2500+ words Essay

Same sex & opposite sex

SESS 0042

18th November 2019

Same sex marriage…  is a very new phenomenon  is therefore not very-well documented yet

 BUT:  We know a little more about same-sex dating

market  We can fall back on some of the theory to

address similarities and differences  Still, a lot will be speculative

2

3

How are same-sex couples just like hetero-couples?

 They operate in a dating/marriage market, searching for a pairwise match

 They, too, face frictions and search costs  They face competition from other

potential partners  Their gains from relationship also

include public goods, specialization etc.

4

How are same-sex couples different from hetero- couples?

 Potentially higher search costs  They are rarer in the population  But this may be mitigated by technology

 Potentially lower gains from specialization  They cannot have own children (but can

have all the other gains)  Stigma (reduces enjoyment from

relationship)  They cannot get married (in some places)

Impact of higher search costs

 Assume a proportion g of gays/lesbians and s of heterosexuals in pop

 Assume every person has a trait  Traits have a distribution  U = U(θi, θ-i, θi – θ-i)

 At any point, one can stay with their current partner or abandon him/her in search of someone better

 The new relationship yields utility in the next period (discounted at discount rate δ)

 But you must draw someone of matching sexual orientation or you remain alone

5qi

g

s

q - i

F (q - i )

qi F (qi )

 

EUt+1 =d (1 - s) ´0 + U(q i,q- i,t+1) fs(q- i )dq- i 0

q

ò é

ë ê

ù

û ú

EUt+1 =d (1 - g) ´0 + U(q i,q- i,t+1) fg(q- i )dq- i 0

q

ò é

ë ê

ù

û ú

q - i

Impact of higher search costs

 Assume population has proportion g of gays/lesbians and s of heterosexuals

 Assume every person has a trait  Assume traits have uniform distribution  Relationship utility:

6

qi

U =q1 2 - (qi - q- i )

2 =2qiq- i - q- i

2

 

fs(q i ) = s q

; fg(qi ) = g q

q - i

 

q1 q- i

 

q2

 

U(qi,q- i )

At which point will the search stop?

 Keep searching until:  U(current relationship) ≥ EU(further search)

7

F (q - i )

q - iq

- i

 

˜ q i

g

s

At which point will the search stop?

 Keep searching until:  U(current relationship) ≥ EU(further search)

8

F (q - i )

q - iq

- i

 

˜ q i

 

˜ q - i

Range of improvement

P ro

b a b il it

y t

h a t

a b

e tt

e r

m a tc

h w

il l

c o m

e u

p

 

EU =d 2q iq- i,t+1 - q- i,t+1 2( )

0

2q i

ò s q

dq - i ³U(q- i,t )

For

 

qi £ q

2

 

qi > q

2 For

 

EU =d 2qiq- i,t+1 - q- i,t+1 2( )

0

q

ò s q

dq - i ³U(q- i,t)

At which point will the search stop?

9

 Keep searching until:  U(current relationship) ≥ EU(further search)  Specific solution for  Where

U =q1 2 - (qi - q- i )

2 =2qiq- i - q- i

2

 

qi £ q

2

 

EU =d 2q iq- i,t+1 - q- i,t+1 2( )

0

2q i

ò s q

dq - i =

= ds q

qiq- i,t+1 2

- 1

3 q

- i,t+1 3

é

ë ê

ù

û ú

0

2q i

=

= ds q

qi (4qi 2 -

1 3

(8qi 3 ) - 0

é

ë ê

ù

û ú =

4 3 dsqi

3

q

 

U(qi ) =2q iq- i - q- i 2

 

2q iq- i - q- i 2 ³

4

3

dsq i 3

q

 

0 ³ 4

3

dsqi 3

q - 2qiq- i +q- i

2

10

When will the search stop? Comparing solutions

 

2q iq- i - q- i 2 ³

4

3

dsq i 3

q

 

2qiq- i - q- i 2 ³

4

3

dgqi 3

q

q - iq- i

 

q1  

4

3

dsq i 3

q

 

4

3

dgqi 3

q

 

U(qi,q- i )

For hetero daters For same-sex daters

 Overall, fewer same-sex daters will be in relationships (frequent draws of 1 – g)

 Same-sex relationships will be unhappier on average (lower threshold)

 Proportions g and s may change with technology and social liberalization

 

4

3

dsq i 3

q >

4

3

dgq i 3

q Clearly, because s > g

Specialization – lower gains?

