Family Lab International

Household structure in the EU

2011-06-08T00:16:00.000-07:00

1. Introduction

Household structure is an interesting area for cross-national study for several reasons. Cross-national differences in household structure reflect important differences between societies: in culture and norms; in the cost and availability of housing; in the economic means available to different groups in society; and in social policy, where differences in tax and benefit regimes may lead to radically different patterns of household structure.

Household structure is also interesting in terms of its relationship to a number of

important outcomes. Poverty, for example, is intimately related to household

structure. Poverty rates are calculated on the basis of household equivalent income

(the sum of the incomes of all household members, divided by a factor related to the

number and ages of these same household members) and household composition is

therefore liable to affect both the numerator and the denominator of this calculation.

There is a large literature dealing with the relationship between household

composition and the risk of poverty (Bane and Ellwood 1986), particularly relating to

vulnerable groups: families with children (Bradbury and Jantti 1999); young adults

(Aassve et al 2007) and older people (Rendall 1995). Of course, poverty is not the

only outcome related to household composition: children’s later outcomes, in terms

of educational achievement, future earnings and so on, are affected by the

composition of the households in which they grow up (Boggess 1998; Francesconi et

al. 2005), even after accounting for the effects of poverty associated with certain

household structures, while older people’s health status is also related to household

composition (Hays 2002).

Household structures across the pre-enlargement EU-15 have been widely

documented (Iacovou 2004, Tomassini et al 2004, Andersson 2004, Robson and

Berthoud 2003, and many others). There are also several studies which include a

number of Eastern European countries (Hantrais et al 2006, Hoem et al 2009,

Gerber 2009). These are based on surveys such as the Family and Fertility survey

and the Gender and Generations survey, which are excellent sources of data for this

area of research, but which include only a limited subset of the new EU Member

States. A smaller number of newer studies have used data covering most or all of

the countries of the enlarged European Union: Mandic (2008) deals with home2

leaving, Liefbroer and Fokkema (2008) deal with fertility; while Saraceno (2008)

provides an overview of household structure in a number of different age groups, as

well as some statistics on labour market status and time use. Our paper is based on

the European Union Statistics on Income and Living Conditions (EU-SILC), which at

the time of writing covers all countries of the expanded European Union except for

Bulgaria, Malta and Romania. Being a general-purpose data set, the EU-SILC does

not allow for such detailed investigation of family formation patterns as some other

data sets. However, its strength lies in the scope of its coverage, which makes it

possible to draw comparisons of many aspects of family structure, over almost the

entire European Union1. We believe that this paper provides a unique resource in

this respect.

We present detailed figures on household structure separately for each country in

the sample. However, we also consider whether there exist groups of countries

which display similar sets of characteristics, and which may be thought of as forming

clusters. Again, there is a well-developed literature in this area relating to the preenlargement

EU-15, and our focus in this paper lies in integrating the new Member

States into this area. In particular, we are interested to uncover the extent to which

the new Member States may be incorporated into existing typologies of family

structure, or whether behaviour in some or all of these countries differs so far from

behaviour elsewhere in Western Europe that it is necessary to think in terms of an

expanded typology.

The section which follows outlines the typologies which have been used to

conceptualise cross-national variations in family structure; we then move on to a

discussion of the data, before presenting our results in Sections 3 to 8.

1.1 Countries and groups of countries

A great deal of comparative research draws on the work of Esping-Andersen (1990

and 1999), whose threefold typology of welfare states consists of: a ‘socialdemocratic’

regime type, characterised by high levels of state support and an

emphasis on the individual rather than the family, typified by the Scandinavian

countries; a ‘conservative’, or ‘corporatist’, regime type, characterised by an

1 Bulgaria, Malta and Romania are not covered here because data for these countries were not

available from the EU-SILC User’s database (UDB) to which Net-SILC members had access.

emphasis on insurance-based benefits providing support for the family rather than

the individual, and typified by the continental European states of France, Germany,

Austria, Belgium and Luxembourg. Esping-Andersen also assigns the Southern

European countries to this group; and a ‘liberal’ regime type, typified by a modest

level of welfare state provision and a reliance on means-tested benefits, exemplified

by the US, and to a lesser extent by the United Kingdom and Ireland.

Several commentators (Leibfried, 1993; Ferrera, 1996; and others) have argued that

the Southern European countries constitute a regime type in themselves, rather than

belonging to the ‘conservative’ group. The arguments on this issue are mixed (Arts

and Gelissen, 2002) with several scholars arguing that structurally, the Southern

welfare states are not dissimilar to those of the ‘conservative’ group of countries.

However, the Southern countries do form an empirically distinct grouping on a wide

range of social and economic indicators, and thus are often defined separately for

the purposes of empirical research.

In fact, it is not clear that a typology based on welfare regimes is the most

appropriate basis on which to study living arrangements and family formation.

