Social and Demographic Consequences of Unilateral Divorce Law

Thorsten Kneip (MPISoc) Gerrit Bauer (LMU)

Contact: kneip@mea.mpisoc.mpg.de

8th Congress of the European Society on Family Relations, Dortmund, Germany, August 31–September 3, 2016

Motivation

2

Persistent increase in divorce rates in Europe since about 1970s

Family law is part of the social environment in which family processes are embedded

Only few empirical sociological studies Only little evidence other than from the U.S.A. Largely based on aggregate data

Motivation

3

0

1

2

3

4

5 Di

vorc

es/1

000

Mar

ried

1960 1970 1980 1990 2000

Actual divorce rate Counterfactual rate Divorce law effect

Source: Eurostat; EU-15 + CH, IS, NO; own calculations (c.f. Kneip & Bauer 2009)

Motivation

4

Persistent increase in divoce rates in Europe since about 1970s

Family law is part of the social environment in which family processes are embedded

Only few empirical sociological studies Only little evidence other than from the U.S.A. Largely based on aggregate data Different micro mechanisms might mediate the effect Selection into population at risk Changes in investment behavior

Summary of findings

5

The introduction of the possibility to unilaterally divorce has increased the risk of divorce … for couples who got married before the reform; … for couples who have married under unilateral divorce law

(the effect is offset by a changed selection into marriage with respect to match quality);

… for parents.

Beyond this, the introduction of unilateral divorce law has … contributed to an increase in the age of marriage; … contributed to an increase in female labor force participation; … contributed to a decline in (marital) fertility; … long-term negative consequences for children who grew up under

unilateral divorce law (beyond an effect of parental divorce).

Theoretical background I

6

“Naive“ view: P(divorce) ↑ when divorce gets easier

Sociological view: P(divorce) ↑ when norms allow for dissolution in pursuit of individual

happiness (e.g. Roussel 1980)

Family Economics (Becker 1981): Switch to unilateral divorce → shift in property rights No effect on divorce under Coasian bargaining

Why Coasian bargaining might fail: Transaction costs (e.g. Stevenson & Wolfers 2006) Marital assets as collective goods (e.g. Zelder 1993; Chiappori et al. 2007)

Empirical evidence for unilateral law effect is usually interpreted as non-applicability of the Coase Theorem

Theoretical background II

7

Unilateral divorce laws may affect Selection into marriage (with respect to match quality)

Lower match quality due to reduced exit costs Raise in marriage rates (Alesina & Giuliano 2007)

Higher match quality due to reduces marital gains Decrease in marriage rates (Rasul 2003)

Investments in outside options (e.g. FLFP) Due to lack of compensation for reduced human capital (Parkman 1992)

Marital specific investments (e.g. children) Due to time constraints at given labor supply

Match quality and specific investments affect marital stability

Data

8

1915

1970

1977

1976 1978

1981

-

1983

2000

1975

1971

16.6%

60.3%

23.1% Divorce regime over course of marriage: always mutual consent consensual/unilateral always unilateral

Retrospective biographic information (marriages, births, employment histories, divorces)

Variation in the timing of introducing unilateral divorce

Suitable sample composition:

n ≈ 20.000

Method

9

Estimation model:

uni: unilateral divorce law c: country fixed effects yob: cohort fixed effects x: individual characteristics

Country-specific trends

Marriage

10

0

.2

.4

.6

.8

1

Surv

ival

Rat

e

20 30 40 50 60 70 Age

consensual

unilateral

The transition to marriage by divorce law

~ 3 years; ~ 8 months can be attributed to the introduction of unilateral divorce law

Divorce

11

Haz

ard

rate

(sm

ooth

ed)

Always consensual Consensual/unilateral Always unilateral

Not controlling for age at marriage

0

.002

.004

.006

0 10 20 30 40 50 Marriage duration

Controlling for age at marriage

0

.002

.004

.006

0 10 20 30 40 50 Marriage duration

Divorce law effect by number of kids

12

.5

1

1.5

2

2.5 Co

nditi

onal

effe

ct

3+ 0 2 1

Labor force participation

13

-.4

-.3

-.2

-.1

0

.1Di

ffere

nce

in w

orki

ng h

ours

Before marriage After marriage

Women / consensual Men / consensualWomen / unilateral Men / unilateral

Fertility

14

.4

.6

.8

1

1.2

1.4

Prop

ortio

nal E

ffect

s on

haza

rd ra

te

without with without with without with

Transitions to 1st, 2nd, and 3rd child with and without controlling for age at marriage

Transition to 1st child Transition to 2nd child Transition to 3rd child

15

Education Marriage Parenthood Smoking Overweight Depression

UDL (-) (-) - - - - + + (+) (+) 0 0

+ 0 0 - (-) -

Parental marriage

intact

Long-term effects on children?

