sex and rationality

ly c H A T E R v E
re
X-
:k
et
re
:d
ve
en
of
ly
pt
ns
lscv
ual
tt's
ns.
o-
,rS
,ea
lr-
Sex and Rationality
Tnrs cneprrn and the next present a positive economic theory of sexual
behavior,s howingh ow the rype and frequencyo f differents exualp racdces,
as distinctf rom the sex drive and sexualp reference(i nclination,o rientation),
can be interpreteda s rationalr esponsetso opportunitiesa nd constraintsT. he
model of rational sexual behaviore mphasizetsh e varied goals of sex' the
costso f sexuals earcha, nd the natureo f marriage-whetherc ompanionateo r
noncompanionate-asc ritical determinantso f sexualp racticesa nd amirudes,
along with urbanizationa nd, abovea ll, the occupationaal nd financials ituation
of women.C hapter7 proposesa normativee conomict heory from which
the pattern of regulation that would be optimal if sex were a morally indifferent
subiect can be deduced.
The Benefits of Sex
Let us begin by consideringth e endst hat sex servesa nd then the meanso f
servingt hosee nds,t hat is, the practicest hemselvesT.h e endsf all into three
groups,w hich I shall call procreatiue,h edonistic,a nd sociable.T he first is
obvious. The second has two cells. One is relief from the urgency of sexual
desire;t he analogyi s to scratchinga n itch, or to drinking water when one is
thirsty. The other is arse rotica,t he deliberatec ultivationo f the faculty of
r.*url pl..rute; the analogy is to cultivating a taste for fine music or fine
wine,
The third group of sexuale nds,t he sociablei,s the leasto bvious.I t refers
to the use of sex to constructo r reinforcer elationshipsw ith other people'
sucha ss pouseso r friends.T he RomanC atholicc oncepto f the "marital debt"
(the obligation of each spouse to submit to the other's request for sexual
intercourie)f its nicely here.T he relationshipsc an be explicitly commercial;
thus the ends pursued by the prostitute fall into my category of sociability.
t L 2 A THEORY OF S E X U A L I T Y
Many sexualr elationshipsth at are not classifieda s prostitutionn evertheless
contain a strong commercial element: for example, marriage for money, But
an exchangen eedn ot be commerciatl o be economic.M ariiage is a relationship
of exchange that can be modeled in economic t..ms iue' if neither
spouse'sm otivesa re crasslyp ecuniary,o r indeedp ecuniarya t all. The idea
of male friendships pilling over into a (homo)sexuarle lationshipt har in rurn
cemenrsth e friendshipi s an elemento f the ancientG reekt heoryo f homosexuality;
it is visible, for example, in plato,s Symposium.
The elemento f sociabiliryi s fundamentairo an understandingo f the sexual
differencesb erweenc-o mpanionatea nd noncompanionate." riiag.. (For the
saKeo t srmpllclty,I treat companionatenesass a dichotomousv ariable;a c_
tually it is a continuouso ne.) In noncompanionatem arriage,h usbanda nd
wife are virtual strangersf rom an emotionals tandpoint," nJ th. sexuarr elationshipb
erweenth em is no more rewardingt han ihat between, ,r"rrg..r. Fo.
the husband,t herefore,p rostitutes,c oncub-inesc,a sualm istress.r,i nd.u.n
(in particular circumstancetso be explored)a dolescenbt oys may offer gooJ
substitutesfo r marital sex, as loversm ay for the wife. '0yea re led ,o priaio
that noncompanionarem arriagew ill foiter extramaritals ex on the part of
both spousesc, oupledw ith strinuouse fforts by husbands-motivatede ither
by instinctuali ealousyo r by consciousc oncernl esrt he children, t ry .upfo.
not be theiro wn-to preventw ivesf rom engagingin extramaritals ex.M eans
to this.e nd includep hysicals equesrratio"nl i,i"!r, disparagemenotf female
sexualiry, and the mutilation of the female sex organs in order to diminish
women'sc apacityf or, and hencei ncentivet o seeki,e xual pleasure.r
_ In companionatem arriage,m arital sex is investedw ith affectivee lements,
thereby creating "socioemotionalc losenessa nd exchange,r,a ther than "psychophysiological iusi
pleasurea nd relief."2T hesee lemenish avet he effecio n
many men.ofm aking extramaritals exe ithera n inferior substirutefo r maritar
sex (or at leastn o longer a superioro ne) or a more costlyo ne, because the
wife. in a companionate marriige is more likely to be jeaious and therefore
the husbandm ust devoter esourcesto .on.."ring his ph anderingsE. ven ii
some extramarital sexual relationships_for example, Leeping " ,iirtr.rr, ",
1. Lee Rainwater," Marital Sexualityi n Four Cultureso f povcrry," 26 Jounal of Maniage aad the Family 457 (1954),i s an excmplarys tudyo f the operationo f noncompanionatc ."rri"g'..
f]so.
verl much to the poinr is Fatima M...,irrl, B"yorri the ieil, Uate_Female Modera,Muslim Dynarnicsi , a Society1 1975)A. nd cf. David iohrn, Lo*, Sexualirya nd SocietyT: be Enforce- ment of Moralsi n classicaAl thens1 59-770( 1991).C litoridecto-y, by the *"i i, noi -.r.ty an African tribal practicc;v ictorian doctors resortedt o it (and somctlmes of cven to the removar thc ovaries)a sa lasr-ditchc uref or fcmarcm asturbations. us"n s. M. E dwards, and the Females exuality Law: A Study of Colstructs of FemaleS exualityo s iiey tnfo_, i**r" ;i ["g'it Procedure8 7-90 {1981).
.
2. Lec,Rarnwater",S omeA specrso f Lowcr-ClassS exualB ehavior,.i n rexuat Stydiesin Human Eehautor:t he AmericanS cene1 77,192_193( AilonS hilohc d. 1970).
t
t
T
I
I
s
i
a
r
tl
b
n
b
ls
d
tI
id
n(
e)1
Pr
su
af.
co
riz
m
co
fer
Sht
Pra
ly
al
ty
al
Sex and RationalitY 1 1 3
distinctf rom PatronizingP rostitutes-havet he samea ffectivev alue as marital
;;;;;.t "*'.ostly in iim. "nd other resourcesq' uite apart from costs.of
..*."f..t,, comparcdt o the now inferior alternativeo f a casual iaisonw ith
'ai' iDl;r;o;;slyti ruteo r the equivalent.
i' p"rting anotherp ossiblem otive.for sex' one stressedin
,1,. i.-irir, literatur., thiiterature on homosexualirya mong primates, and
,it. fi,.i"*t. on prison homosexualitya, 1d thal is the assertion of dominance
;;;;;;;;r'over the penetrated'T3h is is frequentlva dduced as a motive
;j'l;,:';;;;l'"Jl.iot.*ual sodomv, for rape' and in some versions of
."di."lf.mini'tthoughtevenforostensiblyconsensualvaginalintercourse.
Th. i.r,r. of dominanceis examinedin later chaptersB' ut mention of primate
,o i, !..,irr.nt herc in suggestinga possibleg enetic,basis for the sociability
dimension of human ,.*uiilty' SJ* ttttnts the bond berween mates and by
;;i;; t" helps ensure that the male will stay around to protect the female
;;;'i;ildt;nj rtoln hunger and predation' But we must be caretul' In one
t.t"..i,f,. ioni i, clor., ii the female enioys scx than if she is indifferent;
;J;l; ;;;;t*"ti.v of affection'I n anotherr especth' owever' it is-w eaker'
;;;;;; ;".. a*ak.n"d to the possibilityo f sexualp leasure,.the female is
-".. ttt*pirile to the ad"antts of other males' Knowing this' he-r mate
;;;;;;pt.ious, jealous, watchful, constraining; and mutuality of affection
i, una.r-in.a. Considerationssu cha s theseh avel ed to questioning
llttht:
the femaleo rgasm' and more broadly the continuouss exuala vallaDlllry or
t.h.'ei; ;h;;u.m* an female' are acnrally ada,ptive'a
mat;h the varieiieso i possibles exualp racticet o.the endsi ust
ia.Jft.a "t being served by sex' It should be apparent that the means are
noi i.,,...t "ngeab"les; ome" te btae' suitedt o one end than to another' For
example,n oncontraceptlvev aginal intercolrse.is well suited to the end of
;;;:;:;;i";, but, unlessa larg"e number of children are desired' not so well
suitedt o the pursuit ot sexuaip leasure:a n unwantedc hild is a heavy tax to
;;;;; ,h; pi."rrrr., of sex' Noncontracepdve vaginali ntercoursem av also'
and for the same reason, undermine sex as the glue of a marital relationship;
."i t"i "*. reproductionm ay, indeed'b e essential to companionatem a.r-
,lgtJ'r.t ,rt.l. to *ho* se* i' meitly a mafter of scratching an itch'
."'r*rU"rion shouldb e an excellenst ubstitutefo r intercourse becauseit is so
much cheaperin termso i ti-t, t*poture to.disease' and other dimensionso f
cost. Homoscxu"t int."ott"t i' *o'thlett for producing children and is inferior
from a pleasure, t"ndpoint for personsw ho prefer heterosexual sex'
3. Scc,f or an cxamplcA, lanJ Davis, 'sexualA ssaults^tihne P hiladclphiPar isonS ystcma nd
Sf,Jfit V"n.," id. at ilo. Cf iohcn, notc I abovc' at 186-187'
--'f.
b."rfa Symons,T bc Evolutiono f Human Sexuality' chs' 3-4 (1979)'
5. Angus McLarcn "na a'ittl iig"t Mclarcn' Th.e Bedroom and the State: The Cbanging
Practicesa nd Potiticso f contr))"piii" o"a Abortion in Canada' 1880-1980 26-27 11986)'
t 1 4 A THEORY OF S E X U A L I T Y
but even for them it may score high on cementing relationships and do okay.
as a method fo, scratci,ing the Jext'al itch' lt may even serve the end of
;il;;;li;;s of the Jpposit. sex are unavailabtei;t s suitabiliryi n this
iegard will vary dependingo n one'sp laceo n the Kinseys cale'
-Th"
.ol. of substitutiona mongs exualP racticesis illustratedb y the changes
in the behavioro f homosexuals-thahta veo ccurredi n the wake of the AIDS
.iiJ.-i.. Homosexualsh ave respondedto the increasedc ost of unshielded
anal intercoursew ith multipl. ,.* p",tnt" by reducingt he number of their
p"ti".tt, in.r."ring th. ur. of conioms, and substitutingf ellatio qa "$:t
;;;;; ih"t ".."1.r. likely than anal intercourse to spread AIDS'6 This
;;;ili; shows that event hough sexualp referenceis not chosen,t he area for
rational choice in sexual pra-tices is a large one. And substitution is not
i,*ii.J ," pra*ices but alio includes the substirution of sexual obiects (for
example, a se*ual partner of the opposite sex for one of the same sex)' of
nonr&ud fo, ,e*.,al activities,a nd of various transactionafl orms, such as
rape,c oncubinagea, nd prostitution.
"Safe sex" is not a perfect substitute for unsafe sex; if it were, it would
have beenu niversala mong homosexualsb eforet he AIDS epidemic'b ecause
unsafe sex was known tJ carry a high risk of contracting other venerea-l
diseasesS.Tin ces afes exi s not a perfects ubstituteI, hypothesizeth at the.A IDS
.piJ.-i., by making homose*ual activiry unavoidably more cosdy (whether
th.ough ih. gr.",.r-.o.t of contracting AIDS or the lesser cost of switching
io r"fZ to),
"has
,edrrced the amount of that activity' One might exPed the
.if"o ,. be especially dramatic among those men whom I termed in the
precedingc h"pi., "opportunistic" homosexualsS. incet he benefitso f homo-
,.*u"l ".ti"ity ".. l.ti to them than to "real" homosexualst,h ey should
respondm ore to an increasein the costso f the activiry'B ut that is provided
5. HcathcrG . Miller ct al., AIDSr The SecondD ecade8 2-83 (1990);M arshallH ' Becker
"nJ;ltt C. Joseph,
'AIDS and BchavioraCl hanget o RcduceR isk: A Rcvicw'"7 8 Ametican
Jounal of bubiic Health 394 (1988); Ron D' Stall, Thomas.J' Coates' and Colleen Hoff'
lBeh""ioral Risk Reduction for HIV lnfection among Gay and Bisexual Men'" 43 Ametican
Psychologis8t7 S (1988);N ational RescarchC ouncil,C ommineeo n AIDS Researcha nd thc
Seiatior;I, Social,a nd StatisticalS ciencesA, IDS; SexualB ebsuiora td IntravenovsD rug Use
132, 134-i35 (CharlcsF . Turner,H eatherG ' Miller,a nd LincolnE ' Mosese ds'1 989);D ennts
Altman, 'AIDS and thc Reconccptualizatioonf Homoscxualiry,"in Altman et al'' Homosexuality'
'Vbich
Homosexuality?In tenaiional Conferenceo n Gay and Lesbi4n Studies3 5' 44 (1989);
R..W.ConncllandSusanKippax,"SexualityintheAIDSCrisis:PancrnsofScxualPracticcand
Pleasurein a Sampleo f AustralianG ay and Biscxual Men," 27 Joumal of S-e-xR eseatch
767
(1990).P rotectivcm casurcasl c lcssc ommoni n citicsi n whicht hc incidcncoc f AIDSi s low; sec
Millcretal.at83;JohnS.Moranctal.,"lncreaseinCondomsalesFollowingAIDSEducation
and PubliciryU, nitcd Statcs,"8 0 Americsnlo umat of PubticH eahh 807 (1990)'T his is not
surprising from an economic standpoint, sinic thc cxpccted cost of unsafe sex is lowcr in those
crtres.
7. SccC haptcr 5. For direct cvidencet hat many homosexualsre gards afes ex as more costly
becauslce ssp Lasurableth anu nsa{cs cx,s eeM iller et al', note6 abovea' t 110'
Sex and Rztionality 1 1 5
the increasei n costs for them is the same.I t may be smallcr.T he riskiest
behavior with respect to AIDS is receptiue anal sex; opportunistic hcterosexuals
usually are penetrators.