Berthoud and Iacovou (2004), in an analysis of Western European family structure,

argue that a typology based on religious affiliation or geography explains family

structure as well or better than one based on welfare regimes, proposing a spectrum

ranging from Northern/Protestant to Southern/Catholic. At one end, the Scandinavian

countries are characterised by small households (particularly single-adult and loneparent

households), early residential independence for young people and extended

residential independence for elderly people; cohabitation as an alternative to

marriage; and an almost complete absence of the extended family. At the other end,

the Southern European countries are characterised by relatively low levels of nonmarital

cohabitation, by extended co-residence between parents and their adult

children, and by elderly people with their adult offspring; this, together with a much

lower incidence of lone-parent families, make for much larger household sizes.

Reher (1998) outlines a typology based on geography and the familialistic legacy of

the Catholic church, in explaining features of family structure across Western

Europe. He describes a “Northern” cluster (Scandinavia, the United Kingdom, the

Low Countries2 and [much of] Germany and Austria), characterised by ‘weak’ family

ties, early home-leaving, and a sense of social rather than familial solidarity with

elderly or weak members of society; and a ‘Southern’ cluster (the Mediterranean

countries, including Portugal) characterised by ‘strong’ family ties, later homeleaving,

and a more family-based sense of solidarity. He notes that Ireland is an

indeterminate case, being geographically Northern, but having much more in

common with the Mediterranean countries in terms of family structures.

Hajnal (1965 and 1982) describes divisions in marriage patterns along a different

East-West axis, with regions east of a line from St Petersburg to Trieste

characterised by near-universal and relatively early marriage, and regions west of

this line characterised by later marriage, with a higher proportion of individuals

remaining unmarried.

In this paper, we use the following fourfold grouping to present our results. The first

group is a ‘Nordic’ cluster consisting of the Scandinavian countries (Sweden,

Denmark and Finland) plus the Netherlands. The second group consists of the preenlargement

countries of North-Western Europe: the United Kingdom, France,

Germany, Austria, Belgium, Luxembourg and Ireland. The third group consists of the

Southern European countries: Italy, Spain, Portugal, Greece and Cyprus. The final

group is an ‘Eastern’ group consisting of the other post-2004 members of the EU: the

Czech Republic, Hungary, Estonia, Latvia, Lithuania, Slovenia, Slovakia and Poland.

Of course, not all countries fall neatly into one or other of these groups. Where there

are intermediate cases, we have positioned these on the edge of a group. The

Netherlands, for example, is defined by Esping-Andersen (1999) as a member of the

Conservative rather than the Social-Democratic welfare regime type; empirically, it is

in some respects closer to our North-Western cluster than the Nordic cluster, and

has been placed on the boundary between the Nordic and North-Western groups.

Ireland has been placed on the boundary between the North-Western group (where

it belongs geographically) and the Southern group (with which it displays a large

number of common features). And Cyprus has been placed on the boundary

between the Southern group (with which it has clear geographical and cultural

commonalities) and the other new EU members.

2 Member States referred to as ‘Low Countries’ are the Netherlands, Belgium and Luxembourg.

As will become clear, the Eastern European countries are very far from forming a

homogeneous grouping. This group may be thought of as consisting of three

subgroups: the Czech Republic and Hungary (which have a good deal in common

with the North-Western cluster); Slovenia, Slovakia and Poland (which are extremely

similar to the Southern cluster; and the Baltic Republics (Estonia, Latvia and

Lithuania), which are in some respects most different to any of the pre-enlargement

countries.

2. Data and methodology

2.1 Data

The data used in this paper are taken from the Community Statistics on Income and

Living Conditions (EU-SILC). The EU-SILC is an annual survey which provides micro

data on a wide range of social indicators including income, poverty, social exclusion

and living conditions. The first release of data (relating to the year 2004) includes

information on 13 Member States (Austria, Belgium, Denmark, Estonia, Finland,

France, Greece, Ireland, Italy, Luxembourg, Portugal, Spain and Sweden), plus

Norway and Iceland. From 2005, Germany, the Netherlands and the United Kingdom

joined, along with the rest of the new Member States (Cyprus, Czech Republic,

Hungary, Latvia, Lithuania, Poland, Slovakia and Slovenia). Finally, from 2007

onwards, the EU-SILC represents all 27 Member States, and includes Turkey and

Switzerland as non-members alongside Norway and Iceland.

The EU-SILC consists of both cross-sectional and longitudinal elements. In most

countries the sample design takes the form of a rotational panel: the sample is

divided into sub-panels, each sub-panel is retained in the sample for a maximum of

four years, and each year one sub-panel is dropped, to be replaced by a new

replication. In most countries the sample is divided into four rotational groups giving

rise to four-year panels; exceptions are France (nine-year panel); Norway (eight-year

panel) and Luxembourg and Sweden (pure panels). The cross-sectional and

longitudinal data are released separately; data for the analysis in this paper are

taken from version 2007-2 of the cross-sectional SILC.

For the purposes of the analysis in this paper, it is important to note that the data

relate to private households only, and that the sample of respondents is drawn from

members of these households currently resident at the time of data collection. Those

living in institutional settings (hospitals, old people’s homes and student

accommodation) are not included as part of the sample. In all countries, a large

majority of individuals do of course live in private households, but for some groups, a

substantial proportion live in institutional settings - most notably older people,

students, and younger men engaged in military service. It is difficult to quantify the

impact of these phenomena; Appendix II attempts to quantify the problem in relation

to the younger groups.