Summary & discussion

16

The introduction of unilateral divorce had far-reaching consequences for marital interaction and, as a consequence, of demographic processes.

The applicability of Coase’s theorem seems reasonable in the absence of children.

The presented findings point to the complexity of familial action against the background of structural conditions, partly prescribed by policy, and the need for a careful analysis of the (unintended) consequences of policy instruments prior to their introduction.

Publications

17

• Kneip T & Bauer G (2009) Did Unilateral Divorce Laws Raise Divorce Rates in Western Europe? Journal of Marriage and Family 71 (3), 592-607.

• Reinhold S, Kneip T & Bauer G (2013) The Long Run Consequences of Unilateral Divorce Laws on Children – Evidence from SHARELIFE. In: Journal of Population Economics 26 (3), 1035-1056.

• Kneip T, Bauer G & Reinhold S (2014) Direct and Indirect Effects of Unilateral Divorce Law on Marital Stability. Demography 51 (6), 2103-2126.

THANK YOU!

18

BACKUP

Robustness

.6

.8

1.2

1.4

1.6

1.8

2.2

1.0

2.0

our m

odel

linea

r tren

d

cubic

trend

incl. t

ime t

rend

incl. l

eads

incl. l

ags

female

s only

reliab

les on

ly

common

SE

boots

trapp

ed S

E

robus

t SE

divorc

e

age/m

arriag

e squ

ared

age/u

nion f

ormati

on

incl. d

e jure

law

de ju

re eff

ect

Model robustness

1.2

1.4

1.6

1.8

1.0

2.0

w/o Aus

tria

w/o Germ

any

w/o Swed

en

w/o Neth

erlan

ds

w/o Spa

in

w/o Ita

ly

w/o Fran

ce

w/o Den

mark

w/o Gree

ce

w/o Switz

erlan

d

w/o Belg

ium

incl. C

Z, PL,

GDR

Country selection

𝑢𝑢�ℎ𝐷𝐷 𝑢𝑢�ℎ𝑀𝑀

𝑢𝑢�𝑤𝑤𝐷𝐷 𝑢𝑢�𝑤𝑤𝑀𝑀

M

D‘ D 𝑢𝑢�𝑤𝑤𝐷𝐷𝐷 ≥

𝑢𝑢�ℎ𝐷𝐷𝐷 Uh

Uw

Bargaining over divorce

𝑢𝑢�ℎ𝐷𝐷 𝑢𝑢�ℎ𝑀𝑀

𝑢𝑢�𝑤𝑤𝐷𝐷 𝑢𝑢�𝑤𝑤𝑀𝑀

M D

M‘

𝑢𝑢�ℎ𝑀𝑀𝐷

𝑢𝑢�𝑤𝑤𝑀𝑀𝐷

Uh

Uw

Bargaining over divorce

𝑢𝑢�ℎ𝐷𝐷 𝑢𝑢�ℎ𝑀𝑀

𝑢𝑢�𝑤𝑤𝐷𝐷

𝑢𝑢�𝑤𝑤𝑀𝑀 M

D

Uw

Uh

Bargaining over divorce

𝑀𝑀𝐶𝐶𝐶𝐶∗

𝑈𝑈𝑤𝑤

𝑈𝑈ℎ

𝑇𝑇𝐶𝐶𝐶𝐶𝑤𝑤 𝑇𝑇𝐷𝐷𝑤𝑤

𝑇𝑇𝐶𝐶𝐶𝐶ℎ 𝑇𝑇𝐷𝐷ℎ

𝑀𝑀𝐷𝐷∗

Bargaining within marriage

𝑀𝑀𝐶𝐶𝐶𝐶

𝑈𝑈𝑤𝑤

𝑈𝑈ℎ

𝑀𝑀𝐷𝐷0 𝑀𝑀𝐷𝐷1𝐷 𝑀𝑀𝐶𝐶𝐶𝐶𝐷

𝑀𝑀𝐷𝐷0𝐷

𝑇𝑇𝐶𝐶𝐶𝐶𝑤𝑤

𝑇𝑇𝐶𝐶𝐶𝐶ℎ 𝑇𝑇𝐷𝐷ℎ

𝑇𝑇𝐷𝐷0𝑤𝑤

𝑇𝑇𝐷𝐷1𝑤𝑤

Bargaining within marriage