Rivista Italiana di Economia Demografia e Statistica Volume LX IX n. 2 Aprile -Giugno 2015 [629941]

Rivista Italiana di Economia Demografia e Statistica Volume LX IX n. 2 Aprile -Giugno 2015

DO CONSENSUAL UNIONS DIFFER FROM MARRIAGES ?
A STUDY OF FAMILY CHARACTERISTICS AND RECEIVED
HELPS IN ITALY1

Romina Fraboni , Eleonora Meli

1. Introduction

In the last twenty years in Italy family formation has changed : marriage has
decline d, while non -marital cohabitation has increased . One of the evidence of the
shift in family formation process is the rise in b irths out-of-wedlock .
In the literature the deinstitutionalization of marriage is linked to the growth of
cohabitation and its acceptance is occurring in stages in western countries : at the
beginning cohabitations is an avant -garde phenomenon, then it is accepted as a test
or a prelude to marriage, then it becomes acceptable as an alternative to marriage,
and finally, it becomes indistinguishable from it (Cherlin 2004, Rosina and Fraboni
2004, Kiernan 2002, Villeneuve -Gokalp 1990) .
The late diffusion of consensual unions, in comparison with other wes tern
countries, can be attributed to low social acceptance, parental support in favor of
marriage, existence of religious norms and lack of a legal acknowledgment.
A number of studies have shown the relation between cohabitation and social
support (Nazio and Saraceno 2012, Di Giulio and Rosina 2007, Eggebeen 2005).
Since 1993, the multipurpose household surveys started to systematically
investigate structure , composition, formation and dissolution of families and
households .
The aim of this work is to study differences and similarities between married
and unmar ried couples, particularly in order to understand whether there is a
convergence process between cohabitat ing and married co uples in Italy , as far as
structural characteristics (age, educational level, employment, presence and
number of children) and behavi ors (home activities share, childcare) are concerned .
Given the strong family ties that goes in support of newly formed unions we
study the differences in the probability to receive help for both consensual and
marital unions.

1 Although the paper is the result of the joint work of the authors, paragraph 2 and 3 are due to E.
Meli, paragraph 4 to R. Fraboni, introduction and conclusions are in common.

106 Volume LX IX n. 2 Aprile -Giugno 2015

2. Method and d ata

We f ocus on couples with women in reproductive ages (15 -49 years),
distinguishing married couples at th eir first marriage and consensual unions of
never married partners.
Firstly, we present a description of the spread of consensual unions in Italy from
1993 to 201 4 and we focus on differences between un married and married couples.
Secondly, we compare ma rried and unmarried couples as far as helps received
from non -cohabiting peoples are concerned.
To this aim w e use data from Multipurpose survey on “Aspects of daily life ”
(1993 -2014) and “ Family and Social Subjects ” (1998, 2003, 2009), led by Istat on
independent sample s of 24,000 national representative household s where all
members are interviewed. Data are representative at regional level of analysis. We
apply a l ogistic regression analysis on the probability to receive helps and its
change across time.

3. The spread of consensual unions in Italy and regional differences

During the last twenty years, unmarried couples have steadily increased from
227 thousands in the early 1990s to more than 1 million in 2013 -2014 (i.e. from
1,6% to 7,5% of all couples) . Consensual unions are differently spread across
Italian regions as they are more frequent in the North -east (13,0%), the North -west
(10,6%), and less frequent in the Islands (4,1%) and in the South (3,3%). Focusing
on couples with woman in repr oductive age s, cohabiting couples of never married
partners have become ten times higher – from 61 thousands in 1993 -1994 to 623
thousands in 201 3-2014, therefore representing 8,9% of overall Italian couples in
most recent year (Figure 1) .
Premarital cohabitation has always played a relevant role in partnership. Indeed,
beyond the current partnership, taking into account the overall amount of people
with an consensual union, also in the past, it emerges that, in 2009 , in Italy , there
were ov erall 5 mi llion 915 thousand s people who were living in a consensual union
or who had experienced it in their past life -cours e: i.e. 11,5% of people aged 15
and over (ISTAT 2011a) . Most of them have, eventually, married. Far from being
perceived as an alte rnative to marriage, cohabitation is more often considered a
step to marriage in Italy .
Over 55% of people who has had a consensual union and who , at the beginning,
did not intend to marry, perceives cohabitation as a test to marriage: they declare
that “living together” has proven the possibility to marry.