2.2 Defining relationships between individuals

When analysing people’s living arrangements, it is necessary to establish the

relationships between members of households. Many household-level data sets do

this by means of a household grid, which records the nature of the relationship

between each pair of household members. Unfortunately, the EU-SILC does not

collect this type of information, recording instead only the personal identifiers of each

individual’s spouse or partner, mother and father, where these are resident in the

same household. This enables us to identify which people are living as part of a

couple, and/or with their children or parents; we are also sometimes able to identify

sibling and grandparent relationships, where a third person is present. However,

many relationships cannot be identified – for example, we are generally unable to

distinguish a co-resident sibling or cousin from a friend or lodger.

In addition, although we are able to identify the parent/child relationship, there is a

degree of uncertainty relating to the specific nature of that relationship. In particular,

the role of step-parents is not always clear. It appears that the use of the “mother”

and “father” identifiers has not been entirely consistent, so that in some cases they

have been used exclusively to indicate natural parents, while in others they have

been used to indicate step-parents as well. Given the increase in stepfamilies over

recent decades, this is a particularly unfortunate limitation with the data.

We return to these limitations at various points in the paper, and they should be

borne in mind while interpreting our analysis. Nevertheless, the EU-SILC allows a

great deal of interesting analysis on household structure, and it remains a useful,

and in many respects unique, source of data on this topic.

2.3 Methodology

The analysis in this paper is for the most part descriptive – the figures and tables in

sections 3-6 present means over the populations of interest, and compare them

between countries. All country means are weighted using the cross-sectional weights

supplied with EU-SILC.

For most of the analysis, we also present the mean across the EU-15 “old” Member

States, the mean across the nine “new” Member States represented in these data;

and the average across all countries in the sample. Analysts normally use one of two

methods for computing these cross-country means. One method involves calculating

a simple average of the country means (which gives each country, large or small, the

same weight). Alternatively, countries may be weighted according to their

populations, so that large countries make a larger contribution to cross-country

means. We use this second method: countries are weighted according to their

populations.

In this type of analysis, the issue arises of whether means which appear to be

different are actually different, in a statistical sense. We have computed standard

errors for all the figures we present; however, to present them systematically would

run the risk of adding further complication to our already very full tables. These

standard errors are sufficiently small that wherever we note systematic differences

between groups of countries, these differences are statistically significant; however,

smaller differences between countries in the same group may not be statistically

significant. Full tables, complete with standard errors, can be found in Appendix I.

The majority of the analysis presents means over individuals. In the section on older

people, we present means calculated over people aged 65 and over; in the section

on children, we present means calculated over people aged under 18. In each case,

the population over which we have calculated the means is stated clearly in the

footnotes. In some cases, for example, when dealing with concepts which relate to

the household rather than to the individual, it is more appropriate to calculate means

over households. Where we have done this, it is stated clearly in the text and

footnotes.

Finally, two sets of results in the paper use different analytical approaches. In Table

5.1, we present the ages at which young people make a range of life transitions

(moving out of the parental home, living with a partner and having children). These

ages may be interpreted as representing the mean ages at which these transitions

are made; however, they are calculated in a slightly different way. The way in which

these calculations are made is explained in the relevant section. Section 8

synthesises the results from the foregoing sections using principal components

analysis. Again, this technique is explained further in that section.

Results are presented in the form of tables, graphs and maps.

3. Household composition

In this section, we discuss household composition at its broadest level. The first

seven columns of Table 3.1 define seven categories of households, and show how

the prevalence of these household types varies across the EU.

The first column relates to single-adult households, where the adult is aged under 65.

These make up 17.6% of households across the EU: this proportion ranges from 6% in

Portugal up to 30% in Denmark, and in general is lowest in the Southern European

countries, plus Slovenia, Slovakia and Poland; rather higher in the rest of Eastern Europe;

higher still in the North-Western group of countries; and highest in the Nordic group.

The distribution of couple-only households where both adults are aged under 65 is

similar (column 3). These households account for 14% of households across the EU,

with the prevalence ranging from only 8% in Slovenia, up to 20% in Finland.

We turn now to households where at least one adult is aged 65 or over. The distribution

of these household types does not follow our country groupings neatly; this is to be

expected, since many factors contribute to household composition among older people:

typical age differences between partners; differences in life expectancy between men

and women; rates of divorce and separation; and the decision as to whether to live with

adult children or other relatives.

Single-adult households among the 65+ age group (column 2) are most common in the

Nordic and North-Western groups of countries (where divorce is relatively common and

where it is relatively unusual for older people to live with children or other relatives) and

least common in the Southern countries (where divorce rates remain low, and where it

is common for older people to live with adult children). Couple-only households where at

least one partner is aged 65 or over (column 4) are most common in the Southern

European countries (low divorce rates) and least common in Eastern Europe (high

divorce rates, and a high incidence of multigenerational households).