I also hypothesizeth at the AIDS epidemicw ill resulti n a reductioni n the
number of illegitimateb irths. A condoms ervestw o functions:c ontraception
and diseasep revention.B ecauseA IDS can be spreadt hroughv aginala s well
as anal intercoursea, nd the dangero f infectioni s dramaticallyr cducedb y the
use of a condom, the epidemich as increasedth e beneFtso f using condoms
to heterosexualas s well as to homosexualsW. e can thereforee xpcctt he use
of condomsi n heterosexuailn tercourseto increase,asn d,a s a by-product,t hc
incidenceo f noncontraceptiveva ginali ntercourseto decline,T he number of
legitimateb irths will probablyn ot fall significantlys, incet hc vastm aiority of
married couplesi n this country alreadyu sec ontraceptiveesx ccptw hen prcgnancy
is desired. Indeed, the number of legitimate births may well rise. Sincc
marital scx is safert han nonmaritals ex,A IDS increasetsh e benefitso f marriage,
a nd marriedp ersonsa rem orel ikely to producec hildrent han unmarried
pcrsons are. AIDS may therefore cause the fraction of illegitimate births to
fall evenf astert han the number of suchb irths. By like reasoningw e should
expectt he incidenceo f AIDS to be highert he more societys ubsidizeisll egitimate
births by welfare programs such as Aid to Families with Dependent
Children, for such subsidiesr educet he bcnefts of contraceptivein tercourse
and therefore can be expected to reduce the use of condoms. All thcse predictions
assumeo. f course.t hat other relevantv ariablesa re held constant.
In relating the various sexualp ractices,v iewed as meanst o ends,t o thc
endst hat I havee numeratedI, haveb eena nalyzingt heb enefitso f thep ractices.
To havc a homosexual preference is to perceive homosexual intercourse as
conferringg reaterb enefits,in terms of pleasurea nd perhapsa lso sociability,
than heterosexuailn tercoursew ould. A personw ho wantsc hildrenw ill derive
greater benefit from noncontraceptive than from contraceptive vaginal intercourse.
T he benefitso f sex may be private or social,d ependingo n society's
attitude toward the endst hat the Dracticess erve.
Thc Costs of Scx
There are, of course,d ifferencesin the costsa sw ell ast he benefitso f various
sexualp ractices;a nd thesed ifferencesa re a particularf ocuso f my analysis'
8. For cvidcnccit has incrcascds,c cJ acquclincD arrochF orrcsta nd SushcclSa ingh' "The
Scxuala nd RcproductiveB chavioro f AmcricanV omen, 1982-1988,"2 2 Family Planning
Perspective2s0 6,213 (1990).F or othcrc xamplcso f hetcroscxualrsc's ponstco AIDS' sccG ina
Kolata," Drop in CasuaSl cxT iedt o AIDSP cril,"N aa YorkT imesM, ay 15,1 991'A 12; Michael
R. Kagay,' Fear of Aids hasA ltcrcdB chaviorP, oll Shows,"N cw York Times,l une 18, 1991'
85.
t r 6 A THEORY OF S E X U A L I T Y
For example,n oncontraceptiveva ginali ntercourseb y a fertile couplei s exp.
nrit, oino,, dependingo n ,"h.tli.t childrena red esireda nd' if not' whether
effeain" methodso f contraceptiona re availablea t reasonablec ost' If few
childrena re desiredb ut contr"ieption is uncertainw' e can exPecta' sJ dalthus
predicted,a n increasei n the age at marriagea nd a reduction (quite apart
?.o- "g.'", marriage)i n the frJquencyo f intercourse.Teh e latter effecth as
f-b"bly been offsJt in recent times by increases in the ef6cacy' safery' and
.o-forr of contraception, resulting in a net reduction in the cost of sexual
irr,.r.o,rrr., norwithitanding the reduced demand for children' and hence in
an increasein the amounto iint.r.ou,,o-a predictiont hat the datac onfirm'10
iik.rrrir., we can expect that sterilization is more-likely to be the contracePtive
-.tioa of .hoi.. the older a person is and the inore children he or she has'
becauseth esec haracteristicrsa iser he cost (or reducet he bene6ts)o f having
morc children and hencer educet he costso f permanenti nfertiliry.T his pred-
iction is also confirmed.lr
,1, tfr.r. examples suggest' the balance of private costs and private beneits
determines the ielative lrequency of different sexual practices' Since that
bal"n.. is bound to differ among societiesa, nd within societiesa t different
;;ri.dt, the relative frequency of th. difftttnt. sexual pracdces will
1a1
iigr,ific"ntly acrosss ocieriesa nd periods.I nstitutionala rrangementss'u ch as
;;ti;h;.;1, will affect the balance of private costs and private bene6ts and
i.tt.. tfr. fi.q,r.n.y of different practices' but for now I abstract from punishment
and even social censure.
The tendencyt o think of sex in termso f biologicalo r psychologicacl om'
puilior, ."itt.t'than of rational choicem akesi t imponant to emphasizeth e
irUttit"t"Uifiry of sexualp racticesa s meanst owardst he variouse ndso f sex'
Here are more examPles.
1. Masrurbationi s more common among middle-clastsh an among lowerclass
youths (see Chapter 2, and also note 70 in the present chapter) because
the averagea geo f firit sexuali ntercourseis higherf or the former'
2. Petting io orgrr* and early marriage are more common in religious
societiesth it discouragep remaritali ntercourseth an in permissives ocietiesin
9. For cmpiricals uPports' ecP aulA . Davida nd WarrcnC ' Sanderson' 'RudimentaryC ontraccptivcM
ethodsa Jthc AmericanT ransitiont o Marital F€rtilityc ontrol, 1855-1915," in
Long-Term Factofs in Atleftcun Ecottomic Grouth 307,335-338 (Stanley L Engerman and
Roben E. Gallman cds. 1985).
10. CharlcsF ' Wcstoff," Coital Frcqucncya nd Contrace ptton'"6 Ftmily Pla"ningP erspectiues
135 (Summer1 974);W cstoffa nd Norman B' Ryder,T be Cottraceptiue.Revolutio6n7 -70^11977):
JamesT russella nd lvesto{f, "ContraceptivcP racticca nd Trcnds in Coital Frequency'"l l famtty
'Planning
Perspectiue2s4 5 (Scptcmbcr/bctobc1r 980)' Cf' Ronald DcmosL cc' "Targct Fertility'
Contr"iption, and AggrcgatcR atcs:T owarda FormalS ynthcsis'7" 4 Demography 455' 455-
466 11977).
11. Westofaf ndR ydcrn, otc1 0 abovca, t 117-1181' 28'
Sex and Rationalin t t 7
which such intercourseis common.12P eoplei n the permissives ocietiesm ay
evenc onsider" heary petting" immoral and themselvetsh e moral exemplars.
They probably wonder at teenageg irls who submit to anal intercoursei n
order to preservet hcir virginiry,a practicen ot unknown in this country and
particularly common in South America (see Chapter 5).
3. One of the better-documentefdin dings in the empiricall iterature on
homosexualityi s that the proportion of male homosexualsw ho marry is
highert he more intolcrantt he societyi s of homosexuals.1ln3t olerancem akes
the practiceo f homosexualitym ore costly,s o therei s substirutionin favor of
a heterosexual alternative. Moreover, the value of marriage as a way of
concealingh omosexuala ctivity is greatert he lesst olerantt he societyi s and
hencet he greatert he benefito f suchc oncealmentO. n this view,h eterosexual
marriagec ould acruallyb e a complemen(t in the economics enseo f the term,
discussedla ter in this chapter)t o homosexualitya s well as a substitutefo r it.
But it would be simplistic to conclude that the marriage rate for male
homosexualsw ill always and everywhereb e low in tolerant societiesT. he
costo f homosexuality( or of beinge xposeda s a homosexuali)s not the only
considerationt;h e costt o the homosexuaol f heterosexuaml arriagem ust also
be considered.I n some tolerant societiesn, otably that of ancient Greece,
marriage was cheap for a homosexual. He was not required to associate
intimately with his wife savef or occasionasl exuali ntercourseT. rue, there
werea lsos trongs ocialp ressuretso marry,a nd lifelongb achelorsw erel ooked
upon with suspicion-not becauset hey were thought to be homosexuals,
however,b ut becauseth ey weres hirkinga sociald uty.T heses ocialp ressures,
and the suspiciono f older bachelorsm, ay iust have reflectedt he easew ith
which a homosexuacl ould marry.H e reallyw ould be an oddballi f he forwent
marriage,f or he would be giving up the chancet o havec hildrena nd gaining
nothingi n return.G reek-stylen oncompanionatme arriagew asn ot a substitute
for homosexualr elations,s o little time and emotionali nvestmentd id it require;
the homosexuawl as not forcedt o choose.I t is in the tolerants ociery
in which companionate marriage is the norm that we should expect the
proportion of male homosexualsw ho marry to be low; modernS wedenis an
example.
In considerings ubstitutabilityin matterso f sex,w e must be carefuln ot to
conhne analysis to the substitution of one sexual practice for another. This
would imply unrealistically that the total amount of sexual activity of an
individual and even a community is a constant.A bstinence,im plying the
12. HaroldT . Christensen'S, candinaviaann dA mericanS cxN orms:S omcC omparisonwsi th
SociologicaIfm pfications,"2 2 loumal of SociolI ssues5 0,72 (April 1966).T hc spccificc omparircn
is bctwccn Utah, homc of thc Mormons, and Dcnmark.
13. MichaelW . Ross,T he Manied HomosexuaMl an: A PsychologicaSlt udy1 10-111a nd
ub. 11.1 (1983).F or cxample,t he pcrccntaSocf malc homosexuailns thc UnitcdS tatcsw ho
marry ir roughly rwicc as high as in Swcden.
1 1 8 A THEORY OF S E X U A L I T Y
substitutiono f nonsexualf or sexuala ctiviry, is a possibler esPonseto an
increasein the costs,o r decreasein the benefitso, f sex.T his is implicit in the
earlierp oint that an increasei n the effectivenesosf contraceptivem ethods
r"sults in a higher frequencyo f coitus;a nd recallt hat Malthus believedt hat
the high cost of chiliren iesulted in delayed marriage, often with sexual
absdnJnceu nril then,T he strongera person'ss ex drive, and the more varied
the possibles ourceso f sexuals arisfactiont,h e lessl ikety is abstinenceto be
the iesponset o an increasein costso r a reductioni n benefitso f a preferred
sexualp ractice.I mprovementsin contraceprionb, y reducingt he_principaclo st
of heterosexuailn tercourseh, avea lsor educedt he benefitso f abstinence.
My emphasiso n substirutioni s likely to be resistedo n the ground that it
makis all sexuala nd indeedf amilial relarionshipas ppearc ommerciailn character
and by doing so distortst heir actual characterI.t may seemt o imply,
for example,t hat in a legal systems uch as ours that permitsd ivorce,- every
happily married man and woman must always be in the market for a becer
,pour.. nu, that is not in fact an implication.r J(es aw in the precedingc hapter
tirat love can be given a precise economic meaning that excludes such implausibleh
ypotheies,i t is a preoccupationw ith the unique particularso f
anotherp erson,p arricularsf or which there is, by de6nition,n o substitutet o
be found in any other person. This is why extramarital sex is often a poor
substitute for marital sex in companionate marriage. Rational choice and
economicd ecisionm aking are not synonymsf or commerciael xchange'
Lovei llustratesth e differenceb erweena bioeconomica nd a purelye cotromic
approach to sex. The idea of the loved one as unique is an aspect of the
Uonaing or imprinting observed in animal as well as human senings and
illustraied by the mutual bonding of mother and infant in a variety of species.
A purely economica nalysisw ould emphasizes uch factorsa s the high costs
of'r.arch for a new mate, the costso f seParationto children and, through
altruism, to the parents, and the benefits to one's reputation of adhering to
commitments.N one of thesef actors,h owever,e xplainst he emotionalc haractero
f the love bond. Their significanceis in making possiblep redictiono f
how durable the bond will prove in various circumstances.
Love, then, is natural in a fairly literal sense; where culture enters is in
deflecting,s utlimating, erasing,o r otherwisec ontrolling love. If choice of
spousei s important to the familieso f prospectives pousest,h e familiesr vill
rry to channelo r if necessaroy verridet he developmenot f love.raS incel ove
is more imponant in companionateth an in noncompanionatem arrlage'o ne
can predici that the greatert he power of familiesi n a society,t he weakert he
institution of companionatem arriagew ill be.
Two rypeso f erotic love shouldb e distinguishedO' ne-headlong, passionate,
romantic, and usually brief (call it infatuation)-is likely to disrupt a
14. William J. Goodc, "The Thcoretical lmponancc of Love,' 24 American sociological
Revieu 38 11959\.
Sex and Rationality 1 1 9
system of companionate marriage both by impeding rational marital search
and by making spousess usceptibleto adulteroust emptationsa fter any infatuation
berween them has faded. It is fairly harmless in noncompanionate
marriage:t he marriagei s likely to haveb eena rrangedb y thes pousesf'a milies;
the husband is not expected to be faithful; and the wife is likely to be under
continuouss urveillanceT. he other rype of erotic love is that which accompanies,
r einforces,a nd is reinforcedb y a relationshipc haracterizebdy emoiional
intimacy, shared values, and constant companionship.
Although noneconomistsm ay believet hat I overemphasizteh e substitutabiliry
of sexual practices,s exual obiects,a nd sexualt ransactionalf orms'
economistsm ay believet hat I underestimatsee xuals ubstitutabiliryb y distinguishings
o sharplyb etweens exuala ctsa nd sexualp referencesC.o nsidert he
analogyo f smoking.A personm ight giveu p smokingb ecauseo f concerno ver
his h;lth yet retain an undiminished desire to smoke; he would be like a
homosexuawl ho was deterredb y fearo f punishmenot r diseasefr om engaging
in any homosexuala cts.B ut most smokers,a fter a period of abstinencel,o se
their tastef or smoking;t heir preierencec hangesin line with their behavior.
It is an empirical questionw hether sexualp reference( of heterosexualsa s
well as homosexualsi)s more like smokingo r more like beingl eft- or righthanded.
A lefr-handedp ersonc an forceh imselft o write with his right hand,
but he neverf eelsc omfortabled oing so' Evidenced iscusseidn Chapter4 and
elaboratedi n later chaptersh as persuadedm e that sexualp referenceis far
closer to 'handedness" than to smoking'
The Kinseys cale,h owever'm ay seemt o imply substitutabiliryo f preference
as well as o? act. Although for clariry of exposition I frequently divide the
world into heterosexualsa nd homosexualsK, inseyw as surely correct that
many peopleh aveb oth heterosexuaal nd homosexuapl referencesin; deed,t he
fa.t ih"t somep eopleo f predominantlyh eterosexuaplr eferencea rew illing to
substirutea homoiexual sex act in a pinch (and vicev ersa)c ould be thought
to mean that they have some degree of homosexual taste or inclination. I
acceprt he Kinsey scale as a useful way of expressingth e distribution of
heterosexuaal nd homosexuapl referencesb,u t I believeth at one'sp ositiono n
the scalei s exceedinglyre sistantt o sociali nfluenceT. he positiond etermines
how tikely one is to substitutea homosexuafl or a heterosexuasle x act, but
does not imply that one can be shifted from that position to even an adiacent
one. ln behavior one might move from a 6 to a 0; in preferenceo ne will
remain a 6.