Rivista Italiana di Economia Demografia e Statistica 107

Figure 1 − Unmarried couples and couples of never married partners per 100 couples
with woman aged 15 -49. Years 1993 -2014

Source : Istat, Aspects of daily life

3.1. Differences and similarities on partners’ characteristics

The analysis of the differences and similarities on partners’ characteristics show
that cohabiting women are younger , have higher education al level and are more
often employed than married women at their first marriage (Table 1) . Indeed, the
majority of women (15 -49) in a consensual union is still younger than 35 years ,
even if it is declining over time (78,7% in 1993 -1994 and 52,7% in 201 3-2014);
among married women at their first marriage only 23,2% of them is below age 35
(it was 39% in 1993 -1994). Thus, postponing family formation affects both first
marriage and first consensual union, with a lower proportion of younger women
than in the past. However, consensual unions show the largest share of younger
women . Age heterogamy is changing quickly and it is generally lower among
consensual unions than marital unions. Married couples at their first marriage are
composed more often by men older than women ( 79,6% vs 66,2% in consensual
unions), while consensual unions of never married partners are composed more
often by women older than men ( 24,1% in 2013-2014) (Table 2) .

02468101214
1993-1994
1994-1995
1995-1996
1996-1997
1997-1998
1998-1999
1999-2000
2000-2001
2001-2002
2002-2003
2003-2005
2005-2006
2006-2007
2007-2008
2008-2009
2009-2010
2010-2011
2011-2012
2012-2013
2013-2014
Unmarried couples Never married couples

108 Volume LX IX n. 2 Aprile -Giugno 2015

Table 1 − Women in couple by couple typology, age, level of education and geographical
area. Years 1993 -1994 and 2013 -2014 (percentage distributions)
Women's characteristics Never married couples First married couples
1993 -1994 2013 -2014
1993 -1994 2013 -2014
Age
15-34 79,2 52,7
39,0 23,2
35-49 20,8 47,3
61,0 76,8
Total 100,0 100,0
100,0 100,0
Education
Graduated 9,4 21,8 7,8 16,2
Upper secondary 38,8 50,4 34,0 45,9
Up to lower secondary – school license 51,8 27,8
58,2 38,0
Total 100,0 100,0
100,0 100,0
Geographical area (a)
North -West 1,0 11,8
93,8 78,3
North -East 1,2 13,9
94,8 76,5
Centre 0,3 8,4
95,3 83,1
South 0,2 3,7
97,4 92,1
Islands 0,9 4,2
96,1 90,4
Italy 0,7 8,9 95,5 83,3
(a) As a % of overall couples in the same area.
Source : Istat, Aspects of daily life

The rise of the level of education attained by women is particularly relevant
among youngest cohorts. However graduated women rise more in consensual
unions than in married couples. The gap in educational levels between partners are
increasing between thes e two union typologies. In 30,6% of consensual unions
women have a higher educational level than men. This occurs in 27, 6% of married
couples.
Unmarried women are more often employed than married ones. Employed
women in unmarried couples have increased mo re during the last 20 years.
Consensual unions are characterized by double -earner partners more often than
married couples (6 5% vs 45,7%). In 201 3-2014 couple with “employed man and
housewife” amount to 8,5 % in consensual unions, while it reaches 27,3% in
married couples .
Couples with young children are increasing, too. The number of c hildren born
out-of-wedlock is increasing: from 8% in 1995 to 25,9% in 2013. As a
consequence t he proportion of consensual unions with children increases from 33%
in 1993 -1994 to 53% in 201 3-2014. Children in consensual unions are younger

Rivista Italiana di Economia Demografia e Statistica 109

(74,5% of children aged 0 -3 years vs 40 ,2% in married couples) . Most of
unmarried couples have only one child ( 68,5%).