Column 5 relates to all other households where children under 18 are not present. In all

countries, the majority of these are households containing both parents and their adult

children; however, in the Southern and Eastern European countries, a substantial

minority of households are composed differently – for example, with a couple plus

another adult of similar age, who may be a sibling. These households are most common

in the Southern European countries plus Slovenia, Slovakia and Poland; they are less

common in the North-Western countries, and much less common in the Nordic cluster,

where they account for only 4% of households in Denmark.

The remaining household types relate to households with children under 18. Those with

a single adult (i.e. lone parent households, column 6) are in a minority everywhere,

being most common in Ireland and the UK (7% and 5% of households respectively), as

well as in Sweden and Finland and the Baltic states. Lone parent households are least

common in Southern Europe plus Slovenia, Slovakia and Poland.

Finally, Column 7 relates to households where two or more adults are living with

children. These are not necessarily two-parent families; some are one-parent

families with adult children as well as minor children; or they may be extended

families with children. These account for 25% of households across the EU-25; while

single-adult families with children are considerably more numerous in the Northern

than the Southern countries, the opposite is true of families with more than one adult,

which are most common in the Southern countries, plus parts of Eastern Europe,

and least common in the Nordic countries.

The final two columns in Table 3.1 are concerned with mean household size.

Column 8 shows mean household size using the individual as the unit of analysis;

Column 9 calculates the mean over households, and thus provides smaller means,

because larger households are only counted once. Mean household sizes are lowest

in the Scandinavian countries, and also low in the North-Western countries, with the

exception of Ireland. The two different methods of calculating mean household sizes

produce slightly different rankings for the largest household sizes. Taking the mean

over households, the largest households are seen in the Southern European

countries, plus Ireland, Slovenia, Slovakia and Poland. If the mean is taken over

individuals, on the other hand, the Eastern European countries are those with the

largest household sizes: this is because the Eastern European countries have more

very large households than the Southern European countries.

4. Children

Children’s living arrangements are of interest to social scientists because of their

relationship to child poverty and to outcomes in later life. We begin this section with a

short discussion of childlessness; we then present statistics on family size and

children’s living arrangements.

Childlessness is increasing throughout the EU, and is contributing to the belowreplacement fertility rates which are evident in many countries, and to an associated level of anxiety related to the dependency ratio (Sobotka 2008). We are not able to calculate rates of childlessness directly from the data we are using here, because while EU-SILC collects information on all children living in respondents’ households, it does not collect information on respondents’ children who are no longer living in the parental home. We therefore estimate a different indicator of childlessness, using a sample of women aged between 33 and 37 years old. These women are old

enough to have had most of the children they are ever going to have; and they are

young enough that only a very small proportion of their children will have left home.

Thus, the proportion of women in this age group who have none of their own children

living with them approximates well to the proportion of this cohort who are childless.

This proportion is lowest in Lithuania at 7%, and generally across the Eastern

European countries plus Ireland. It is highest in Italy and Spain at over 33%, as well

as in a number of Nordic and Northern countries. Thus, although childlessness is

increasing across the Eastern European countries (Philipov and Dorbritz 2003),

childlessness across this cohort has not yet risen to the levels seen elsewhere in

Europe.

Comparing these figures on childlessness with figures on total fertility rates (column

2), there is no clear relationship between the two series. In several countries, the

relationship runs along the lines which one would expect. France, Luxembourg and

Ireland have low levels of childlessness and high TFRs; Italy, Spain and Germany

have high levels of childlessness and low TFRs. However, across Eastern Europe,

levels of childlessness are low, but total fertility rates are also low – this may be

attributed to postponement (and possible permanent reductions) in fertility among

the youngest cohorts. By contrast, in a number of Nordic and Northern countries,

rates of childlessness among the 33-37 age group are very high, while TFRs are

also relatively high (though generally below replacement rate). This has to do with

(a) the postponement of fertility being a less recent phenomenon in these countries,

and (b) the fact that larger families are more common in these countries.

The four columns on the right-hand side of Table 4.1 present figures on the number

of children present in households. These are calculated using the sample of

households where any child under 18 is present; it is important to remember (a) that

these are means over households rather than individuals, and (b) that they do not

include any offspring who are not currently resident in the household, or any

offspring over age 18, even if they are resident in the household. Thus, these figures

will tend to underestimate the proportions of larger families, particularly in those

countries where home-leaving takes place earlier; however, they are indicative of

cross-country variations in family size.

The very largest families are found in Ireland, where 21% of families have three or

more children, and where 5% of families have four or more children. The next largest

families are found in Belgium and the Netherlands, followed by the rest of the Nordic

cluster. The smallest families, based on the percentage of households with three or

more children, are found in Spain, Portugal, Greece and Italy – in these countries,

under 7% of households have three or more children. These countries, in common

with a number of other Eastern European countries, also have a relatively large

number of households with only one child.

