9
121.5
114.1
104.7
Source: Data from Herlihy and Klapisch-Zuber,Les Toscans,p. 472.
REFERENCES
Florence, State Archives (ASF), Catasto; Notarile Antecosimiano.
Becker, Gary S. A Treatiseon the Family. Cambridge,MA: HarvardUniversityPress, 1991.
Boserup, Esther. Women'sRole in Economic Development.London:Allen & Unwin, 1970.
Brown, JudithC., and JordanGoodman."Womenand Industryin Florence."this JOURNAL
40, no. 1 (1980): 73-80.
Chabot,Isabelle. "Lareconnaissancedu travaildes femmes dans la Florence du bas Moyen
Age: Contexte ideologique et realite." In La donna nell'economia secc. XIII-XVIII,
edited by Simona Cavaciocchi, 563-76. Florence: Sansoni, 1990.
Chojnacki, Stanley. "Dowries and Kinsmen in Early Renaissance Venice." Jouirnal of
InterdisciplinaryHistory 4, no. 5 (1975): 571-600.
."Marriage Legislation andPatricianSociety in Fifteenth-CenturyVenice." In Law,
Custom, and the Social Fabric in Medieval Europe: Essays in Honor of Bryce Lyon,
edited by BernardS. BachrachandDavid Nicholas, 163-84. Kalamazoo, MI:Medieval
Institute Publications, WesternMichigan University, 1990.
."The Most Serious Duty": Motherhood,Gender,and PatricianCulturein Renaissance Venice." In Refiguiring Woman:Perspectives on Gender and Italian Renaissance, edited by Marilyn Migiel and Juliana Schiesari, 133-54. Ithaca, NY: Cornell
University Press, 1991.
120
Botticini
Cigno, Alessandro.Economicsof the Family. Oxford:OxfordUniversityPress, 1991.
Cohn, SamuelK. Womenin the Streets:Essays on Sex and Power in RenaissanceItaly.
Baltimore:JohnsHopkinsUniversityPress, 1996.
Cole, Hal L., GeorgeJ. Mailath,andAndrewPostlewaite."SocialNorms,SavingsBehavior, and Growth."JournalofPolitical Economy106, no. 6 (1992): 1092-125.
Goody,Jack."BridewealthandDowry in AfricaandEurasia."InBridewealthandDowry,
editedby JackGoody and StanleyJ. Tambiah,1-58. Cambridge:CambridgeUniversity Press, 1973.
Gregory,Heather."Daughters,Dowries, and the Family in Fifteenth-CenturyFlorence."
Rinascimento,2d ser.,27, no. 1 (1987): 215-37.
Shoshana.Onthe EconomicsofMarriage:A TheoryofMarriage,
Grossbard-Shechtman,
Labor,and Divorce.Boulder,CO:WestviewPress, 1993.
Hajnal,J. "EuropeanMarriagePatternsin Perspective."In Populationin History:Essays
in HistoricalDemography,editedby David V. Glass andDavid E. Eversley, 101-43.
Chicago:Aldine PublishingCo., 1965.
Henderson,John.PietyandCharityinLateMedievalFlorence.Oxford:Clarendon
Press,1994.
Herlihy,David. "TheMedievalMarriageMarket."Medievaland RenaissanceStudies6,
no. 1 (1976): 3-27.
Les Toscanset leursfamilles. Un etudedu
Herlihy,David, andChristianeKlapisch-Zuber.
Catastoflorentinde 1427. Paris:Pressesde la FondationNationalede Sciences Politiques, 1978.
Hughes,DianeOwen."UrbanGrowthandFamilyStructurein MedievalGenoa."Past and
Present 66, no. 1 (1975): 3-28.
. "FromBridepriceto Dowryin Mediterranean
Europe."In TheMarriageBargain,
editedby MarionKaplan,113-58. New York:HaworthPress, 1985.
Africa:FemaleProductivity
Jacoby,Hanan."TheEconomicsof Polygynyin Sub-Saharian
and the Demandfor Wives in Cote d'Ivoire."Journalof Political Economy103, no.
5 (1995): 938-71.
Kirshner,Julius."Wives' ClaimsAgainstInsolventHusbandsin LateMedievalItaly."In
Womenof the Medieval World:Essays in Honor of John H. Mundy,editedby Julius
Kirshnerand SuzanneF. Wemple,256-303. OxfordandNew York:Basil Blackwell,
1985.
. "Materialsfor a GildedCage:Non-DotalAssets in Florence(1300-1500)." In The
Family in Italyfrom Antiquityto thePresent,editedby RichardP. SallerandDavid I.
Kertzer,184-207. New HavenandLondon:Yale UniversityPress, 1991.
Kirshner,Julius,andAnthonyMolho."TheDowryFundandthe MarriageMarketin Early
QuattrocentoFlorence."Jouirnalof ModernHistory50, no. 3 (1978): 403-38.
Klapisch-Zuber,Christiane."The 'CruelMother':Maternity,Widowhood,andDowry in
Florencein the FourteenthandFifteenthCenturies."AnnalesE.S.C. 38, no.5 (1983):
1097-1109.
. Women,Familyand Ritualin RenaissanceItaly. Chicago:Universityof Chicago
Press, 1985.
Kuehn,Thomas."CumConsensuMundualdi:Legal Guardianshipof Womenin Quattrocento Florence."Viator13, no. 2 (1980): 309-33.
. Law,Family, and Women:Towarda Legal Anthropologyof RenaissanceItaly.
Chicago:Universityof ChicagoPress, 1991.
Mancini,Girolamo.CortonanelMedioEvo.Roma:MultigraficaEditrice,1969 (originally
published1897).
Molho,Anthony.MarriageAlliance in LateMedievalFlorence.Cambridge,MA:Harvard
UniversityPress, 1994.
Rao, Vijayendra."TheRisingPriceof Husbands:A HedonicAnalysis of Dowry Increases
A Loveless Economy?
121
in Rural India."Journal of Political Economy 101, no. 4 (1993): 666-77.
Riemer, Eleanor S. "Women, Dowries, and Capital Investment in Thirteenth-Century
Siena." In The Marriage Bargain, edited by Marion Kaplan, 59-80. New York:
Haworth Press, 1985.
Trexler, Richard C. "Le celibat a la fin du moyen-age: Les religieuses de Florence." Annales E.S.C. 27, no. 6 (1972): 1329-50.
Uccelli, Paolo. Storia di Cortona. Arezzo: Bellotti Editore, 1935.