The CentralitY of Seatch Costs
The different types of cost rhat sex involves, and the major factors that bear
on these. ortr, h"u. now to be consideredO. ne is the cost of search'lsI t is
15. The subiecto f an enormousl itcrature in economicss, omeo f it conccrncdw ith marital
t20 A THEORY OF S E X U A L I T Y
zero for masturbationc, onsidereda s a solitarya ctiviry'w hich is why it is the
cheapesot f sexualp ractices.( The qualificationi s important: "murual masturbation,'
heterosexuaol r homosexuali,s a form of nonvaginali ntercourse,
and its search costs are positive.) I do not mean that (solitary) masturbation
is, necessarilyc,o stlessto ut court. Apart from the trivial time costs,t herea re,
if the practicei s sociallyd isapprovedc, ostso f concealmenat nd often of guilt.
The economic theory of search may help explain the curious but wellconfirmed6
nding that marriagesi n which the spousesc ohabitedw ith each
other before marriagea re no more stablet han marriagesin which they did
not. evena fter correctioni s made for other differencesb erweent he cohabitators
and the noncohabitators.lT5h e reasonm ay be that couplest hat need
more information about each other delay marriage and use cohabitation to
generate that information, while couples that already have the necessary
information skip that preliminary. The rwo rypes of couple may neveftheless
have the same amount of information when they marry. If premarital cohabitation
is forbidden or discouragedh, owever,t he first set of couplesw ill
experienceh igherc ostso f marital searcha, nd asa resultm akes omewhLalet ss
satisfactorym arriages.
Searchc osts explain why, among prostitutes,s treerwalkersp rovide the
lowestq ualiry of sexuals ervicesa nd call girls the highest.lTS treetwalkersa,s
distinctf rom bar girls,p rostitutesw ho work in brothelso r in massagep arlors
or for escorts ervicesa, nd, particularly,c all girls, imposev irtually no search
costso n Dotentialc ustomersa. nd as a result minimizet he full cost of their
servicesI.f searchc ostsa re relativelyi nvariantt o qualiry of servicet,h ey will
be a larger percentageo f the full cost the lower the qualiry, and hencet he
lower the other costs,o f the servicea, nd so there will be more reluctanceto
bear them when qualiry is low and more willingness when it is high, The
analogy is to transportation costs. High-value goods are shipped a greater
distancet han low-valueo nesb ecauses hippingc ostsh avea smallere ffecto n
the 6nal price of the former. Men will incur considerables earchc ostsf or a
call girl-and evenh igher ones,o f course,f or a mistresso r a wife-but few
or nonef or the lowestq ualiryo f sexuals ervicesw, hich thereforea rep rovided
searcha nalogizetdo job searchin the employmenmt arket.S eeb, csidesB ecker'sw ork citcdi n
Chaptcrt , DalcM ortcns€n',M atching: Findinga Panncrf or Lifc or Othcrwis,e" 94 American
Jounal of Sociology5 215 (1988);M ichaclC . Kecley", The Economicosf FamilyF ormation,"
15 Economicln quiry 238 (1977)F. or a lucidi ntroductionto thc cconomicosf searchs,c eG eorge
J. Stiglcr7, he Theoryo f Pricel -5 andc h. l4 {4th ed. 1987).
16. Sec,f or cxampleA, l{redD eMarisa nd GcraldR . Lcslic,' Cohabitationw ith the Futurc
SpouscI:t s Influenceu pon Mariral Satisfactioann d Communication,4"5 lounal of Mattiage
and the Family 77 (1984).
17. For a goodd escriptiono f thc structurco f the markcti n prostitutionin the UnitedS tates
today,s ccM iller ct al.,n ore5 above,a t 263-269.
Sex and Rationalin 121
by streetwalkerst,h e zero-search-cosptr ostitutes.A narogouslnt he lowestqualiry
goods are consumed at home, not shipped at all.
search costs are positive for any rype of interactive sexuar activity. and
sometimesth ey are, for all practicep urposesi,n finite, In a well-run prison,
the costo f fnding a sexualp aftner of the opposites exm ay approachi nnniry.
But is this really a searchc ost?o ne might supposeth at if the imale)p risoner
knows therca re no women in the prison,h is costo f heterosexuasle archw ill
be zero-not infinite.B ut if "search"i s understoodto includen ot only looking
for a match but also making it, the secondc haractrei zation-infinite searci
cost-is more accurate.
Not all prisons are well run. A prisoner can try to bribe a guard to bring
in a woman; he can try to break out of the prison; he can try to assault I
femalee_m ployee.s ome prisonse vena llow coniugalv isits.N everthelesst,h e
costo f heterosexuasl earcha, nd henceo f heterosexuasle x,i s much higherf or
prisoners than for free persons, and we can therefore expect hom;sexual
relations to be more common within prisons than outside ihem. even if the
distribution of homosexualp referencea mong prisonersi s identicalt o that
amongf reep ersons.lsIn Iike vein we can surmiseth at thc riseo f monasticism
in the early christian period was a factor in the preoccupationo f theologians
with masturbation and homosexualiry.
The prison examplec an be generalizedC. onsidera hypotheticaml an who
setsa valueo f 20 on sexw ith a woman of averagea ttractivenesasn d a value
of 2 on sex with a male substitute-perhapsa n effeminatem ale or a male
made up to resemble a woman. (I intend these valuations to reflect all alternative
uses of a man's time other than those involving a sexual relationship.)
If the cost of sex with a woman is 30 and with a man only 1, our hypothetical
man will prefer sex with the man rather than with the woman, even though
his strong preferencefo r women over men as sexualp artnersm arks him as
heterosexualH. e might make this substitutione veni f he could haves ex with
some women at a cost lower than the bene6ts.S upposeh e could have sex
with a woman he valued at only 1 at a cost between 0 and 1; he would still
prefer sex with the man. Of course, a man of even stronger heterosexual
preferencem ight set a valueo f 0 on sexw ith anotherm an,B ut we know that
many men will accept a man as a sexual partner even though they greatly
prefer (equivalentw) omen to men,
Generalizing further, we can assume that our hypothetical man has a
downward-slopingd emandc urve for heterosexuasl ex acts.I t slopesd own-
18. On the incidcnceo f homosexualiriyn prisons,s ecR ichardT cwksbury", Mcasureso f
ScxualB chaviori n an Ohio Prison," 74 Sociologya nd SocialR esearch3 4 (1989); D. J. Vest,
Homosexuality Re-Examined 233-240 (7977); C. A. Tripp, The Homoscxual Matrix 222n,
(1975);M ikhail Stern,Seixn the USSR2 17 (1980)S; hakuntalDa cvi,T DeWorldo f Homosexrals,
ch, 8 (19771;D avis,n otc 3 abovc.W hcthcr the distributiono f homosexuapl rcfcrenceis the same
within prisonsa s outsidc thcm is considcredin thc ncxt chapter.
L22 A THEORY OF S E X U A L I T Y
ward not only becauseth ere is diminishingm arginal utiliry of sex, but also
becauseth e malc taste for sexualv ariery impliest hat homosexuals ex acts
become bener substitutes for heterosexual sex acts the more sated with the
latter a man becomes. Thc higher the cost of heterosexual sex, the more the
man will substitute other activities-some of them nonsexual, such as watching
televisiono, therss exual,s ucha s masrurbationo, r beingf ellatedi n a restroom
encounter,
The more attractive someone is to a given class of potential sexual parmers,
the lower his costs of search for such partners will be; he will not have to
search as far to find a parrner whom he likes and who likes him. We would
thereforc expcct unattractive men to patronize prosdrutes more than attractive
men do, and effeminate men-€ven if they are hetcroscxual-to engage in
more homosexuala ctivity, on average,t han noneffem inate men, if we issume
that cffeminate men are on avcrage less attractivc to women than more
masculinem en are.
The suggestion that effeminate men who are no more homosexual in prefererce
than other men will nevenhelesse ngageo n averagei n more homosexual.
b eh.auioirs suffciently counterintuitiveto warrant eiaboration.I ts plausibiliry
hinges on the meaning of effeminacy. The term crops up constantly in
discussionosf maleh omosexualirya, nd is resistedb y homosexuarbse cauiei t
is pejorative( althoughI do not usei t peioratively)a nd becauseb y comparing
male homosexualst o women it assimilatesth em to what is still the subordinati
sexi n our sociery;f eministsw ould say that this is uhy thet erm is pejorative.
The term is also vague, and for my purposes that is a more serious concern.
C_A. . Tripp, however,h as offered a rypology of effeminacies:" nelly," "swish,"
"blas€," 'camp.'rr Roughly thesem ean,r espectively, feminine, a masculine
caricarure of feminine, regal (like a fashion model), highly affected. Tripp was
writing in the mid-1970s, and today one might wish to add a fifth-rype:
boyish, androgynousa, PeterP an. He was writing about homosexuals,'but
some hcteroscxual malcs fit into one of the five types I have described.
_As one moves along the spectrum from the conventionally masculine to the
effeminatet,h e percentageo f men who are homosexuarl ises,a nd if women
are not surew hich effeminatem en are heterosexualt,h ey will tend to apply
a discount facor to all such men--especially in the age of AIDS-with ihe
result that thc averagee ffeminate( but heterosexualm) an will incur higher
costso f heterosexuasl earcht han other heterosexuaml en.2oH e will havi to
19. Tripp, notc 18 abovc, at 177-t90,
20. For somcc vidcncct hat most womcn do in fact prcfcr masculinc-appcarinmg cn, secp aul
J. I:vrakas, "FcmalcP rcfcrcnccfso r Malc physiqucs,"9 loumel of Reseirchin Fersonarity3 24,
329, 331 (19751i Edward K. Sedalla, Douglas T. Kenrick, and Bcth Vcrshurc, "Dominancc and
HctcroscxualA ttracdon," 52 Jourral of Peronality and Socialp sycbology7 30 (19g7); Michacl
R. cunningham, Anita P. Barbcc, and carolyn L. pikc, 'what Do womcn vant? Facialmctric
Sex and Rationali* L23
look longer and harder to find a woman with whom to have a marital or
other sexualr elationship,a nd this will increaseth e likelihood of his substituting
a man. The analogy is to two persons who ha"e the same degree of
preferencefo r vanilla over chocolatei ce cream,a nd therefora-if the prices
are the same--If the price of vanilla to the 6rst person rises, he will buy less vanilla and
more chocolatet han the secondp ersone vent houghh is preferencefo r vanilla
is intense.
To avoid any misunderstandingle, t me emphasizeth at I am speakingi n
marginalt erms. I do not suggesth at everye ffeminateh eterosexuael ngages
in some homosexual behavior, but only that, given two groups of men of
identicallym otile sexualp referenceo, ne group being more effeminatet han
the other,t herew ill be more homosexuaal ctiviryi n the first group.
A further implicationi s that handsomeh eterosexuaml en will, on average,
engagei n more homosexuala ctivity than men who are not handsome.W e
know from the preceding chapter that men tend to find good looks more
sexuallya rousingt han womend o.21T his is true of homosexuaml en asw ell.22
Hencea handsomem an will be relativelym ore attractivet o homosexuaml en
than to women, As a result, the costs of homosexual sex to the handsome
man will be lower relativet o his costso f heterosexuasle x than they are to
homelym en.B y the samet oken,h omelyw omen shouldh aver elativelyb ener
lesbiant han heterosexuaol pportunitiesb ecausew omen tend to place less
value on good looks in a sexual partner than men do.
Still anotheri mplicationi s that, givent wo heterosexuaml en,t he one with
the strongers ex drive is likely to engagein more homosexuaal ctiviry than
the one with the weaker drive. By stronger and weaker I mean a greater or
lesserd emandf or sex, in the economics ensein which A's demandf or some
good is greater than B's if at any given price A will buy more of the good
than B will; only in the present context price includes time, risk of disease,
and other nonpecuniaryc osts.T he strongera man's sex drive, it is plausible
to speculatet,h e broadert he rangeo f sexuals timulit o whichh ew ill respond.2r
Assessmenotf Multiplc Motivesi n thc Pcrccptiono f Malc FacialP hysicaAl nractivcncss,'59
loumal of Personalirya nd SocialP sycbology5 1,58-70 (1990).
21. For additional evidcncc, not rclated to biological thcory, scc Jcffrcy S. Ncvid, "Scx
Diffcrcnccsin Factorso f RomanticA nraction," 11 Se.rR olcs4 01 (1984),a nd studicsc ited there.
22. Judith A. Howard, Philip Blumstein, and Pcppcr Schwam, "Social or Evolutionary Thcorics?
S omc Observationso n Prefcrenccisn Human Matc Selcction,"5 3 Journal of Personality
and Social Psychology 194, 196-197 and tab. 2 (1987).
23. Glcnn D. Wilson, "Thc Ethological Approach to Sexual Deviadon," in Variant Sexuality:
Researcha nd Theory 84,92-93 (Wilson cd. 1987).T his may cxplain thc positivc corclation
notcd in Richard S. Randall, Freedom and Taboo: Pomography and the Politics of a Self Diuidcd
101-104( 1989),b etweenc onsumptioonf pornogaphya nds cxuacl xperiencc.
1 2 4 A THEORY OF S E X U A L I T Y
A prerryb oy may leavet he tepid heterosexuaclo ld but providea n acceptable
albeit not ideal outlet for the sex drive of the superheatedh eterosexual.2Ita
is like a passion for apples.T he passionatea pple eater is more likely to
considera slightly wormy or overripea pple an acceptables ubstitutef or an
oranget han the personw hosee nthusiasmfo r applesi s more measuredR. eplace
apples with sex and oranges with television and the point should be
clear.T he superheatedh eterosexuaml ay not be able to find enoughw omen
to satisfyh is sexuald esiresa nd, facedw ith a choiceb erweenm asturbation
and boys, may choose boys.
What may seemt o cut the other way is that if the heterosexuawl ith the
strongest sex drive is also the one most aftractive to women, his costs of
heterosexuasl earchw ill be lower, and this may balanceh is more intense
demand for sex. But betweenr wo noneffeminatem en, the one with the
strongers ex drive may be /essa nractivet o women.2sH e is likelier to seek
multiple sex paftners, thereby reducing the amount of protection that he can
offer to any one of his partners and her children. And if he is thus less
aftractive to women, this will further reduce his costs of homosexual sex
relativet o heterosexuasl ex.