We turn now to a “child’s-eye” view of living arrangements. Declining marriage rates,

rising rates of cohabitation and high rates of union dissolution –trends which have all

been a feature of recent decades - mean children may spend time growing up in a

number of different household types (e.g. lone parent households, cohabiting couple

households). Table 4.2 shows the proportions of children (i.e. those under age 18)

living in four such situations: living with one parent; with two parents who are

cohabiting but not married; and two parents who are married to each other. There

are also a small number of children who are not living with either natural parent; we

include these in the table for completeness.

Because the EU-SILC data do not allow us to distinguish fully between natural

parents, “official” step-parents, and other co-resident partners, we have allocated

children to these categories as follows. The “one parent” category includes all

children living with only one parent, where that parent does not have a co-resident

partner. The “two parents, cohabiting” category includes children living with two

parents who are cohabiting rather than married, as well as children living with one

parent who is cohabiting with a partner who is not defined as the child’s parent. The

“two parents, married” category includes children living with two parents who are

married, as well as children living with one parent who is married to an adult who is

not defined as the child’s parent. Despite these limitations, our findings are similar to

those of (e.g.) Perelli-Harris et al (2009), who cover fewer countries with better data.

Starting with the left-hand column of Table 4.2, we observe that very few children in

any country live with adults who are not recorded as their parents – the figures range

from 0.3% in Luxembourg and the Netherlands to 3.3% in Latvia. Table 4.2 is based

on a sample of all under-18s, and some of those recorded as living with no natural

parents will be teenagers who have moved out of their parents’ home. These

account for about one quarter of those recorded in this column.

The second column of Table 4.2 presents the percentage of children living with a

lone parent. This percentage ranges from 7% or less in Greece, Spain and Cyprus,

up to 24% in Ireland and 27% in Latvia. As we have seen before, there is a high

degree of heterogeneity within the Eastern European group: in the Baltic republics,

the rates of lone parenthood are among the highest in Europe, while in Slovenia,

Slovakia and Poland, they are among the lowest. These figures may usefully be

compared with the figures on lone parent households in Table 3.1: in general, those

countries with a large proportion of lone parent households are the same countries

with a large proportion of children living with a lone parent, and vice versa.

Looking at the percentage of children living with two parents in a cohabiting union

(column 3), this follows a broadly similar pattern, with high proportions of cohabiting

partnerships in the Nordic countries and the Baltic republics, and low proportions in

the Southern European countries plus Slovakia and Poland.

Finally, we look at the percentage of children living with two parents who are

married. Over 90% of children in Greece and Cyprus live with two married parents;

the proportions are also high across the rest of Southern Europe and in Slovakia and

Poland. By contrast, only half of all children live with two married parents in Sweden,

largely thanks to the high proportion whose parents are cohabiting rather than

married; this is also the case in Estonia and Latvia, where high rates of lone

parenthood also play a role.

The final column of Table 4.2 shows the percentages of children who live in

multigenerational households (defined here as households where grandparent(s) as

well as parent(s) are present). There is a clear regional gradient here. Well under 1%

of children in the Nordic cluster live in multigenerational households; 1-5% of

children live in multigenerational households in all other North-Western countries

except for Austria (where the figure is higher); and around 6% of children live in

multigenerational households in Southern European countries (except in Portugal,

where the figure is 11.6%). However, in Eastern Europe, the figures are much

higher: over 10% of children live in multigenerational households in all countries

except the Czech Republic, and this rises to over 20% in Poland and Latvia.

5. Young adults

The transition from childhood to adulthood is characterised by a number of

transitions: from the parental home to living independently; from the single state to

living with a partner; and from childlessness to parenthood. Not all young people

make all these transitions, and some never make any; however, the majority do

make some of these transitions in their twenties or thirties. These transitions have a

direct relationship with young people’s wellbeing and life chances: making these

transitions at an early age is associated with early independence, but may also

(particularly in the case of early home-leaving or early childbearing) be associated

with an increased risk of poverty and disadvantage (Aassve et al. 2007). By contrast,

the very late transitions observed in the Southern European countries, while being

protective against poverty, may delay independence and may also be burdensome

for the parents of young people (Schizzerotto and Gasperoni 2001).

Because some of these transitions are reversible – young people may leave home

and move back in again, or they may live with a partner for a short time before

subsequently splitting up, it is difficult to calculate the mean or median ages at which

these transitions are made by observing the transitions themselves. Instead (taking

home-leaving as an example), we assume that young people who are currently

observed living with their parents have not made the transition out of the parental

home, and we assume that those currently observed as living independently have

made the transition. Of course, we will count some young people who have left home

and come back again as not having yet made the transition; and we will count some

people who are living away from home but for whom the transition is not permanent as

having made the transition. But these errors are likely to cancel each other out. We

then use non-parametric regression techniques to calculate the age at which 50% of

all young people are observed living away from home, or living with a partner or with

children, and consider this analogous to the median age of making the transitions.

Before discussing these figures further, it is worth pointing out that they are based

only on young people living in private households – those living in institutional

settings such as military barracks or university residences will not be sampled. We

believe our results are reasonably robust to these issues: see Appendix II.