Table 2 − Never married couples and first married couples by partners' characteristics .
Years 2013 -2014 (percentage distributions).
Partners' characteristics Never married couples First married couples
Age difference
She is younger 66,2 79,6
Same age 9,7 8,4
She is older 24,1 12,0
Educational level
She is more educated 30,6 27,6
Same education 53,2 55,3
She is less educated 16,3 17,1
Occupational status
Both employed 65,0 45,7
He is employed, she is housewife 8,5 27,3
He is employed, she is unemployed 11,3 10,1
He is unemployed, she is employed 4,6 4,9
Other 10,5 12,0
Presence of children With a t least one child 53,0 86,1
With c hildren aged 0 -5 74,5 40,2
Source : Istat, Aspects of daily life

Home activities sharing and childcare are changing slowly , even among
unmarried cohabitants. Quite al l of domestic activities is done by women, but in
unmarried couples men are more often involved in daily care and home duties : in
17,1% of the cases men in consensual unions work at least as woman vs in 11,1%
of the cases of married men. In presence of 0 -3 years old children, the involvement
of unmarried fathers in daily childcare is larger than that of married ones, and it is
rising over time. Among all the considered activities unmarried fathers do more
than marr ied ones.

110 Volume LX IX n. 2 Aprile -Giugno 2015

Figure 2 – Children aged 0 -3 years by father’s involvement in daily care activities and
father’s marital status, per 100 children aged 0 -3 years. Years 1998, 2003,
2009

Source : Istat, Family and social subjects

4. Inequality in the support to families with children

The possibility to receive informal support (such as economic aid, health
assistance, personal care, help in study or work, childcare, elderly care or help with
disabled people, domestic chores, accompanying, companionship,
bureaucratic /administrative duties …) during the four weeks preceding the
interview is very relevant in a society where strong family ties are observed.
Previous studies have shown that, apart from families with a disabled person
that have the highest rankin g in received help s, families with children and
employed mother are the privileged target of informal care received from n on-
cohabiting people (ISTAT 2011 b). Besides that, we include public help s provided
by the municipality or by charity or health assista nce institutions, as well as,
private services paid by the family (such as baby -sitter, domestic help, assistance to
elderly or disable d people ). Overall, informal support covers three fourth of the
total amount of received helps. Comparing the two couples typologies with
children 0 -13 years we find that r eceived h elps are higher for unmarried couples 051015202530354045
Married
Unmarried
Married
Unmarried
Married
Unmarried
Married
Unmarried
Married
Unmarried
Feed Sleep Dress Bath Diapers1998
2003
2009

Rivista Italiana di Economia Demografia e Statistica 111

than for married ones. Received helps increase in both cases: from 42,4% to 43,7%
in presence of unmarried mothers and from 30,7% in 1998 to 38,8% in 2009 in
presence of married ones. In order to better understand the determinants that affect
the probability to receive help in married and unmarried couples with children we,
then, apply a logistic regression model . We focus on couples at first marriage and
consensual unions of never married partners with women aged 15 -49 and we study
the probability to receive helps. We include partners’ characteristics ( such as age,
education level, employment status ); couples’ characteristics (geographical area ,
demographic dimension of municipalities, number of 0 -3 children, presence of lone
people in the family or extended families); family of origin’s characteristics: both
partners parents’ education. We pool together the three survey editions to
understand the evolution o ver time of the phenomenon under study .

4.1. Main results of a model on the probability to receive helps
The probability to receive helps is significantly higher for women aged 30 -39
than aged 15 -29 years and it decreases with increasing age of the man. As for the
territorial effect we observe a higher probability to receive helps in the North -east
and North -west than in the South and Islands. In the suburbs of the metropolitan
area helps are slightly more frequent than in the centre of the main municipality.
Education is highly associated with helps received by the couples. Indeed, t he
probability to receive helps increases with higher educational level of each
partners . The same effect is shown by the social status of the family of origin: the
probability to receive helps increases with the highest level of education of the
woman’s parents. As for the occupat ional status we notice that, in comparison with
double -earners couples, those characterized by the male -breadwinner model
receive lower help. On the contrary, higher probability to receive help is shown by
couples where none of the partners works, particu larly where the man is
unemployed and the woman is housewife , and, evidently, this is a selected set of
couples in need of help. The family structure plays a relevant role in the support,
particularly when more than one family cohabit, i.e. extended family , as this
reduces helps from other non -cohabiting people. The presence of young children
(aged 0 -3 years) in the couple has the strongest effect in the analysis: it positively
affects the helps received by the couple and it rises with the number of young
children. As for the different typology of couple , it emerges that, all other things
being equal, marriage is associated with higher probability to receive any kind of
support in comparison with consensual union . Finally, the trend over time shows an
increase in the trend of helps received in 2009 with respect to 1998.