There is only a superfcialp aradox in suggestingth at the averagew oman
may find both effeminatem en and highly sexedm en lessa nractivet han other
men. These are both exrreme rypeF-the 6rst failing to emit unambiguous
signalso f heterosexuapl referencea nd competencet,h e seconds ignalinga n
intensity of sexual desire likely to undermine his commitment to the woman.
Both disadvantagehsa veb eena mplifiedb y the AIDS epidemic.
The fact that homosexualp referencea nd homosexualb ehaviora re only
looselyc orrelatedh elpse xplaina curiousd ualismi n popular thinking about
24. On the useo f boysb y hctcroscxuaml ena s females ubstituress,e e,f or examplc,R andolph
Trumbach," SodomiticaAl ssaultsG, endcrR olca ndS cxuaDl cvclopmcnint Eightccnth-Century
London," in Tbe Putsuit of Sod,omyM: ale Homosexualityi n Renaissancaet d Enlightenmeat
Europe4 07 (KentG crarda ndG cn Hekmae ds.1 989);G eorgeA ustinC haunceyJ,r .,
.Gay New
York: Urban Culturc and the Making of a Gay Malc !9orld, 1890-1940" 77-97 (Ph.D. diss.,
YaleU niversity1, 989);A lan Bray,H omosexualitiyn RenaissanE.en glatd7 8 (1982);J amesM .
Saslow," Homosexualiryi n thc RcnaissanccB: ehavior,I dcntrry and Anistic Exprcssion,"i n
Hiddn ftom History: Reclaiming the Gay and Lesbian Past 90, 92 (Manin Bauml Dubcrman,
Manha Vicinus,a nd GeorgeC haunceyJ,r .,e ds.1 989);M crnissin, otc 1 abovc,a t 53; Villiam
H. Davcnport," An AnthropologicaAl pproach,"i n Theorieso f Human Sexualityl9 7,Z3Z llames
H. Gcer and William T. O'Donohucc ds. 1987).O n rcsidcntiabl oys' schoolsa s hotbcdso f
homosexualirys, eeJ ohn Chandos,B oysT ogether:E nglishP ublic Scbools1 800-1864 307-311
(1984);V ern Bullougha nd BonnicB ullough," Homosexualiriyn Nineteenth-CentuEryn glish
Public Schools," in Homosexualityi n lnteraational Perspectiue1 23 (JosephH arry and Man
SinghD as eds.1 980).E vidcncet hat oppornrnistich omosexualsin deedh avea snongcrs exd rivc
than cither exclusivch omosexualso r cxclusivch cterosexualiss notcd in Glcnn'Wilson.L oue and
Irctinct 197, 199 (1981).
25. For somee videnccs,c eS adallaK, enrick,a nd Vershurcn, otc2 0 abovc,a t 737.
Sex and Rationalin 125
homosexuality. In sexually permissive societies a homosexual tends to be
thought of primarily as a personw ho has a strongp referencefo r same-sex
relationsw, hile in sexuallyr epressivseo cietietsh e veryc oncepto f homosexual
preferenceis weak or absent,a nd a homosexuails likely to be thought of as
a heterosexuawl ho commitsu nnaturala ctsb ecauseo f lust or other wickednessT.
he explanationi s as follows.I n permissives ocietiesh,e terosexualhsa ve
plenty of opportunitiesf or heterosexuailn tercourse.2O6p pomrnistich omosexuality-
the homosexualiryo f heterosexuals-isth ereforer are.A t the same
time, homosexualsa re free to associatew ith one another-and do so, to
reducec ostso f search-forming visibleh omosexuasl ubculturesfr om which
hetcrosexualsle arn that there is such a thing as homosexuapl referenceI,n
repressives ocietiesh, omosexualsa re under strong pressureto concealt heir
preference(a s by marriage),a nd homosexuasl ubculturese itherd o not form
or are clandcstineB. oth factorsr etard awarenesso f homosexuapl reference,
sometimesa mong homosexualsth cmselvesA. t the samet ime, the sequestration
of unmarriedy oung women that is characteristioc f sexuallyr epressive
societiesr educesh eterosexuaol pportunitiesa nd thereforep romoteso pportunistich
omosexualityw, ith the resultt hat homosexuaal ctivityi n repressive
societiest ends to be dominated by opportunistich omosexualsI.t is even
possiblet hat, being only an occasionapl articipanti n the market for homosexuala
ctiviry,t he opportunistich omosexuaml ay be lesss killful at concealment
of sucha ctivity and hencem ore likely to be apprehendetdh an a "real"
homosexuawl ould be.27In that eventm ostp ersonsp unishedf or homosexual
behaviori n a repressiveso cietym ay be heterosexualsa,n d the evil-will theory
of homosexualityw ill have a definitep lausibiliry.T he perversityo f such a
punishments ystem( perversefr om the standpointo f repressingh omosexual
preferencei)s enhancedw hen, as is common,t he penetratori n homosexual
sodomyi s punishedm ore heavilyt han the receptor2s----evtheonu gh the laner
(if an adult) is more likely to be a "real" homosexual.
This discussionm ay explainw hy the word homosexuails of recento rigin
(it was coinedi n 1869) and also helpst o rebut a predictablec riticismo f my
suggestiont hat the highly sexedh eterosexuails more likely to engagei n
homosexuaal ctst han the lessh ighly sexed( but noneffeminateh)e terosexual.
The criticismi s that the self-identiryo f the highly sexedh eterosexuaisl likely
25. This is provided,o f courset,h at the socictyis permissivwe ith rcspectto femalea s well
as to males cxuala ctivity.l f thcrei s a vcry stront doubles tandardm, caningth at thc societyis
highly pcrmissivcw ith rcspect o malc scxuala ctivirya nd highly rcpressivwe ith rcspect o fcmalc
scxuala ctivity, thc analysisi n the text is changeda, s wc shalls cci n Chaptcr5 .
27. Michacl Schoicld,S ociologicaAl spectso f HomosexualityA: ComparativeS *dy of Thrcc
Types of Homosexuals 148-149 (1965). But sec Paul H. Gebhard ct al., Ser Offerders: An
Analysis of Types 357 (1965).
28. Guido Ruggicro,T he Boundarieso f Eros: SexC rimea nd Sexualityin RenaissancVeo ice,
ch. 5 ( 1985).
I
i
I
II
I
t26 A THEORY OF S E X U A L I T Y
to be, if anything,m ore bound up with reiectiono f homosexualiryth an the
self-identityo f other men. Not necessarilyT.h e experienceo f ancientG reece
and modern South America alike is that, as Foucault and others have argued,
homosexualirya s a normativer ather than a descriptivec ategoryi s cultural
rather than natural. If one is macho, one is indeed not a fairy, a poof' a
faggot; but whether sodomizing boys is the act of a fairy, a poof, a f.aggot
dependso n cultural factors-such ast he availabiliryo f youngw omeno f one's
social class.
An important determinant of search costs is the extent of the market. lt is
not o.rly important but fundamentali;t can subsumeb oth the prisone xample
and the effeminacye xample.I t explainsn ot only the obviousp oint that human
sexuali ntercoursew ith animalsi s more commoni n rural than in urban areas
but also the slightly less obvious point that incest, too, is more common in
rural areas.2Teh e largert he sampleo f potentials exp aftnerso f a givent ype,
readilya ccessiblteo the searchert,h e lower the costso f search.I n the prison
example,t he accessiblesa mpleo f potentials ex Paltnerso f the opposites ex
was assumedto be zero,a nd in the effeminacye xamplei t was assumedto be
smallerf or effeminateth an for noneffeminatme en,w hile in the inceste xample
the correspondings ample-the sampleo f accessiblPe ersonso f the opposite
sex who are not relatives-is smaller in rural than in urban areas'
Urbanization
It is when a particular rype of sex partner is rare in the society as a whole
that the searchc ostsf or that type of sex will be most stronglyi nfluencedb y
urbanization.S upposeth at a villageo f one hundredp ersonsc ontainsa single
homosexualA. s long as he confinesh imselft o the village,h is searchc ostsf or
a homosexuarl elationshipw ill be infinite,u nlesso ther homosexualsv isit the
village.H e can travel to other villages,b ut his searchc ostsw ill still be high
sincet hey includet he costso f travel.S o eveni f homosexualsa re initially no
larger a fraction of ciry dwellers than of village dwellers, our homosexual
villagerw ill be likelier to 6nd a homosexuapl armer at reasonablec ost if he
moves to a ciry. So he is apt to move. And as a result of the migration of
homosexualsto cities,t he percentageo f homosexualsin citiesw ill be higher
than in rural areasw, hich will makec itiese venm orea ttractivet o homosexuals
by further enlargingt heir market.l nsteado f one homosexuailn a villageo f
one hundredp ersons( 1 percent),t here may be 2,000 homosexualsin a city
of 100,000p ersons( 2 percent)T. he deartho f largec itiesi n colonialA merica
may explaint he widespreadim pressionth at in that periodh omosexualiryw as
29. JosephS hepherI,z ceslrA BiosociaYl iew 129 (1983);D . JamesH endcrson, "lncest: A
Synthesiosf Data," in TtsumaticA busea nd Neglecot f Childrena t Home4 23,424( GenrudeJ .
Williamsa ndJ ohnM oneye ds.1 980);N eil Ellion,S ezsraliriyn Scandinavi2a7 1 ll970l.
Sex and Rationality t 2 7
exceedinglyra re here,c omparedr o England,:oA furtherp oint is that therei s
more privacyi n the senseo f anonymityi n citiest han in rural areas.a nd hence
lessl ikelihood that unlawful sexuala ctsw ill be detectedw, hich reducest he
expectedc ostso f punishment,3l
The effecto f urbanizationi n increasingth e sampleo ver which a member
of a sexual minority can search at low cost for a sexual paftner is likelv to
be especiallyp ronouncedi n the caseo f maleh omosexualiryb ecausea, s I shall
explorem ore fully in Chapter 11, male homosexualsra rely form permanent
unionsa kin to old-fashionedm arriage.I f they did, the marketf or unattached
homosexualsw ould be reduced,p erhapsd rastically,a s homosexualsp aired
off and withdrew from that market.S inceA IDS encouragetsh is pairing,c ities
may becomel essa tEactivet o homosexualsth an they oncew ere.
By facilitatingt he creationo f marketsf or homosexuaal ctivity,u rbanization
affects not only the geographic distribution of such activiry but also its
amount. The village homosexual who has no potential male sex parrners
cannot be a practicing homosexual; when he moves to the city. he can be.
Justa st he creationo f a marketi n a conventionagl oodw ill leadi o an increase
in the supplyo f that good,s o the creariono f a marketi n homosexuarle lations
will (with a qualificationt o be noted)i ncreaseth e supplyo f homosexuaal cts.
The total numbero f practicingh omosexualisn a societyw ill thereforein crease
with the rise of cities,32m aking homosexualirys eem,a nd in a valid if partial
senseb e, a by-product of economicd evelopmenat nd moderniry.E vin if a
city, having attracteda largen umbero f homosexualss, topsg rowing, homosexuals
may continue to migrate to it, attracted by the large number of
homosexualsa lreadyt here;i f so,t he percentagoef homosexualws ill continue
to rise.T his is possiblee veni f whateverf actorsh avec ausedth ec ity to stagnate
operatea s powerfully on homosexualsa s on heterosexualsb,e causeh eterosexuals
can be expected to leave the city at a faster rate than homosexuals
sincet herei s no sexuala dvantagein staying.T he fact that the two citiest hat
are generallyc onsideredth e centerso f homosexualiryin the united states-
New York and SanF rancisco-havel ong beens tagnantin overallp opulation
is,consistenwt ith this analysisA. nother bit of evidenceis that the percentage
of Catholic priestsi n the United Statesw ho are homosexuahl as apparently
increaseds ignificantlys ince the 1950s,33a period during which the total
numbero f Catholicp riestsi n the nation has beenl evel.3Wa e shall seei n the
ncxt chaptert har rhep riesthooda ppearsto bed isproportionatelhyo mosexual.
30. David F. Greenbcrg,T he Constructiono f Homosexuality3 46 (199g).
31. Cf. Bran notc 24 abovc, at 43.
32. Alfred C. Kinscn Wardcll B. Pomcroy, and Clyde E. Manin, Sexual Behauior in the
Human Male 455459,530-531 (1948).
33. A. W. Richard Sipc,A SecretW orld: Sexualitya nd the Searcbfo r Celibacy1 07 11990).
34. Id. at 7 (tab. 1.1).
1 2 8 A THEORY OF S E X U A L I T Y
I haves aid that urbanizationi ncreasetsh e numbero f practicingh omosexuals.
I t doesn ot increaseth e numbero f personsw ith homosexuapl reference'
and it may not even increase the numbei of homosexual acts' Opportunistici"-
or"**Uty should decreasew ith the rise of cities becauset he privacy of
ciry life faciliiates nonmarital heterosexual sex, including prosdrudon.3s It is
i-. ,i", the costs of search for opportunisdc homosexual contacts are lower'
Sut t "- assuming that these aie second-best contacts' as they are for all
Ii."l" h",.ror.xuals, though not for the relativeh andfulo f genuineb isexuals'
If my analysis is correcti the precise sense in which urbanization fosters
ho-or.ru"ii.y is that it increasis the total number of homosexual acts committed
by "real" homosexuals.
The eifect of urbanization on homosexuality explains how advocates of
lro^*ou"t rights can make a credible, although almost certainly exaggerated'
claim that 10 percent of the American population is homosexual' No one
kno*, ho* many homose*ualsth erea re i[any society-the very concepto f
a .homosexual" is vagueb ecauseo f the equivocals taruso f act versusp referencea
nd becauseo ithe range of possiblep rcferences.B ut we shall see in
Chapter1 1 that the beste stimite ofthe percentageo f adult malesw ho have
" ,ir'ongty homosexual preference (thus excluding oppom.rnistic homosexuals'
-o., oi whom consideit hemselvehs eterosexualsis) 3 to 4 percent,a nd of
adult females1 percent,f or an averageo f 2 to 2'5 percent'B ecauseh omosexuals
are corr."ntr"t"d in a handful of cities in which they are a significant
fraction of the population,b ecauseo ur demographicim pressionsa re shaped
by massm edia'that are themselvecso ncentratedin one of thosec ities( New
iork), and becauseth e lingeringt aboo statuso f sexr etardst he dissemination
of accuratein formation about sexualp racticest' he public is open to persuasion
that there are many more homosexualsin the societya s a whole than is
really the case.