The results of these calculations are shown in the first six columns of Table 5.1.

Results are shown for men and women separately, because women tend to make all

these transitions at an earlier age than men.

The first two columns show the age by which half of all young people have left the

parental home. There is a strong divide here between the regional groupings we have

defined: for both men and women, the countries where the transition takes place

earliest are those of the Nordic cluster; the remainder are in the North-Western group

of countries. The transition takes place latest in the Southern European countries, plus

three of the Eastern European countries: Slovenia, Slovakia and Poland; homeleaving

is also relatively late elsewhere in Eastern Europe. The range is very large

indeed: 50% of women have left home by age 20 in Finland and Denmark, while the

corresponding age in many Southern European countries is 27 or 28.

The next two columns show the age by which 50% of young people are living with a

partner. The regional patterns are essentially the same as those for home-leaving,

with partnering taking place relatively early in the Nordic and North-Western

countries, and relatively late across Southern and Eastern Europe. However, the

differences are not so stark in terms of the ages at which the transitions are made. In

the Nordic countries, the median age at partnering is several years higher than the

median age at leaving home, indicating that a prolonged period of living alone is the

norm in these countries; while in the Southern and Eastern European countries, the

mean ages at leaving home and partnering are much closer together, typically

around only one year apart. In the case of Poland and Slovakia, partnership on

average occurs earlier than home-leaving, indicating that it is common for young

adults to remain living with their parents while they also live with a partner.

Finally, we look at the age at which young people live with their own children

(columns 5 and 6 of Table 5.1). For women, this approximates well to the median

age at first birth; for men, the approximation is less good, because some men father

children they do not live with. Here, the pattern of cross-national variation is different,

with the earliest childbearing evident in Cyprus plus the Eastern European countries;

childbearing is relatively late in the Nordic cluster plus some of the North-Western

countries, but latest of all in Italy and Spain, where the median age for a first birth

calculated in this way is 32 for women and 36.5 and 35.5 respectively for men.

The last two columns in Table 5.1 indicate the percentage of men and women aged

between 18 and 28 who live alone in each country. These figures range from only a

few percentage points in Slovenia, Slovakia and Poland, to over 20% in the

Scandinavian countries, and over 30% for both sexes in Denmark. This reflects the

difference in the median ages of leaving home and living with a partner which we

mentioned before: where the gap between these two ages is small, the percentage

of young people living alone is also small, and where the gap between the two ages

is large, this is reflected in a high proportion of young people living alone.

6. Partnerships: cohabitation and marriage

One area in which there are substantial differences between Northern and Southern

European countries is in the prevalence of cohabitation as a substitute for marriage

(Kiernan 1999): non-marital cohabitation is far more common in Northern than in

Southern European countries, particularly in the Nordic countries, where it is very

much the norm among childless young people.

Table 6.1 shows the percentage of opposite-sex partnerships which are reported as

cohabiting rather than marital partnerships in each country, for four age groups:

couples where the woman is in her twenties, her thirties, her forties and her fifties. For

each age group, two sets of figures are reported: the first for partnerships where there

are no co-resident children, and the second for partnerships where the children of one

or both partners are resident in the household. It should be noted that this is not a

perfect indicator of couples who have children – many couples in their fifties, and

some in their forties, will have children who have moved away from the parental home,

and will thus not be counted as having children in the data.

It is clear that there is a substantial age gradient in all countries, with couples in their

twenties substantially more likely to be cohabiting than couples in their forties and

fifties. These figures do not allow us to separate out age effects (sample members in

their twenties have not got married yet, but many will) from cohort effects (people born

in the 1980s are less likely to get married, ever, than people born in the 1950s).

However, some combination of these two effects is leading to a strong gradient:

across the EU as a whole, 63% of childless partnerships among people in their

twenties are cohabiting, compared with just 8% of childless partnerships among those

in their fifties; for partnerships where children are present, the corresponding figures

are 28% for those in their twenties, against 3% for those in their fifties.

A steep north-south gradient is also evident from Table 6.1. In the Nordic countries,

well over half of all childless couples in their twenties and thirties are cohabiting; in the

other Northern European countries, the proportion cohabiting is lower, but still high,

while it is much lower in Southern Europe ranging from 7% of childless couples in their

thirties in Greece to 29% in Portugal. Levels of non-marital cohabitation in the Eastern

European countries are rather heterogeneous, being as low as Southern European

levels in Poland, Slovakia and Lithuania, and comparable with Nordic levels in

Estonia.

There are also strong differences between couples with and without children: in all

countries, for all age groups, couples with children are less likely to cohabit than

couples without children, and in nearly all cases these differences are large. However,

in some countries these differences are systematically larger than in others. Figure 6.1

shows the percentage of cohabiting unions by age group for Sweden and the

Netherlands. For childless couples, the percentage in cohabiting unions is somewhat

higher in Sweden than in the Netherlands, but the numbers are not hugely dissimilar:

91% against 86% for those in their twenties, and 45% versus 38% for those in their

forties. However, when we turn to look at couples with children, much larger

differences emerge: for those in their twenties with children, 69% of Swedish couples

are cohabiting, compared with only 35% in the Netherlands; for people in their forties,

the figures are 29% versus 9%.