 Two men/women may be more similar in terms of productivities than mixed couples

 Consider a world with two goods, X and Y  Time is fully allocated to produce these two goods,

so that tY = 1 – tX  Men and women differ in their productivities when

producing X and Y:  Men: X = mX(t); Y = mY(1 – t)  Women: X = fX(t); Y = fY(1 – t)

 All men are identically productive, as are all women  All people have the same utility fn: U = U(X, Y) 11

General formulation

12

Optimization while single:

 

max t

U(X,Y)

 

X =mX (t);Y =mY (1 - t)

subject to

 

¶U ¶X

¶mX ¶t

- ¶U ¶Y

¶mY ¶t

=0

¶U ¶X

¶mX ¶t

= ¶U ¶Y

¶mY ¶t

In a same-sex relationship:

 

max t1 ,t2

U(X,Y)

 

X =mX (t1) + mX (t2 )

Y =mY (1 - t1) + mY (1 - t2 )

subject to

 

¶U ¶X

¶mX ¶t1

- ¶U ¶Y

¶mY ¶t1

=0

¶U ¶X

¶mX ¶t2

- ¶U ¶Y

¶mY ¶t2

=0

 

¶U ¶X

¶mX ¶t1

= ¶U ¶Y

¶mY ¶t1

¶U ¶X

¶mX ¶t2

= ¶U ¶Y

¶mY ¶t2

Identical  t1 = t2

In a hetero relationship:

 

max t1 ,t2

U(X,Y) subject to

 

X =mX (t1) + fX (t2 )

Y =mY (1 - t1) + fY (1 - t2 )

 

¶U ¶X

¶mX ¶t1

- ¶U ¶Y

¶mY ¶t1

=0

¶U ¶X

¶fX ¶t2

- ¶U ¶Y

¶fY ¶t2

=0

Different equations yield different

results  t1 ≠ t2

Consider an example:

13

 

U =ln X + AlnY

X =ta ;Y =(1 - t)b

 

max t

U =ln ta + Aln(1 - t)b =a ln t+ Abln(1 - t)

 

¶U ¶t

= a

t -

Ab 1 - t

=0

t = a

Ab +a

 

max t1 ,t2

U =ln X + AlnY

 

X =t1 a + t2

a ;Y =(1 - t1) b + (1 - t2 )

b

 

t1 : at1

a - 1

t1 a + t2

a -

Ab(1 - t1) b - 1

(1 - t1) b + (1 - t2 )

b =0

t2 : at2

a - 1

t1 a + t2

a -

Ab(1 - t2 ) b - 1

(1 - t1) b + (1 - t2 )

b =0

subject to

Again, t1 = t2

ata- 1

2ta =

Ab(1- t)b- 1

2(1- t)b

Þ a

t =

Ab 1- t

Þ t = a

Ab +a =t1 =t2

While single:

In a gay relationship:

Notice: not only identical allocation among the two men but also identical to singlehood.

14

How are same-sex couples different from hetero- couples?

 Potentially higher search costs  They are rarer in the population  But this may be mitigated by technology

 Potentially lower gains from specialization  They cannot have own children (but can

have all the other gains)  Stigma (reduces enjoyment from

relationship)  They cannot get married (in most places)

Why is same-sex marriage so new?

 Humans have tried almost everything else:  Polygamy and polyandry, and even monogamy  Marriage with concubinage  Wife-sharing (though rare and usually short-lived)  Temporary marriage  Levirate marriage  Child marriage  Marriage with bride price and with dowry  Marriage regime with no divorce & with unilateral divorce, with no-

remarriage and with some  Marriage with neo-localism and marriage into parental household  Cohabitation

 Yet: same-sex marriage is, historically, almost unheard of: why?

15

Short history of marriage as an

institution

 Defining the terms:  Neo-localism = moving to a new

household  Inheritance:

 EPI = equal partible inheritance  PLI = patrilineal inheritance (e.g.

primogeniture)  Family of procreation vs family of

orientation 16

Short history of marriage as an

institution

17

Hunter-gatherers Pastoralists- herdsmen

Subsistence agriculture

Mod. industrial societies

Motivation for instituting marr.

R, A R, A, I R, I, A C, A?, R?

Inheritance EPI PLI PLI EPI

Involvement of kin

Yes in formation, not in operation

Both in formation & in operation

Both in formation & in operation

Weak in either

Family status What status? Somewhat important

Crucial, formalized

Decline in importance

Focus Family of procreation

Family of orientation

Either (some neo- localism)

Family of procreation

Divorce? Acceptable Not acceptable Ask Henry 8 Acceptable

Source: Quayle, Robina. A History of Marriage Systems. Greenwood Press, 1988

Emerging patterns  Once societies produce wealth, marriage matters for

inheritance:  Large cross-sectional variation in handling marriage

 Diplomatic affair among ruling families  Some negotiations among propertied classes  Relatively unstructured among the poor

 It affects who is involved and what rights and duties couples have

 First-born’s marriages are more micromanaged  Great impetus for formalization of marriage

 UNTIL: the wealthiest times in history  i.e. NOW!