From an economic standpoint, prostiotion is a parallel phenomenon to
homoscxuality' Consider a io"i.ry in which there is a very large number of
bachelors.T his could be due to polygyny,a high cost of marriage,f.e male
infanticide, higher adult mortaliry among women than among men because
35. On the concentradono f prostitutcsi n cities,s ec,f or cxamplc,M ary Gibson, Prostitutiotl
dnd the state in ltaly, 1850-191J 102-103 (1985); Richard J. Evans' "Prostirution, Statc and
Socictyi n Impcrial Gcrmany,' 70 PastZ r Presmt 105 11976)l'n addition to thc greatcrp rivacy
in citics, thcie is a largcr number of transient malest hcre, which incrcascst hc demand for
prortitoics. A Iactor thal might secmt o cut thc other way is that thc scquestrationo f womcn is
morc costly in agrarianc ommunitiesb, ecauscw omcn arc valuablca s farm workcrs' Secn otc 37
bclow and accompanyingtc xt; and K. J. Dover, Grcek Popular Motality in the Time of Plato
and Aristotle9 8,i09 $t7q, But scqucstcringw omcn frccst heir husbandsfo r sexualc scapadcs
outsidc the homc. A socictyi n which womcn are scquesterccda n thercforcb e licentiousb y our
standardsc:o nsidctrh e examplco f ancicntG rcecco ncca gain'
Sex and Ratiorality t29
of high mortality in childbirth, or some combination of these factors. All but
the first were present in medieval Europe, and so it is not surprising that a
high percentage of young men were bachelors.35V hy might marriage be
costly? Suppose that in an era of limited contraceptive know-how, high infant
mortality, and no bonle-feeding, a woman who married thcreby disabled
herself from pursuing remunerative work because of the burden of frequent
pregnancies and of having to breast-feed the infants resulting from those
pregnancies. In the Middle Ages, and indeed until much later, the disability
was gre ater in towns and cities than in agricultural communities, where a
woman could work closet o home and on an interminentb asis.3I7f the wife
cannotw ork, the husband'sin comew ill havet o supportn ot only himselfb ut
alsoh er asw cll ast heir children,a nd this may be impossibleu nlessh is income
is high or the woman comes to the marriage with a dowry.38 A poor family
may not be able to afford a dowry, however, in which event marriage may be
out of the question for the man and, given a scarcity of high-income men, for
the woman as well.
The high cost of marriage, coupled with the fact that women frequently
died in childbirth, meant that a wealthy man might run through a series of
young women-losing his first wife in childbirth, remarrying, losing that wife
in childbirth, remarrying again. This was de facto polygynn and, interestingly,
was opposed, although not condemned' by the Church-and by young bachelors
through the practicc of charivari mentioned in Chapter 2. Like other
forms of polygyny it was sustained by inequalities of wealth. As between two
men otherwise alike, an older one with children by a previous marriage and
a younger who has never been married, the younger man would pay more for
a wife if both men had equalr esourcess,i nceh e would havea strongerd esire
36. On the mcdieval dcarth of women and its causcs,s eeJ ohn Boswell' Thc Kindnesso f
Strangers: Thc Abandomnetrt of Childrea in Westem Europe from Late Antiquity to tb. Rcnais'
sance2 58,409 n.34 (1988); "Mcdieval CcmctcryP atternsP: laguca nd Nonplagtc," in Medinal
Demograpby: Essaysb y losiab C, Rlssell 148 (1987); Marcia Guncntaga nd Paul F. Secord,
Too Many Women? The Sex Ratio Question 54-55 (1983); David Herlihy, The Social History
of ltaly and VestemE urope,700-1500C: ollecteSd tudiesc,h s.1 3,1 4 (1978)J; acqueRs ossiaud,
MedievaPl rostitation1 8 (1988).
37. Hcrlihy, note 36 abovc, ch. 13 at 15; Rae ksscr Blumberg, "A Paradigm for Predicting
thc Position of Womcn," in Ser Rolcs and Social Policy: A Complex Social Scimce Equation
ll3, 123 (Jcan Lipman-Blumcn and Jcssic Bemard cds. 1979). Morcovcr' womcn arc morc
valuable thc morc children arc valucd, and childrcn tend to bc highly valucd in farming communitics
becausct hcy can do usefulf arm work from an early agc. Sec,f or example'M ark R.
Roscnzwcig',T he Dcmandf or Childrcn in Farm Households,"8 5 lounal of PoliticalE conomy
123 tr977t.
38. On thc opcration of a dowry systcm, scc Julius Kirshncr and Anthony Molho' "Thc
Dowry Fund and thc Marriage Markct in E^ '/ Quatnocento Florcncc," 50 lounal of Modcm
Hktoty 403 ll978l.
1 3 0 A THEORY OF S E X U A L I T Y
both for a child (assuming that there is diminishing marginal utility of childrcn
as of othcr goods)a nd for sex (assuminga, lso realisticallyt,h at intcnsityo f
males exuald esired eclinesw ith age).B ut if the older man was wealthier,h e
might outbid the younger man even if he did not have as strong a desire to
marry. The result would be another unhappy bachelor.
Enter prostirudon. The existenceo f a large number of young bachelors
createsa demandf or paid sexuals ervicesin substitutionf or marital sex;t he
existenceo f a number of unmarriedy oung women createsa supply.I t might
seem that if every bachelor were supporting a prostitute by paying for her
servicest,h ere would be no cost savingsc omparedt o marriage.B ut the premise
is incorrect, not only because the customer of the prostitute will not support
the prostitute'sc hild (and for this reasona lone,a prostitutew ill take extreme
measuresto avoid becominga mother),b ut also becausee acho f the prostitute's
customersp ayso nly a small shareo f the costso f supportingh er. Since
the average prostitute has many more than one customer, there are fewer
prosdrutest han there are prostitutesc' ustomersa, nd the costo f maintaining
a woman is spread over a number of men, reducing the cost to each. A man
who frequentsp rostitutesi s not supportinga couple,l et alone a family, but
insteado ne man (himself) plus maybea hundredtho f a woman.
'What
we seeh ere is de facto polyandry,3ew hich in medieval times blotted
up someo f the surplusb achelorsa, nd in doing so offsett he serialp olygyny
just mentioned.O f coursef or bachelorsw hoses exualp referenceru ns to men
an)rwayh, omosexualitym ay providea cheapera s well as a more pleasurable
substirutef or marriaget han patronizingf emalep rostitutes.B ut they are related
practices, and this may help explain their bracketing in the English
debates over public policy toward sex. The famous Wolfenden report, for
exampfe, was a srudy of the laws against homosentality and the laws regtrlating
prostitution.aoT here is a more direct relationship,w hich incidentally
supportst he emphasisin this chaptero n the substitutabiliryo f sexualp ractices:
one does not have to go back to fourth-century B.c. Athens to find
39. "Enabling a small numbcr of womcn to take carc of the needs o{ a largc numbcr of men,
it [prosdrution] is the most convcnient scxual outlet for armics and for the lcgions of strangcrs,
pcrvcns,a nd physicallyr cpulsivei n our midst." KingsleyD avis, "ScxualB ehavior,"i n ContemporaryS
ocialP rcblems3 13, 351 (RobcrtK . Menon and RobcnN isbcte ds.,3 d cd. 1971)." A
young man, lacking a lcgitimatc wife, must satisfy himself with thc dcvalucd woman of tlc
brothcl . . . Thc disordcr of fornication was allowcd to young mcn to insure a superior ordcr,
the stability of thc family." Madelcine Jcay,
"Scxualitya nd Family in Fiftccnth-CcnturyF rancc:
Are Litcrary Sourccsa Mask or a Mirrcr?" 4 lountal of Family History 328,3 40 (1979).F rontier
socictics,s uch as thc ninctccnth-ccnturyA mericanV est, also tcnd to bc thick with prostitutcs
becauseo f thc high ratio of men to women.J ohn F, Deckcr,P rostitution:R egulationa nd Control
s9 (1979).
40. Reporto f Committeeo n HomosexualO ffensesa nd Prostitution,G reat Britain lU.S.e d.
1 9 5 3) .
en
of
he
to
rs
te
ht
er
5e
rt
le
:l-
:e
Sex and Rationality 1 3 1
heterosexualws ho have considereda boy or a young man to be a tolerable
substitute for a female prostitute.'{t
The demand for prostitution can be expected to be greatei the lower the
average wages of women relatiue lo the average wages of men, for then men
can easily afford to patronize prosdrutes. This helps explain the extraordinary
incidenceo f prostitutioni n modem Bangkok,w heret he prostitutesp' rincipal
customers are foreign tourists from countries in which wage levels a.e many
timesh ighert han thoseo f rural Thai women.a2lT ithout internadonatlo urism.
however, a society will encounter a problem in trying to use prostitution to
cope with the problem of a surplus of bachelors.T hc more extreme the
shonageo f women, the highert heir wagesw ill be; therefore,t he more bachelors
there are, the fewer will be able to afford prostitutes. So we can expect
that in a society in which the rario of men ro available women is high,
oppomrnistich omosexualitya sw ell asp rostitutionw ill be usedt o equilibrate
the demand for and supply of sexual services.( Later, however,w e shall
consider an imponant qualification when the shonage of women is due to
polygyny.)O ther methodso f copingw ith a surpluso f bachelorsa re monasticism
and eunuchism-the former a characteristicm edievalt echniqueo f
Roman Christians, the laner of Byzantine Christians and their Muslim successorsI.
discussm onasticismin the next chapter.
In describingp rostitutiona s a substitutef or marriagei n a socieryt hat has
a surpluso f bachelorsI, may seemt o be overlookinga fundamentadl ifference:
the "mercenary"c haractero f the prostitute'sr elationshipw ith her customer.
The differencei s not fundamental.I n a long-termr elationships uch as marriage,
t he pafticipantsc an compensate acho ther for servicesp erformedb y
performing reciprocal services, so they need not bother with pricing each
service, keeping books of account, and so forth. But in a spot-market relationship
such as a transaction with a prostirute, arranging for reciprocal
services is difficult. It is more efficient for the customer to pay in a medium
that the prostirute can use to -purchases ervicesf rom others.
The percentageo f men,e speciallyh eterosexuaml en,w ho are unmarriedi s
lower in modern society than it was in the Middle Ages. Yet prostitution
persistst,h ough on a smallers caleT. he declineis consistenwt ith the economic
model,b oth becaused emands houldf all when the numbero f bachelorsfa lls.
sincc they value prostitution more than married men do, and becauset he
41. Chauncry, note 24 abovc, at 91-93.
42. PasukP hongpaichit,F rom PeasantG irls to Bangkoh Masseuse7s4 -75 (l9BZ). For an
examplcf rom ninctecnth-ccnrurAy mcrica,s ceC hristincS tansell,C ity of
'Women:
Sex and Class
in Neu Yorh, 1789-1850 174-180 (19861. Marxistslcncralizing, as it were, from these examplcs-
wcrc wont to sayt hat thc dcmandf or prostitutionc omcsf rom the bourgcoisica ad the
supply from thc prolctariat,B ut this is not quite accuratea, s wc arc about to scc.
:r
rg
rt
f
n
I
t
132 A THEORY OF S E X U A L I T Y
supplyo f women for prostitutions houldf all as alternativejo b opportunities
for women emerge.ar \(/e also expect-and we find#-that the demand for
prostitutions hifts toward the lower classesa s the middle classm oves( more
rapidly than the lower classes)ator ward companionatem arriage.T his is becausep
rostirution is a poorer substirutef or sex in companionatem arriages
than for sex in noncompanionatem arriages.
That prostitution should persist at all may seem surprising in light of the
sexualr evolution,w hich hasv astlyi ncreasedth e numbero f women available
for casual iaisonsT. he economicm odelo ffersa possiblee xplanationS, tudents
of medievalp rostitutionh aveo bservedth at the prostituteso f the time offered
their customers "normal," ordinary sex; students of modern prostinrtion
observe,c onfirmingp opular impressiont,h at today'sp rostirutess pecializein
"kinky" sex.a6T he reasonm ay be that in the medievalp eriod the principal
customerso f prostitutesw ereb achelorsw, hereasto day,w ith the sharpd ecline
in the number of heterosexuabl achelorsu nablet o form sexualr elationships
with unmarried women at reasonable cost, most customers of prostitutes are
perforce married men.a7 Married men and men with steady girlfriends have
all the "normal" sex they want, at home, more or less for free, and if these
are companionater elationshipsi,t is better sex than with prostitutes.S ince
prostirutesc annotu nderpricet he wiveso r girlfriends,t hey havea n incentive
to differentiate their services-to offer something for which married men and
men with steadyg irlfriendsw ill pay becauseth ey do not have accessto the
identicals ervicei n the (free)h ome market.
43. Vcrn L. Bullough, The History of Prostitution 90 (1964). So it is no surprise that there is
linlc prostitution in Swcdcn, even though it is not illegal. Deckcr, note 39 above, at 130-132.
For thcrc is considcrablcs exualf reedomi n Swcden,a nd, as we shall seei n thc next chaptcr,
women's opportunities in thc job market arc probably better thcre than in any other country.
44. TheodoreN . Ferdinand." Sex Behaviora nd thc AmericenC lassS tructure."i n Sttrdiesin
Humon Sexual Behavior, norc 2 above, at L66, 774.
45. Rainwatcr. note 2 abovc.
45. CompareR ossiaud,n ote 36 above,c h. 8, and Leah Lydia Otrs, Prostitl,tioni n Medieual
Society: Tbe History of an Urban lnstitution in Languedoc 100-104 (1985), with fuchard
SymanskiT, he Immoral LandscapeF: emaleP rostitutioni n WestentS ocieties6 6-68 (1981),a nd
Jcnnifcr James,
'Prostitutcs and Prostitution," in Deuiants: Voluntary Actors in a Hostile World
358,402409 (EdwardS agarina nd Fred Montanino eds. 1977),J amesa lso emphasizesm, uch
in thc spirit of my analysis,t hc frequcncyo f cripplcd mcn, impotcnt mcn, and travclcrsa mong
a prostitutc'sc ustomcrs--rll beingm en with eitherp ermancntlyo r temporarilyh igh scarchc osts
in thc markct for "frcc" scx.
47. Rossiaudn,o tc3 6 abovc;S ymanskni,o tc4 5 abovc,a t 54, 73 (tab.4 .3);D eckern, otc3 9
abovc, at 169, 218; David A. J. fuchards, "Commcrcial Scx and the Rights of thc Person: A
Moral Argumentf or thc Dccriminalizationo f Prostitution," 127 Uaiuersityo f PennsylvaniaL aw
Reuiew1 795,lZ70 (1979).R ossiaudp ointso ut (at 39-41) that marricdm enw eref orbiddento
patronizc the medicval brothels and that prostirutesd cscribcdt hcir reladonshipsw ith rheir
customcrass 'Ve arc marricdf or thc night."