This difference between couples with and without children does not follow predictable

regional lines. The difference does tend to be smaller where cohabitation rates are

higher (Sweden, Denmark, Estonia and Slovenia) – but the difference is large in the

Netherlands, Germany and Austria (where cohabitation rates are high) and also in

Cyprus, Greece and Slovakia (where cohabitation rates are low).

7. Older people

Increasing life expectancy and declining fertility mean that the elderly are set to form

a progressively larger proportion of our population over future decades. Older

people’s living arrangements are of key interest to policy-makers: as well as being a

key determinant of older people’s well-being, living arrangements are related to

levels of social expenditure on elderly people.

Table 7.1 shows the proportion of older people living in four situations: alone; without

a partner but with other people; with just a spouse or partner; and with a spouse or

partner plus other people. Before commenting on the table, it is worth noting that

these figures relate to older people in private households: older people in institutions

such as nursing homes are not sampled by the EU-SILC and are not included in this

analysis.

Each set of figures is calculated separately for men and women, and the differences

between the sexes are starker here than elsewhere in this report, because of

differences in life expectancy between men and women, and the consequently

higher proportion of elderly women who are widowed. As we mentioned in Section 3,

the proportion of older people who are living with and without a partner is also

related to the prevalence of divorce and separation in each country.

Two “ideal types” are visible. In the Scandinavian countries plus many Northern

European countries, in particular Germany and France, the predominant living

arrangement for older people is either with a spouse or partner, or alone. Typically,

living in a household with anyone except a spouse or partner accounts for only 10%

or less of older people.

In the Southern European countries, by contrast, it is much more common for older

people to live with people other than a partner: in Spain, 42% of older women and

40% of older men live with others. This type of living arrangement is also relatively

common in the new Member States, particularly Latvia, Slovenia and Poland.

Using EU-SILC data it is not possible to determine the relationships of older people with

the others with whom they live in every case. However, in every country, the large

majority of older people who are observed living with people other than a spouse or

partner, are observed living with at least one of their adult children. These are generally

not the same households which form the group considered in Section 5, of young adults

living with their parents; in most cases, the parents in these households would be too

young to be included in the analysis in this section. The relationship between these

groups is worthy of further analysis. In one sense, the groups are clearly related, in that

they are both composed of adults in the same household as their parents; moreover, they

both tend to be found in the same groups of countries. However, there is a conceptual

difference between the two household types. In the case of adult children living with their

parents, it is the parental generation who may be thought of as providing support for their

children; in the case of elderly people living with their adult children, the locus of support

may often be the other way round, with the younger generation providing not only the

majority of the income, but also an element of care for the parents.

8. Synthesising the differences: Factor analysis

From the figures in the preceding sections, a number of patterns have emerged. One

way in which these may be synthesised is via the use of factor analysis. Principal

components analysis identifies three main factors, which together explain 83% of the

variation between countries in the factors explored. Factor loadings are given in

Table 8.1, with the most important loadings being highlighted via shaded cells. We

identify the first factor as being related to the importance of the extended family: the

variables contributing positively to this factor are young adults living at home; older

people co-resident with their own children, household size, and multigenerational

households. Negatively related to this first factor are young adults living alone and

prime-aged people (i.e., adults aged 35-64) living alone.

If the first factor relates to the extended family, the second factor may be thought of

as relating to the stability of the intimate relationship. The only variables which are

related to this factor are babies living with a lone parent; children living with a lone

parent; prime-aged people who are divorced or separated (and not living with

another partner) and old people living alone. This variable does appear to be related

to the stability of the intimate relationship rather than to notions of social liberalism,

since cohabitation as an alternative to marriage makes no contribution to this factor

at all.

The third factor relates to fertility, with childless women making a negative

contribution, and the number of children per woman making a positive contribution.

Factors 1 and 2 are plotted on Figure 8.1. Six clusters of countries have been

identified. Clearly, there is no unique way of identifying these clusters – clusters

towards the centre of the graph could be combined, as could the two clusters in the

north-east of the graph. First, we note that the “old” EU-15 form the clusters which

might have been expected based on previous research. The social-democratic

countries (including the Netherlands) form one group, scoring low on the extended

family and high on the relationship stability axis. The Southern European countries

score high on both the extended family axis and the relationship stability axis, while

the remaining countries of North-Western Europe occupy an intermediate position on

the extended family axis, and score generally lower on the relationship stability axis.

Ireland occupies a position slightly apart from this group, scoring almost as high on

the extended family axis as the Southern European countries, and low on the

relationship stability axis.

The new Member States are rather heterogeneous. Cyprus falls very close to the

other Southern European countries, which is to be expected given commonalities of

geography, language and culture. Three of the Eastern European countries display

similar, but more extreme, characteristics to the Southern European group, scoring

even higher on the extended family axis and at similar very high levels on the

relationship stability axis. These countries are Poland, Slovakia and Slovenia, all of

which have maintained a Catholic tradition through the Communist years (see

Appendix II).