 What gives? 18

Impact of economic growth

 Main determinants of social status and living standard:  In stagnant societies: inheritance  In growing societies: human capital (& inheritance)  The old Gould, Moav & Simhon (2008) argument

 For young person, economic growth implies:  Future earned income over lifetime increases in importance relative to return

on accumulated family wealth  This weakens the bargaining position of kin in partner choice (i.e. “Marry X, or

I’ll disinherit you” loses power)  Note that this affects even youngsters currently earning little, as it is the

expectation of growth that matters  Empirical predictions:

 Autonomous choice of partner should spread in dynamic societies  Autonomous choice of partner should spread from below, i.e. from poorer strata  Poorer, stagnant but stratified societies will adhere to arranged marriage

practice  The practice will weaken with the onset of persistent economic growth 19

Two-fold impact of economic growth

 It weakens importance of inheritance  Switch from arranged marriage to

autonomous-choice marriage  It leads to convergence of M/F productivities

 Both in the market and at home  Decline in the importance of specialization  Switch from joint production to joint

consumption  Emergence of marriage as a love match

20

What does that have to do with same-sex marriage?

 As long as marriage was about inheritance  Same-sex marriage cannot deliver

 Same-sex marriages could not produce an heir

 As long as marriage was about specialization  Case for same-sex marriage was rather weak

(assuming specialization along gender lines)  Only when both trends came together

 Same-sex marriage becomes a close substitute to heterosexual marriage

21

Empirical predictions

 Demand for same-sex marriage will start in:  Dynamic growing societies  That have reached certain technological

stage  Stagnant and technologically lagging

societies will resist same-sex marriage

22

23

How are same-sex couples different from hetero- couples?

 Potentially higher search costs  They are rarer in the population  But this may be mitigated by technology

 They cannot have own children (but can have all the other gains)

 Stigma (reduces enjoyment from relationship)

 They cannot get married (in most places)

24

Empirical predictions  If search costs are higher…

… matches will be worse than among heterosexual couples

…less stability  If own children are unavailable…

…some kind of gain from relationship is also unavailable…

…less stability  If stigma affects the gains from relationship

…again, stability will suffer

Jepsen & Jepsen (2002)

 Use 5% PUMS of 1990 census  Collect data on age, race, education,

investment income, hours worked for  Married opposite-sex couples  Cohabiting opposite-sex couples  Gay couples  Lesbian couples

 Compare degree of PAM and specialization

25

Jepsen & Jepsen (2002)

26

Jepsen & Jepsen (2002)

27

28

29

30

31

How do same-sex couples operate?

 Sort of similar to hetero couples  They argue about the same stuff  They get together and break up

 Sort of different  They have more equal split of tasks  They face tougher social environment  They rely on friends more than hetero

couples

32

Conflict: 1=never, 5=always

Source: Kurdek (1994)

Kurdek (2006) on relationship dynamics

33

34

Andersson, Noack, Seierstad & Weedon- Fekjaer (2004)

 Data from Sweden and Norway on same-sex and hetero couples

 Finland introduced registered partnership in 1993, Sweden in 1995

 Very close to marriage (although divergent legal framework for adoption)

 They have then about 10 years of data

35

36

Norwegian & Swedish couples

7.78% 11.26% 14.33% 20.03% 8.01%

37

Wassup w/ the lesbians?

 Lesbian couples consistently turn out to be least stable of the various combinations

 It’s not children (own or otherwise), however

38

Kurdek (2004)

 126 gay, 101 lesbian & 483 married couples

 Assessed annually up to 12 times  70 married couples (15%) broke up  24 gay couples (19%) broke up  24 lesbian couples (24%) broke up  Married were stat. significantly different

from same-sex in terms of stability

39

Blumstein & Schwarz (1983)

 1021 married, 233 cohab hetero, 493 gay & 335 lesbian couples

 18-month follow-up survey  4% marrieds broke up  14% cohab hetero couples broke up  13% gay couples broke up  18% lesbian couples broke up  Marrieds were stat. significantly different from

others  No other statistically significant difference

reported

40

Selection issues  Stigma prevents some same-sex couples from going public  So, census may misreport roommates vs cohabiting partners  Most samples oversample white, educated, cosmopolitan gays

& lesbians  Samples are often drawn from magazine readership  Same-sex couples differ from hetero couples in age gap & age

at union  Gay couples are very urban; lesbian couples less so; marrieds

even less  Same-sex couples are more international  Same-sex partnerships are often “second marriages”  Lesbian couples involve a parent more often than gay couples

 Presumably these are not own children

41

Conclusions overall

 Long terms prospects are likely positive:  If social stigma eases, there will be more

enjoyment of relationship  If same-sex marriage wins out, there will be

more stability (and investment in relationship)

 General patterns correspond to what we would expect from theory  Except for the lesbians