Ser and Rationalitv 1 3 3
If cities attract prostitutcs and homoscxuals-if, moreover, as the economic
theory of search suggests, 4zy soft of minority sexual taste (any minoriry
tastc,p eriod) will be bettcr scrvcdi n citiesa nd thereforeit s practitionersw ill
be attracted there-it becomes easy to understand why "vice" is traditionally
rcgardeda s an urban phenomenona nd why the averageu rbaniteh as a more
tolerant and permissivev iew of sex than the averagen onurbanite.aP8e ople
revoltcd by sexual excess or deviance (as it seems to them to be) will tend to
leave cities, and their deparnrre will reinforce the moral and attitudinal diffcrences
berween city and country, more precisely between large cities and
small towns,4ei n matterso f sex. Within cities,m oreover,g roups of sexual
deviantsw ill tend to clump in order to reduces earchc osts.H ence" red light"
districts will cmerge even without zoning regulations and make deviants more
conspicuous, reinforcing the cities' reputation for vice. Also, when the cost of
marriagei s higher in citics than in rural areasb ecausem arried women in cities
find it more difficult to get work, there will be not only more prostitution and
oppoffunistic homosexuality but also more couples "living in sin,' more
abortion, more illegitimacy.50
Income and Wealth
Urbanization is one explanatory variable in the economic theory of sex;
income is another. I have already noted the role of income in relation to the
incidence of marriage. But thcre its effect is indirect; to the extent that poverty
reducest he marriager ate, it increasesth e number of bachelorsa nd so the
demandf or prostitution,a nd it alsoc reatesa supplyo f unmarriedw omen to
be prostitutes. Income may have a direct effect as well. Although a person
who has more money can afford more of anything than a poorer person,
including sex, some goods are what economists call inferior goods, meaning
that an increasein consumersi'n comew ill-provided that price and quality
remaint he sam*reduce the quantityo f the good demandedS, exm ay be an
48. For illustrativc studies, see G. Edward Stephan and Douglas R. McMullin, 'Tolerance of
Scxual Nonconformity: City Sizc as a Situational and Early Lcarning Detcrminant," 47 American
SociologiczRl cuieut4 ll (19821;T homasC , T9ilson,' Urbanism and TolcranccrA Testo f Somc
HypothescsD rawn from Viah and Stouffcr,"5 0 AmericanS ociologicaRl euiew1 17 (1985);b ut
scc Mark Abrahamson and Valcric J. Carter, "Tolerancc, Urbanism and Region," 51 American
Sociological Rcview 287 (19861.
49. Not only are incest and bcstiality morc common in rural atcas than in citics and womcn
in farming communitieslc sss equctcrcd,b ut, as wc arc shonly to scc,i ncomcm ay bc negativcly
rclatcd to amount of sexual activity-and rural dwcllers oftcn havc much lower incomes than
city dwellcrs.F or all thcscr casonss exualm oralsi n thc countrysidca, sd istinctf rom smallt owns,
may bc lax by middlc-classs tandards.G . R, Quaifc, Vatton Varches and Vayutard Vives:
Pcasantsa nd lllicit Sexi n Ea y Sevcttt.enthC eflturyE ngland1 79,245 (1979).
50. Cf. John R, Gillis, For Benar, for'Worse: British Maniages, 1600 to thc Present 16l-166
(1985).
1 3 4 A THEORY OF S E X U A L I T Y
inferior good.rl It is time-intensiveb, ecauseti me is an important component
of search and search an imponant componenr of sexual activity, and the cost
of time rises with income earned from work.
This is far from a complete analysis, however. To begin with, sex is a
leisure-timea ctiviry( ir is not doned uringw ork), and leisureis usuallya ssumed
to be a superior good-people demand more of it as their incomes rise. Sex
competesw ith other leisurea ctivities,t o be sure; but even if, becauseit is
time-intensiveit,s shareo f the total time that peopled evotet o leisurea ctivities
is falling, the absolute amounr of sexual activity may be rising.
Prediction becomes easier if we distinguish berween wealth and (earned)
income.T he demandf or leisure,i ncludings ex,i s an uncomplicatedfu nction
of wealth, becauset he cost of time, and henceo f sex, does not rise with
wealth,a si t doesw ith earnedin come.A personw ho possessewse alth,p erhaps
throughi nheritanceo r paste xertion,n eedn ot haveh igh hourly earningsa, nd
thereforen eedn ot incur high time costst o engagein sex.( Converselya, very
poor person may lack the health and energy required to be sexually active.)
Since carrying on an adulterous liaison takes more time than patronizing
prostiruteso r obtaining sex from one's spousew, e can expectw ealthy aristocratic
men to choosea dultery concubinagea, nd simultaneouso r serial
polygyny over monogamy and prostitution, and bourgeois men to choose
monogamy and (or) prostitution over adultery.
Other factors as well lead one to expects ystematicd ifferencesin sexual
behavior between wealthy and nonwealthy people-female as well as male.
Noncompanionatem arriagep, articularlyt he form designedto cemenpt olitical
allianceso r to shoreu p decliningf ortunes,is more desirableto wealthyp eople
and aristocrats( who are more likely to be wealthyt han commonersa re)t han
to middle-classp eople-and noncompanionatem arriagef osterse xtramarital
intercourse.JM2 oreover, a wealthy man is more aftractivet o a variery of
potential sex partners than an ordinary man is, so his costs of sexual search
51. Staffan Burenstam Linder, The Haried Leisure Class 83-89 (1970). A significant quali"
fications houldb c noted:a personw ith extremelylo w wagesm ay lack the wherewithalt o conduct
an activc scx life. The general thesis of Linder's book-that modem people havc less leisure,
bccauseth cir cost of time is high-is challengcdw, ith spcci6cr efercnccto sleep,i n Jcff E. Biddlc
and Daniel S. Hamcrmcsh, "Sleep and thc Allocation of Timc,. 98 loumal of political Economy
922,939 (1990).T hcy point out that sleepm ay yieldu tiliry,i n whiche ventp coplcm ay .buy"
morc of it (by giving up remunerativea ctivitiest hat would cut into their slecp)a s their incomes
rise even though the 'price" of slecp (in altcrnative activities forgonc) is rising too. ln thc text
belowI considctrh e parallclp ossibiliriyn regardt o scx.
52. "Vhat was decisive [in arisrocratic marriagc] was the intcrest of a family, and bcyond
that-thc class.w c would shivcra linle at the coldnesss, cvcriry,a nd calculatingc larity of such
a noblc concepto { marriagca s has ruled in evcry healthya ristocracy. . . preciselyt his is why
love as a passion-the great meaning of the word-was invented f,ot thc atistocratic world and
in it, whcrc constraint and privation were greatcst." Fricdrich Nictzsche, The Wit! to power
$732,a t 388 (WalterK aufmanna nd R. J. Hollingdalcrr ans.1 958).
Sex and Rationality 1 3 5
are actually lower eveni f his time is varuable.A nd an aristocraticw oman,
insofar as she has her own wealth, is lessd ependenot n her husbandt han a
middle-classw ife and thereforel essp rone to comply with his desiret hat she
bec haste.'The apparentp roclivityo f upper-crasislo .rn *o..n for adultery
may have beenr elatedt o the securew eaith thesew omenp ossessebdy virtue
of Roman law governingd owry and inheritancee, specialiys incet his'wealth
was not an accumulationo f the woman's" earned' incomea nd henced id not
imply a high valueo f the woman'st ime and thereforea high costo r "-orou,
liaisons.sa
. This analysism ay explains omeo f the tensionb etweent he middlec lassa nd
the aristocracyt hat explodedi n the EnglishR evolutiono f 1640,p uritanism
is a middle-classm oral stancer ather than an aristocratico ne. The fact that
thc incomeo f the middle classw as rising in the seventeentche nruryw ould,
if my analysis is correct, tend to increas. ih. dirt"nc. between ,r-r. *. J"rr.,
in anirudes.towards ex, becauseth e rise in middle-classin comes* .dd, l;
increasingth e value of time for middle-clasms en, reinforcet heir puritanii
values.
We can extendt he analysis.tot he presenta nd predictt hat young lawyers
at high-pressurlea w firms will havei ewer love af?airsa, nd commit"adultef
lessf requently,t han personsw ho possessin heritedw ealth but have lower
oppomrniryc ostso f time. More broadly,t he ambianceo f a highlyc ompetitive
capitalist society will be less romantic, less given to sexual iritriiue, ffi;;i;
courting, and elaborate flirtation, than that of a sociery in wiict "u.."g.
hourly wa.gcs are low, especially if wealth is less equalry distributed in tie
static,t raditional societyt han in-thed ynamicc apitalisto ne.T his point bears
on the question,e xplored in chapter 7, whether sexual moraritryh as anv
macroeconomic effects.
It would be wrong to conclude,h owever,t hat as incomesi n a societyr ise,
sex- must inevitably become less frequent or more hurried. Apart from the
earlier point about leisure being a superior good, and the faci that s.* is "
positivef unction of health and that standardso r health rt"u. ro,p-u.J, *.
must not ignore the fall in the costs( unrelatedr o time) of sex that has been
broughta bout by improvementsin contraceptionW. e mustn ot rgnoreq uality
either.R ising earnedi ncomesc auser he demandf or time-intens-iug.o od, ,o
fall if qualiry remains the same; but qualiry may not stay the ."-.,"b.."rrr.
the demand for qualiry is itself a positive iun..lon of income. Wl,; ;;;;i;
53' cf' Julian Pitt'fuvcrs, The Fateo f schechemo r The poriticso f sex: Essayisn tbe Antbro- pology of tbe Meditctrancan 4S4G ,19771.
54. Bcryl Rawson, 'Thc Roman Family," in The Family in Arrcient Romc: New perspectives
1, 19, 27 (Rawsonc d. 1985);J . A. crook, "womcn in ioman Successioni,n" The F)mity in
Ancient Rornca t 58,68-69; Suzalne Dixon, "Family FinancesT: ercntiaa nd Tullia," in The
Familyi n Ancient Rome* 93; SarehR.p omcroy,G odiessesW, horcs,W iuesa, nd Slauei,.Womea
in Cbssical Antiquity t63, lgt-tl} lt97S\.
t 3 6 A THEORY OF S E X U A L I T Y
earn more, they do not iust buy more cars' they buy better cars. Consider the
caseo f children.T he rearingo f childreni s dme-intensives'o as incomesr isc,
the averagen umber of childrenf alls. But the qualiry of childrenr ises.T hat
is, parents invest more in the care and education of their children. We might
similarly expect that rising incomes would, other things being equal, reduce
the amount of sexuala ctiyity in a socieryb ut increaseit s qualiry.T his may
be an important causeo f the growth of companionatem arriage.
Tbe Effectiue Sex Rztio
Implicit in my emphasis on the costs of search is the importance of rhe effectivc
sex ratio in explaining sexual customs, The sex ratio is the ratio of males to
femalesin the population.B y "effective' sex ratio I meant he ratio of males
to auailable females. The higher the effective sex ratio, the higher the costs to
men of heterosexuals ex. Searchc osts are higher becausei t takes longcr for
each man to find a women who is not already spoken for; the analogy is to
fishing in a pond that has very few fish. Moreover, the price of each woman
is bid up because there are more demanders, and if price is not used to clear
the market, we can expect queuing, illustrated by the long period of bachelorhood
for most men in polygamouss ocieties.
The difference between the pure sex ratio and the effective sex ratio can be
dramatic. The sex rario may be 1, but if half the women are the wives of one
man, the effective sex ratio is 2, and there will be a tendency toward both
desperate competition among men for wives and rebellion against the polygamist.
That compctition can be muted, however, and the stability of the
polygamist'sp osition enhancedb, y the encouragemenotr at leastt oleratiorr
of prostitution and homosexuality, which provide safery valves for unmarried
men.ri One can predict, therefore, that these practices, and (for a reason about
to be noted) perhapse speciallyh omosexualiryw ill be more common,o ther
things being equal, the more polygamy rhere is in a sociery. This hypothesis
can be testedw ith data collectedm any yearsa go by ClellanF ord and Frank
Beacho n sexualp ractices,m ainly in primitive societies.rI5n 53 of the 83
55. That a high ratio of malcst o fcmalcsc ncouragcms alch omoscxuabl chaviorh asf rcquently
bccn notcd. Scc,f or cxamplc,J . M. Carricr, "HomoscxualB ehaviori n Cross-CulturaPl crspcctivc,"
in Homoscxual Behauior: A Modem Reapprakal 100 (Judd Marmor cd. 1980). (Onc is
thcreforc not surpriscd to discover that ancient Grcccc and Rome appear to have had a large and
pcrsistcnt surplus of malcs. Pomeroy, notc 54 abovc, at 154-155, 227-228.) On thc tolcration
of such bchavior by old mcn cager ro prcscrve polygyny, scc Barry D. Adams, 'Age, Strucnrrc,
and Sexuality:R cflectionso n thc AnthropologicalE vidcncco n HomoscxualR clations,"i n,{zthropology
and Homosexysl Behauior 19, 21-22 (Evclyn Blackwood ed. 1985). And on thc
dcmand for prostitution as a function of thc sex ratio, scc Dcckcr, notc 39 abovc, et 72.
56. ClcllanS .F orda ndF rankA . RcachP, anenso f SexusBl ehavio1r 29n . 1, 130n ,2,268-
292 (1951).1 hevc supplemcntcda nd in placcsc orrccrcdF ord and Bcach'sd ata with data hom
Sex and Rationality 137
;
l
I
I
societiesfo r which the requisited ata are reported-almost two-thirds-male
homosexuala ctivity is commono r approved.B ut the percentageis lowest in
the 10 societiesth at are strictly monogamous( 50 percent)'h igher in the 30
societiesin which polygamyi s rare (50 percent)a, nd highestin the 43 societies
in which polygamy is common (70 percent). The order is as predicted, although
a richer study would use finer distinctionsa mong degreeso f both
polygamya nd tolerancef or homosexualitya nd would also control for other
poti"tial explanatory variables such as the sex ratio and the difference in
"u.."g. age of spousesa t marriage( a proxy for whether marriagei s companionate
or noncompanionate).