112 Volume LX IX n. 2 Aprile -Giugno 2015

Table 3  Main results of the logistic regression on the probability to receive helps

Source : Istat, Family and social subjects

b S.E. Exp(b) p b S.E. Exp(b) p
Age woman Higher level of
(ref=15-29) education of Her's
30-39 0,249 0,062 1,282 <,0001 parents
40-49 0,032 0,078 1,033 0,682 (ref=Up to primary)
Age man Upper secondary or
(ref=15-29) tertiary 0,239 0,051 1,270 <,0001
30-39 -0,163 0,086 0,849 0,057 Lower secondary 0,112 0,046 1,119 0,015
40-49 -0,185 0,097 0,832 0,056 Couple's
50 and more -0,488 0,113 0,614 <,0001 employment status
Geographical area (ref=Both employed)
(ref=North-West) He is employed, she
North-East 0,195 0,051 1,216 0,000 is housewife -0,589 0,043 0,555 <,0001
Centre 0,030 0,052 1,031 0,562 He is employed, she
South -0,206 0,051 0,814 <,0001 is unemployed -0,289 0,083 0,749 0,001
Islands -0,126 0,064 0,881 0,047 He is unemployed,
Municipality she is employed 0,034 0,129 1,035 0,791
(ref=Metropolitan He is unemployed,
area – Centre) she is housewife 0,745 0,109 2,106 <,0001
Metropolitan area – Other -0,022 0,076 0,978 0,772
suburbs 0,303 0,067 1,354 <,0001 Couple's Typology
until 10,000 (ref=Unmarried)
inhabitants 0,072 0,059 1,075 0,222 Married 0,192 0,091 1,212 0,035
10,001 inhabitants Nr. Children 0-3
and over 0,067 0,055 1,070 0,222 (ref=0)
Woman One 0,726 0,043 2,067 <,0001
educational level Two or more 1,199 0,102 3,316 <,0001
(ref=Graduated) Presence of lone
Upper secondary people in the hh.
diploma -0,321 0,058 0,725 <,0001 (ref=No)
Lower secondary Yes -0,088 0,089 0,916 0,321
school licence -0,580 0,069 0,560 <,0001 Presence of more
Primary school than 1 cohabiting
certificate -0,649 0,098 0,523 <,0001 family member
Man educational (ref=No)
level Yes -0,733 0,133 0,480 <,0001
(ref=Graduated) Survey edition
Upper secondary (ref.=1998)
diploma -0,505 0,060 0,603 <,0001 2003 0,020 0,041 1,020 0,625
Lower secondary 2009 0,156 0,044 1,169 0,000
school licence -0,610 0,068 0,543 <,0001
Elementary school
licence -0,716 0,094 0,489 <,0001

Rivista Italiana di Economia Demografia e Statistica 113

Moreover, running models separate ly by kind of couple , we find that having
young children , especially if more than one, is the most important determinant of
receive d helps either for married couples and for unmarried ones. The model
concerning married couples is consistent with the overall model. Instead, w hen
compared with the general one, the model o f unmarried couples shows no effect of
the partners’ and their family of origin’s level of education , as well as no increase
over time. Moreover, the female age effect is significant also at a ge 40 -49 years
where it increases. The effect of living with lone people is significant and negative.
The strongest and positive effect is associated with the absence of employment of
both unmarried partners, which makes this group of people in need of hel p.

5. Conclusions
In Italy differences on partners’ characteristics between cohabiting and married
couples still persist as in the former partners are usually younger , more educated
and with higher female employment rate , more gender -balanced than in the latter ;
moreover, the number of children in cohabiting couples is increasing , but it is
usually lower and they are younger than in married couple s. Consensual unions are
not uniformly spread over the country. The study of received helps show s that
couples with young children have the highe st probability to receive helps,
regardless of union typology. Moreover, from the model it emerges that married
couples have higher probability to be supported than cohabiting couples and that,
only in the first case helps have risen across years .

114 Volume LX IX n. 2 Aprile -Giugno 2015

References
CHERLIN A. J . 2004. The deinstitutionalization of American marriage. Journal of
marriage and family , Vol. 66, pp. 848-861.
DI GIULIO P., ROSINA A . 2007. Intergenerational family ties and the diffu sion of
cohabitation in Italy. Demographic Research , Vol. 16(14), pp. 441-468.
EGGEBEEN D. J. , 2005. Cohabitation and exchanges of support. Social Forces ,
Vol. 83, N. 3, pp. 1097 -1110.
KIERNAN K. 2002 . Cohabitation in Western Europe: trends, issues and
implications, in Booth, Alan and Crouter, Ann. C, (eds.) Just Living Together:
Implications of Cohabitat ion on Families, Children and Social Policy . Lawrence
Erlbaum Associat es Inc, pp. 3 -32.
ISTAT 2011 a. Come cambiano le forme familiari. Comunicato stampa.
www.istat.it .
ISTAT 2011b. Rapporto annuale. La situazione del Paese nel 2010.
http://www3.istat.it/dati/catalogo/20110523_00/Avvio2010.pdf
LACHANCE -GRZELA M., BOUCHARD G. 2009 . La cohabitation et le marriage,
deux monde à part? Un examen des caractéristiques démographiques,
individuelles et relationelles , Canadian journal of behavi oural science , Vol. 41,
pp. 37-44
NAZIO T. , SARACENO C . 2012. Does cohabitation lead to weaker
interge nerational bonds than marriage? A comparison betwe en Italy and the
United Kingdom, European Sociological Review , pp. 1-16.
PIRANI E., VIGNOLI D. 2014 . Are spouses more satisfied than cohabitors? A
survey over the last twenty years in Italy, Working paper – Dipartimento di
Statistica, Informatica, Applicazioni "Giuseppe Parenti" – Università degli studi
Firenze, N. 9.
POORTMAN A. -R., MILLS M. 2012 . Invest ments in marriage and cohabitation:
the role of legal and interpersonal commitment, Journal of marriage and family ,
Vol. 74, pp. 357-376.
ROSINA A. , FRABONI R . 2004. Is marriage losing its centrality in Italy?
Demographic Research , Vol. 11, pp. 149-172.
SALVINI S. , VIGNOLI D. 2014 . Convivere o sposarsi?. Il Mulino, Bologna.
VACULÌK M., JEDRZEJCZYKOVÀ V. 2009 . Commitment in unmarried
cohabitation, Studia psychologica, Vol. 51, pp. 101-117.
VILLENEUVE -GOKALP C . 1990. Du mariage aux unions sans papiers: histoire
récente des transformations conjugales. Population , Vol. 2, pp. 265-298.

Rivista Italiana di Economia Demografia e Statistica 115

SUMMARY
Do consensual unions differ from marriages ?
A study of family characteristics and received helps in Italy

In the last 20 years in Italy marriage has lost its exclusivity in the traditional model of
family, at the same time consensual union has been more and mo re accepted as a form of
union and, as a result, at least a quarter of births are out -of-wedlock.
This paper exam ines the spread of consensual unions and the structural characteristics
of partners in comparison to married couples. Then , we focus on the differences in the
support received in presence of children.
Results show that consensual unions of never married p artners preserve distinctive traits
related to age, education and labor force participation of both partners in comparison with
more traditional couples. Moreover, cohabiting couples have less and younger children than
married ones. Finally, the most relev ant determinant of received help is the presence of
young children. However, couples at their first marriage show a higher – and increasing
over time – propensity to get help than cohabiting couples.

_________________________
Romina FRABONI , Istat – Italian National Institute of Statistics , fraboni@istat.it
Eleonora MELI , Istat – Italian National Institute of Statistics , elmeli@istat.it