One would also expect-and therei s evidenceo f this as well-that polygamoush
ouseholdsa, nd more generallya ny seftingi n which women are Persistentlyd
eprivedo f male companionshipw, ould foster oppornrnisticl esbianism,
s inci a man with plural wivesi s likely to neglects omeo f them's7B ut
the toial amount of opporonistic lesbianismin a polygamouss ocietym ight
well be lesst han in a -ottog"tnon. one, sincet herea re fewer spinstersin the
former and the averagea geo f marriagef or women is lower.T his point helps
explain,i ncidentally*, h1 fot men homosexualirym ay be the preferreds afety
u"i". to prostitutioni n a polygynouss ocietyP. rostitutesm ay bev erye xpensive
in such a society (unlessi t is a slave socierya nd marriagew ith slaves_i s
forbidden).B y increasingt he demandf or femalesa nd therebyr educingt he
average age of women at marriage and the fraction of women who are
unmairied,p olygynyr educesth e supplyo f womenf or prostitution.T his.can
be expectcd to diivc up the price and thus make substitutes more attractive.
we should not take a sex ratio of 1 for granted. Female infanticide has
beenc ommon in many societiesa, nd has somedmesle d to a sharplys kewed
sexr atio,ssw hich in turn skewst he effectives exr atio. In Americat oday black
women significantlyo utnumbera vailableb lack men.T his is partly becauseo f
thc Human Rclations Arca Filcs. Thc findings rcponed in thc tcxt should bc taken with more
than a grain of salt, not only becauseo f thc failuret o correctf or otherv ariablesb, ut alsob ccausc
thc cthiogaphic data frcqucntlyd isagrcea boutt hc properc haractcrizatioonf a socicty'sa ttitudc
toward h-omosexualitya nd (or) polygamy.c ompare with Ford and Beach'sc haractcrizations
JohnV . M. Vhiting, ;Effcctso f Climatco n CcrtainC ulturalP racticcs," in Cultural Anthropology
"stt
(wr.d H. Goo-Coai, on Twcnry Scxuai Anitudcs and Practiccs," 75 Etbnology: An Intertational Joumal of
Cuhsra! and Social Anthtopology 409 11976).
57. Evclyn Blackwood,; Bt."king thc Mirror: The Constructior of Lesbianisma nd the Anthropological
Discoursc on Homoscxuality," in Anthropology and Homosexual Bchauior' rrotc
5-55 a8b. ovc,a t 1, 11-13.
Se;, for cxamplc, Lloyd deMause, "The Evolution of Childhood," in Tbe History of
chitdhood 7,25-29 (dcMausce d, 1974).1discussin fanticidci n thc last scctiono f this chaptcr.
whcnlsay"skewed,"Isimplymcanmorcorlessthanl'Inasocia|scnsc,asocictythatpracticcs
female infanticidc regards a sex ratio of 1 as "skcwcd'"
1 3 8 A THEORY OF S E X U A L I T Y
higher fetal and infant mortality of black than of white males, which is due
in turn to the interaction of the tendency for pregnancy and neonatal care to
be poor among blacks with the fact that male infants are more vulnerable
than femaleo nes; and partly becauseo f the high death rate of black males
from homicide and their high rate of imprisonment.5Ien addition, white
women are more likely to date black men (and vice versa) than white men
are to date black women.5OE conomica nalysisp redictst hat with so favorable
an effective sex ratio, black men (outside of prison) will be less likely than
white men to engagei n opponunistich omosexuabl ehavioro r to patronize
prostitutesb ut more likely to havem ultiple sex paftners,t o be initiatedi nto
sex early,a nd to father illegitimatec hildren,t endencietsh at make black men
seemm ore promiscuoust han white men.51In like vein, Mormon polygamy
has beene xplaineda s a responseto a surpluso f womeno ver men,c ausedb y
the absenceo f Mormon men on missionaryw ork.52
59. The sexr atio for blacksb crwecnth c agcso f 25 and4 4 is 0.87;f or whitcsi t is 1.01.U .S.
Dept. of CommerccB, ureauo f thc CensusS, totisticaAl bsttacto f the UnitedS tates1 987 17
(1985)( tab.1 8).M uch of thc diffcrencet,o be surc,m ayb c duc to thc undcrcountinogf young
black malcs.K ristin A. Moorc, Margarct C. Simms,a nd CharlcsL . BetscyC, hoicea nd Circunstance:
R acialD ifferencesin AdolescenSt exualitya nd Fenility 172-ll4 (1985),B ut the scx ratio
is only one determinanto f the effectiues ex ratio. More than 3 perccnto f all blick malesa re in
jail or prison,w hilc the correspondingp crccnragco f black fcmalcsi s prcsumablym uch smaller,
sincc fcwcr than 5 percent of statc prisoncrs and fewcr than 10 percent of jail inmates are (cmalc,
U.S. Dcpt. of Commcrcc, Bureau of the Ccnsus, Statktical Abs*act of tbe United States 1990
187 (1989)( tabs.3 23, 325).M ost prisoncrsm, oreovcra, rc youngrf or cxamplc,m orc than 90
percent of statc prisoners are berween 18 and 44 (morc than 50 perccnt betwccn 25 and 44); sec
187 (tab.3 25). For thcsea nd other reasonst,h crc is linlc doubt that thc cffcctivcs cx ratio among
Amcricanb lacksis far belowu nity.M oorc, Simmsa, ndB etscya t 112-114;O sci-MensaAhb orampah,
"Black Malc-FemalcR elationshipsS: omc Obscrvations,"1 9 Jouraal of Black Studies
3Z0,3Zl-322 (1989);G unentaga ndS ccordn, otc 36 above,c h, 8.
60. Manin S. Weinberg and Colin J. Williams, "Black Scxualiryr A Tcst of Two Theorics,"
25 toumal of Sex Research1 97, Zl4 (1988), and sourccsc ited there. \,)9hcrea6s3 perccnto f
black-whitcm arriagesin volvc a black male and a whitc female,o nly 37 pcrcenti nvolve a white
malea nd a blackf cmale.G uttennSa ndS ecordn, otc3 5 abovc,a t 225 (tab.8 .8) (1970d ata).
61. I have no cvidence about opportunistic homosexuality among blacks or about black
patronagc of prostitutes. Thc othcr behavioral diffcrcnccs arc wcll documcnted, howevcr, and
thcy pcrsist cvcn aftcr such variables as cducation and incomc arc controllcd for. Guncntag and
Sccord, notc 36 abovc, at 275-220i Moorc, Simms, and Betscn notc 59 abovc; lfeinbcrg and
Villiams, notc 60 above; Sandra L. Hoffcrth, 'Recent Trcnds in thc Living Arrangcmcnts of
Children;A Cohort Life Tablc Analysis,"i a FamilyD emograpby:M ethodsa nd Theit Application
158 (John Bongaarts, Thomas K. Burch, and Kcnncth W. Wachter cds. 1987); Thomas J.
Espcnshade",T hc Rcccnt Dccline in Amcrican Marriagc: Blacks and Whitcs in Comparativc
Pcrspcctivc,"i n ContemporaryM ariage: ComparstiueP erspectiueosn a Changing lnstitution
53 (KingslcyD avis ed. 1985); Ira L. Rciss," PremaritalS exualP crmissivcncsasm ong Negrocs
and Whitcs," 29 Americat SociologicaRl euiew6 88 (79541,
52. Villiam l-awrenceF oster,'B ctwcenT wo Worlds:T he Originso f ShakcrC clibacy,O neida
Communiry Complcx Maruiage, and Mormon Polygamy' 223-224 (Ph.D diss., Univcrsiry of
Chicago1, 975).
Sex and Rationaliht 1 3 9
Two other hypothesesg eneratedb y my theory are that black men commit
fewer heterosexual rapes than white men, after allowance is made for other
variablest hat explain differencesin crime rates,b ecauseb lacksh ave lower
costso f heterosexuasl earcht han whites; and that black men are lessp rone
to abusec hildren sexuallyt han white men are. The first hypothesisc an be
testedw ith the aid of IsaacE hrlich'sr egressionosf crime rateso n a variety
of potentially explanatoryv ariables,i ncludingr ace.I t is not supponed:t he
coefficient of the nonwhite variable is positive for rape, as it is for all the
crimcsi n Ehrlich'ss tudy; it is, however,m uch smalleri n the caseo f rapet han
in the caseo f any other crime againstt he person.6T3 he secondh ypothesis
(black men commit less child sexual abuse than white men) is supported by
data showingt hat, at all incomel evels,b lacksa re indeedl essl ikely to abuse
children sexually than whites are.6a
63, "Participationi n IllcgitimatcA ctivitics:A n EconomicA nalysis,"i n Essaysin tbe Economics
of Cimc and Punishment 68, 96-97, 100-101 (Gary S. Becker and Villiam M. Landes cds.
1974) (tabs.2 -5). For cxamplc,i n onc of thc rcgrcssiontsh e coefficienfto r rapc is 0.055 and for
murder 0.542 (sce 101 tab. 5-thc data, incidcntally, arc for 1960), implying that a 1 pcrcent
incrcasc in thc numbcr of blacks would rcsult in an incrcasc of more than half of 1 pcrccnt in
the numbcr of murders but only about one-fiftccntho f 1 percenti n the numbero f rapcs.T he
crimesa gainstt hc pcrson in Ehrlich'ss tudy arc, besidcsm urdcr and rapc, assaulta nd robbcry.
Ehrlich classifiesro bbery as a crimc againstp ropcrry,l ikc larccnyo r auto thcft, the other rwo
crimcs against propcrty in his study, rather than as a crime against the person, likc murder,
assault,a nd rapc. But acruallyi t is thc lancr. Robbcryi s taking propertyf rom a personb y thrcat
or usco f forcc. Thc cocfficicntso n thc nonwhitcv ariablca rc low in thc casco f larcenya nd auto
thcft.
The qualificationin thc text-"after allowanccis madcf or otherv ariablcsth at cxplaind ifferences
in crime ratcs"-is very imponant. Without such correction, the rape rate is far higher
amongb lackst hana mongw hitcs.l n 1988o net hird of all personsc onvictedo f rapew ercb lack,
although blacks wcre only 11 pcrccnt of thc adult U.S. population. Patrick A. Langan and John
M. Dawson, "Fclony Sentcncesin StatcC ourts, 1988" 4 (Burcauo f JusticcS tatisticsB ullctin,
U.S.D cpt. of Justicc,D ccember1 990) (tab. 5). (The pcrcentagco f blacksc onvictcdw as higher
for alJ thc offenscsin the study,r angingu p to 53 percenti n the casco ( robbery.)I llustrativeo f
the needt o correctf or otherv ariablesis the fact that rapei s oftena n incident o a nonsexual
crimc such as robbery or burg.laryD. iana Sculln UnderstandingS exual Violezce:A Studyo f
Conuictcd Rapists 147-142 (1990). This implics that sincc blacks commit a disproportionatc
number of such crimes, their costs of raping arc lowcr than the cost to whites. Another cxamplc
of the nccd to distinguish bctwccn raw and intcrprctcd data is that, despite my conjccnrrc that
opportunistich omoscxualityis lcssc ommona mongb lack than amongw hitc malcs,t hc incidcncc
of AIDS conractcd throughh omosexuacl ontacti s twicc ash igh pcr capitaa mongb lackA merican
thana mongw hiteA mericanm alcs". Centcrsfo r DiseasCe ontrolS tatistics4," A IDS 1307( 1990).
But this is beforc corrcctingf or other factorsb caringo n the incidenceo f homoscxual-induccd
AIDS, such as knowledgea bout thc discasc,w hich may wcll be bclow avcragci n thc black
communityB. eckera ndJ osephn, otc6 abovc,a t 405-405.
54. U.S.D ept. of Hcalth and Human ScrvicesO, f6ceo f Human DevclopmcnSt crviccs',S tudy
Findings; National Srudy of thc Incidencc and Scverity of Child Abuse and Neglcct" 28-29 and
69,6 (DHHS Publicationn o. (OHDS)8 1-30325,S cptcmbc1r 981).T his studyr ccognizctsh c
possibilityt hat child abusci s lcssf rcquentlyr cponcd to the authoritiesb y blackst han by whitcs
1 4 0 A THEORY OF S E X U A L I T Y
Marcia Guttentag and Paul Secord's book on sex ratios (note 35) is an
ambitiouse ffort to derivet he implicationso f an imbalancei n the sex ratio
for sexualm ores.5B5 ut it is weakenedb y the authors'l ack of a cogentt heory
of human behavior.T hey speculateth at lesbianismw ill be more commont he
lower the ratio of malest o femalesa nd rhat menw ill tendt o sequestewr omen
more rhe highert he ratio of malest o femalesT. he ideai s that excessd emand
of men for *o-ttt puts female chastiry at greater risk by raising the- "price"
that unmarried men will pay for accessto women, married or otherwise.
Thesec onjecturesa re congenialt o economica nalysis,b ut not the authors'
overarching thesis that an imbalance in the sex ratio dict^tes the character of
a society'sie *ual mores-repressiveo r at leasts traitlacedif therei s a shortage
of women, liberal or permissiveif therei s a shortageo f men,T he sex ratio is
only one determinanto f the relevantc ostsa nd benefitsG. ucentaga nd secord
would be forced by the logic of their analysis to predict that the ancient
Greeks and Orthodox Jews-both cul$res with a persistent surplus of
males66-had similar sexual mores, which they did not.
Nevenheless the sex ratio, in particular the effective sex ratio' is highly
relevantt o the economica nalysiso f sexa nd is especiallvp ertinentt o explaining
differences berween white and black sexual behavior that remain after
coirection for differences in income, education, and other nonracial variables.
Besidesth e higher rate of black illegitimateb inhs, the lower averagea geo f
sexuali nidation of blacks,t he highera veragen umbero f sexualp artnersP er
black male, and the lower incidenceo f child sexuala buseb y blacks,b lack
men are far less likely than white men to undergo voluntary sterilization or
to use condoms; blacks are less likely to seek treatment for fenility Problems
than whitesa re; and black men arem ore disapprovingo f homosexualityth an
white men are,57S inceb lack men are scarcea nd thereforeh ave the rrpper
but concludcst hat corrccting for this d.iffcrcnccw ould not aflcct thc study's conclusion.T he
-p ossibilityt hat rape is undcrrcponedi n black communiticsm ust also bc acknowlcdgcd'
5J. Scca lso Moorc, Simms,a nd Betscy,n otc 59 above,a t 136; David M' Hecr and Amyra
Grossbard-schcctrnan"T, hc lmpact of thc FcmaleM arriagc Squeezea nd the contraccptive
Rcvolution on Scx Roles and thc Women's Libcration Movcmcnt in thc United Statcs, 1960 to
1975,' 43 Jouraal of Maniage ond the Family 49 (1981); Grossbard-Schcctman' "Marriagc
Squcczcasn d thc Marriagc Markct," i, ContemporaryM aniage, notc 51 abovc,a t 375'
66. On Grcccc.s ccr cfcrcnccsin notc 54 ebovc'O n Judaism,s ecG uncntaga nd Sccord,n otc
36 abovc, at 42 and ch. 4.
57. Elizabcth Hervcy Stcphcn, Ronald R. fundfuss, and Frank D. Bcan, "Racial Diffcrcnccs
in Contraccptivc Choicc: Complcxity and lmplications," 25 Demogtaphy 53 (1988)' csP' 67r
l:rry L. Bumpass,'T hc Risk of en UnwantcdB inh: The ChangingC ontext of Contra,ceptirc
Stcrilizationi n the U.S.," 4l Population Studies3 47,350 (1987); Debra S. Kalmuss, "The Usc
of Infcnility Scrviccsa mongF cnility-ImpairedC ouples,"2 4 Demography5 75' 582-583 (1987)i
Grcgory M, Hcrck and Eric K. Glunt, 'AIDS-Rclated Anitudcs in thc unitcd statcs: A Prcliminary
Conicpoalization, ' 28 loumal of Sex Research9 9, 111 (1991); Harlon L. Dalton' "AIDS in
Sex and Rationality t 4 l
e
c
c
hand in bargaining with black women, they can shift the burden of contraception
to them. since fertility problems more commonly afflict women than
men, a man who has many women to choose among may decide to abandon
a woman with whom he is failing to havec hildrenr ather than cooperatei n
seekinga solutiont o the problem;i ndeed,a traditionalr ationalef or iolygyny
is that it is a way of gening around female fertiliry problems withoui di;..;
and remarriage'A nd finally,t he needf or a homosexuasl afetyv alve is minimized
in a- communiry in which the ratio of men to women is low, so
homosexualityis more likely to seemu nnaturali n sucha communiw.
Rcligion and Education
A.powerful variablee xplainingd ifferenceisn sexuala ftirudesis , as I noted in
chapter 2, religiosiry-though only in societiesin which the iominant religions
are_h ostile to particular sexualp ractices,a s christianiry (particularly
Roman catholicism) and to a lesser extent Judaism and Islam "... p..ron,
who.b elieveth at they faces upernatura,l" .r.iionr, or eveni ust that they are
displeasing.Godb, y engagingi n particulars exualp racticesa re lessr ikeiy to
engagein them, or at leastt o expressa pprovalo f ihem, than the irreligious.
This may be one reasonw hy better-educatepde ople are more tolerant in
sexual matters than less-educatepde ople are, for educatedp eople in our
society are less likely to be religious.58
_ Another reason may be that educated people tend to be more aware of
human sexual variety. They have heard of Freud and Margaret Mead, and
maybe even of Malinowski. A person who knows that Jamesi, Francis Bacon,
Oscar Wilde, Henry James, Marcel proust, Gertrude Stein, Virginia Vootf,
John Maynard Keynes, E. M. Forsrer, pyotr llich Tchaikoviky,
"G.o.g.
S.n_
tayana, T. E, Lawrencc, Alan Turing, and Ludwig Wittgenstein *...1orno_
sexuals,a-n d that SophoclesS, ocratesp, lato, ShakespeareC,h ristopherM ar_
lowe, Alexander the Great, Julius caesar, and Richaid the Lion-Hearted mav
have beeq is not so likely to believet hat homosexualiryis merelya ghastly
blight.' Then too, as this list suggests, a number of tire culrure hetes of
intellecnralsh ave beenh omosexual-but they are not heroest o nonintellectuals.(
For the nonintellecnralh, owever,t herea re,a mongm any other homo_
Bl1:\fecc," 718 Daedalus:J oumal of the AmericanA cademyo f Arts and Sciences2 05,213 ( 1989).
68. Scc,f or cxamplc,G allupR eportn o. 259,A pril 1,987,ll,7+.
59. Somc of tlesc illustrious figures, howcvcr, may havc bcen opponunistic homoscxuals
(hcncc" rcally" hcterosexualo)r, genuincb isexualsr,a thert han "real" homosexualHs.o mosexual-
rightsa dvocatesa rc apt to labcl famousp ersonsh omoscxuaro n scantyc vidence,b ut the
Enqclogelia of Homosexuality (Vaync R. Dyncs cd. 1990) revicws thc cvidcncc pro and con
thcsc labclings with appropriate caution. See under panicular names.
142 A THEORY OF S E X U A L I T Y
sexuals who might be mentioned, William Tilden, Cole Porter' and Rock
Hudson.)D ifferJncesb erweene ducateda nd uneducatedp eoplee xist at the
levelo f sexualb ehaviora s well as of attitudes.F or example,e ducatedp eople
haveb eens aid ro masturbater noret han the uneducatedb ecauseth eir imaginative
resourcesa re greater,m aking fantasys ex a better substitutef or real
sex.70
Is religion an ultimate variable,o r, as this brief discussiono f education
,ugg.rrr"-ry be possible,is it epiphenomenalA?s we saw in Chapter1 , the
baiis in the i.*t oi th. Old Testamenat nd of the Gospelsfo r the characteristic
Christiana ttitudest oward sexi s infirm. Sexw as not a subiecti n the forefront
of concerno f either the Hebrew prophetso r Jesus.T here are the Levitical
prohibitionsa gainsth omosexualirya nd adultery,o f course,b ut sinceC hrisiianiry
abandoiedm anyo f the other Leviticalp rohibitions,t he onesi t retained
could not have been retained under the compulsion of textual authority. of
course,t ext is not everythingT. extuals ilencesc an be pregnant'F rom the fact
that the Ten commandments do not forbid incest we should not infer that
the ancient Jews condoned the practice, iust as we should not infer that by
failing to specifyt he sex of the spousest he authors of a marriages tatute
inteni to authotir. homosexuaml arriage.N everthelessth, e Christianv iew of
sexw asi nvented-the decisiont o emphasizJee susc hrist's (apparentc) elibacy,
for example,w as just that, a decision-and it is possibleth at the endsi n view
were practical. what those ends might be is a difficult question for. economic
"n"lyrir. I do not pretendt o have a completea nswer'b ut somee lementso f
one, revolving around the role of the Church in promoting companionace
marriage,w ill emergei n the next chapter.
Complementariryo f SexualP ractices
one good is said to be a complemenot f anotheri f a fall in the price of the
firrt i-nc.e"re,t he quantity demandedo f the second.A n exampleo f complementaryg
oodsi n t1:,.. r." of sex is contracepdona nd vaginali ntercourse..If
we think;f the contraceptive.. product" not as the physicalg ood but as_the
serviceo f preventingb irih, therebya diustingf or improvementsin the effec-
70. Morse Peckham, Art and Pornography: An Experiment in Expllnation 155,231 (1969)'
(compare the discussiono { pornographyb elow.)A consistenbt ut bencr-documcntcdfi nding is
that middlc-clasasd olescent.s"r *rb"r. moret han lowcr-clasasd olescentSs'e e,b esidesre fcrencesi
n Chaptcr2 , Roben J. Havighurst, .cultural Factorsi n Sex Expression,i"n sexual
Behauior in A^erican Society: An Appraisat of the First Two Kinsey Reporrs 191 (Jeromc
Himelhocha nd SylviaF leisF avae ds. 1955). In addition to educationadl ilferencesb ctwecnt hc
two gloups, thc """rage agco f first intcrcourscis lower in thc lower class'F inally,o n the impact
of eJu."tion on the effectiveu se of contraceptions, eeR oben T. Michael, 'Education and the
DerivcdD cmand{ or children,. 87 Journalo f PotiticaEl conomys 128,S 140-s161( 1973).
ock
the
rple
aE-
:eal
ion
the
stic
ont
rcal
ns-
:ed
of
tact
hat
by
ute
' o f
the
rle-
. I f
rhe
ec-
ICY,
ew
nic
of
ate
Sex and Rationality 1 4 3
tiveness as well as in the comfort and the nominal price of contraceptive
devicesw, e will quickly realizet hat the cost of contraceptionh as fallen by
staggeringa mounts.T hat is why one would expect,o ther thingsb einge qual,
the amount of vaginal intercourse to have risen-as it has (see note 10).
Considerationso f complementarityth us enablea dditionalp redictionst o be
madec oncerningt he effecto f socialc hangeo r differenceo n sexualp ractices.
Abortion might seemt o be a substitutef or contraceptiona, nd it is, but it
is alsoa complemento it. By backingu p contraceptiona, bortionr educesth e
costso f engagingi n vaginal intercourseI.f abortion is made very costly by
legal regulations,t hen, since contraceptioni s not 100 percente ffective,a
couplem ay be inducedt o substitutes omem ethodo f nonvaginali ntercourse
(sucha s anal intercourse)o, r event o abstain.C onverselyif, abortioni s cheap,
vaginali ntercoursew ill be more frequenta nd, dependingo n people'sk nowledgeo
f and the efficacyo f contraceptivem ethods,m ay generatem ore unwantedp
regnanciesn,o t all of which will be aborted.T his shouldh elp us to
understandth e combinationo f cheapc ontraceptivesfr,e quenta bortions,a nd
yet a high rate of unwanted births in our society, although a more refined
analysisw ill be presentedin later chapters.
A relatedr easonf or believingt hat the number of abortionso verstatesth e
number of net live births lost as a result of abortion is that when used as a
methodo f family planningi n the senseo f planning/ or a family,a bortion may
affectt he timing more than the numbero f births.S incei,n affectingt he timing
of births,a bortionc ouldr esulti n an increasein the averageq ualiryo f children,
and since, as Chapter 7 will point out, modern people have a tendency
(fosteredb y the declinei n infant and child mortaliry)t o substituteq uality for
quantity as a desiredc haracteristicin children,w e can understandw hy the
force of the moral objectionst o abortion has weakenedi n recentd ecades.
"The apparently paradoxical idea that birth control could be encouraged
through an increasedc oncernf or children"7lm akese conomrcs ense,
Like abortion, infanticide, when viewed, as it should be, as a method of
famify planning 72r ather than a specieso f motivelessm alignancy,d oes not
reduce the population by the number of infants killed. For in a poor sociery,
the fewer children a woman has, the likelier they are to survive to adulthood.T3
This is true even when most or even all infants killed are girls, the "ef6cient"
71. Andr| Burguiire, "From Malthus to Max Webcr: Belated Marriage and the Spirit of
Entcrprisc," in Family and Society: Selections from the Annales Economies, Soci4tiis, Civilisations
237. 239 (Robcn Forster and Orest Ranum eds. 1975).
72. Clearly the dominant forrn of infanticidei n the human speciesS. ccs tudiesa nd refercnces
in lnfanticide: Comparatiue and Euolutionary Pelspectives, pt. 4 (Glcnn Hausfater and Sarah
Blaffer Hrdy cds. 1984).
73. For evidence from Japan, sec Susan B. Hanlcy, 'The Influencc of Economic and Social
Variableso n Marriagc and Fcniliry in Eightcentha nd NineteenthC enturyJ apaneseV illages,"i n
Pobuldtion Patteras in the Past 165,176, 199 (Ronald Demos Lee ed. 1977).
t9).
g l s
fcrual
rmC
thc
act
tnc
t 4 4 A THEORY OF S E X U A L I T Y
form of infanticideb ecauseit limits the furure growrh of the population.T.
(Reducing the number of males need not reduce the number of children in
the next generation,s inceo ne man can fertilizea n indefinitelyl argen umber
of women.) In a sociery in which women need dowries in order to marry, a
couplem ay havet o chooseb erweenk illing one or more of its daughtersa nd
having few or even no grandchildrenb ecauset he cost of raising all the
daughtersw ould preventt he family from doweringa ny of them. Hence-not
today, of course, but in the radically different social conditions prevailing in
earliers ocieties-a practiceo f femalei nfanticiden eedn o more bespeakh ostiliry
toward or a disvaluing of women than the thinning of trees in a forest
signifiesa dislike of trees.TIrt would be irrational from a genetics tandpoint
for a father to be indifferent to the procreative potential, and hence to the
survival, of his daughters.
While contraceptiona nd abortion are complementsa s well as substitutes,
infanticidea nd abortion are substitutesA. dvocateso f abortion rights do not
like to view the relationb etweenin fanticidea nd abonion in this light, because
infanticidei s abhorredi n our society.B ut the relationshipi s unmistakable,
both being methods of preventing unwanted births after conception has occurred.
If this is so, however, the question arises why the crueler methodinfanticide-
was until recently more common than abortion. The obvious
answeri s that until recentlya bonion was a dangerousp rocedurefo r a woman
to undergo,b ut this answeri s incompleteb ecauseu ntil recentlyc hildbinh
was also a dangerous procedure. The subtle answer is that if only female
infantsa re unwanteda nd the sex of the fetusc annotb e determineda, borrion
will kill on average twice as many infants as infanticide will, and onty half
those infants will be unwanted.
The interplayo f substitutabilirya nd complementarityis further illustrated
by the relation betweenc ontraceptiona nd prostitution.O n the one hand, by
reducing the probability that the prostitute will become pregnant, effective
contraception reduces the cost of prostitution and therefore the price. On the
other hand, by reducingt he costo f marriage,c ontraceptione ncourageesa rly
marriage,w hich reducesth e demandf or prostitutionb y providingm ore men
with an alternative source of sexual gratification. Pornography is another
good whoser elationt o sexc ombinese lementso f substitutiona nd of complementariry.
When used in coniunction with masturbation to create a closer
fantasys ubstitutef or sexuali ntercourseth an masturbationa lonew ould do,
74. Susan C. M. Scrimshaw, "lnfandcide in Human Populations: Societal and lndividual
Conccrns," in Infanticide, notc 72 above, at 439, 454.
75. The Japancsc word for infanticide, mabiki, is the samc word used to denotc thc thinninq
out of plants. Hanley, note 73 above, ar 176.
,71
in
er
rd
le
in
sst
rt
le
s,
)t
ie
:
ts
n
h
e
n
tf
Sex and Rationality 1 4 5
pornography operates as a substitute for intercourse and reduces the demand
for it.76 But when used to stimulate sexual desire, pornography may have the
effecto f increasingth e amount of intercourseT. he balanceo f thesee ffectsi s
the focal point of the debate over the role of pornography in rape, discussed
in Chapter 13.
75. ln this vcin it has bccn argued that as pomography, formcrly marketcd only to men,
bccomcsm arketcd morc and morc to women,w omcn will masturbatcm ore, lcadingi n rurn to
a grcater locus by women upon genital sexuality as distinct from tlrc rather morc diffusc scxual
interestt hat has beenc haracteristico f Vcstern women.J ohn H. Gagnon," The Interactiono f
GcnderR olcs and SexualC onduct,' in Human Sexuality:A Coftpdrdtiuca nd Dcvclopmmtdl
Pecp.ctiue 225, 242 (Hcrant A. Katchadourian cd. 79791. This is a typically consrructjonist
hypothcsisT. hc sociobiologisti,n contrast,d oesn ot cxpcct pomography,i n the scnsco f crotic
litcrarure that emphasizcsth c depiction of scxual organs,c ycr to find a large markct among
womcn. The economisti s ncutral on thc oucstion.
d
Y
v
n
r