The remaining countries include the three Baltic states – Latvia, Lithuania and

Estonia – and the Czech Republic and Hungary. All these countries occupy a more

“south-easterly” position on the graph than the other countries, scoring high on the

extended family axis, but low on the relationship stability axis. Ireland – previously an

outlier in relation to the other North

the Czech Republic and Hungary.

Figure 8.1: Clusters arising from Principal Components Analysis

Source: EU-SILC Users’ database. See Table 3.1

These results are fairly robust to the particular variables included in the analysis. In

particular, we experimented with different formulations of the variables indicating

divorce, since it was unexpected (to us at least) that the Scandinavian countries,

which score rather high on the relationship

highest divorce rates in the world. In fact, it appears that this factor does not relate to

divorce per se, but rather to the proportion of people living alone followin

separation (and similarly, to the proportion of children living with an unpartnered

parent following divorce or separation). It seems that the Scandinavian countries,

while having high divorce rates, also have relatively high rates of subseque

repartnering, and thus have a much lower proportion of divorced or separated adults

still living alone. We also explored the phenomenon of cohabitation in some detail;

we had been expecting this analysis to generate a factor indicating social liberalism

which would be explained by cohabitation as well as by divorce and lone

North-Western countries – occupies a position close to

blic : stability axis, while they have some of the , 29 following divorce or subsequent liberalism, parenthood. However, we were unable to formulate any indicator of cohabitation which contributed significantly to any such factor; the second factor remained stubbornly as an indicator of partnership breakdown without subsequent repartnering.

9. Conclusions

In this paper, we have mapped a range of indicators of household structure across

the expanded European Union. One of our main aims has been to focus particularly

on the newer Member States of the EU, and to assess the extent to which household

structures in these countries display similarities and differences to household

structures in the “old” EU-15.

Of the new Member States, we find that Cyprus is extremely similar to the Southern

European countries, as might be expected with reference to cultural, geographic and

religious factors. We also find that there is a great deal of heterogeneity among the

Eastern European countries. One group of countries – Slovenia, Slovakia and

Poland – are consistently very similar to the Southern European countries. In these

three countries, the extended family is the norm: young adults leave home late, older

people often live with their adult children, three-generational households are

common, and lone-parent families are relatively uncommon. In terms of mapping

onto a geographical/religious spectrum, Slovenia is the only one of these countries

which is geographically Southern, but all three of these countries remain strongly

Catholic or Orthodox.

The Czech Republic and Hungary, by contrast, have more in common with the

countries of the North-Western cluster. On a large number of indicators, these

countries occupy an intermediate position between the Nordic cluster on the one

hand, and the Southern/Catholic cluster on the other; and in the factor analysis, they

occupy a position close to the other countries of the North-Western cluster –

particularly Ireland.

Of the Eastern European countries, it is in the Baltic countries where family patterns

diverge most widely from the geographical/religious spectrum. These countries

display a number of features in common with the Southern European countries;

chiefly, a large number of large and multigenerational households. However, they

also display a number of striking dissimilarities with the Southern European

countries, particularly in terms of the very large numbers of lone-parent families, and

other single-adult households. In many respects, the Baltic states are very

heterogeneous: for example, non-marital cohabitation is much more common in

Estonia, and very much less common in Lithuania; while lone parenthood and

multigenerational households are more common in Latvia than in the other two Baltic

states.

In this paper we have answered a number of questions, but these in turn raise

further questions. One question, which we raised in Section 7, relates to the nature

of multigenerational households. We have shown that, in a swathe of countries

across Southern and much of Eastern Europe, co-residence between generations is

very common, particularly so in contrast to the Nordic group of countries, where it is

extremely unusual. We have shown that this co-residence is manifested both by

young adults remaining in the parental household, but also by older people living

with their adult children. The question which this raises, and which in this analysis we

have not yet been able to answer, is whether the second household type is merely a

persistent form of the first (i.e. that the young adults whom we see living with their

parents become the same prime-age adults who live with their elderly parents) or

whether the two household types are in fact drawn from different social groups.

Two other questions also arise relating to multigenerational households. The first is

the extent to which they arise as a result of social and cultural preferences (people

actually like living with other family members, and make a positive choice to do this)

as opposed to arising as a result of economic constraints (young people who would

like to leave the parental home but cannot afford to; or older people who cannot

afford to live alone). There is limited evidence to suggest that in Southern European

countries, at least part of young people’s extended residence in the parental home

arises from preferences (Manacorda and Moretti 2006). However, this has not been

systematically addressed for other age groups, or across the new Member States.

Another question which we might pose in relation to multigenerational households is

the degree to which individuals are supporting each other, both economically and in

other ways, by living together. The assumption tends to be that when young adults

are living with their parents, it is the parents who are supporting the young adults;

while when very elderly individuals are living with their adult offspring, it is the elderly

people who are receiving that support. However, this too has not been systematically

tested, particularly in the context of the expanded European Union.

Finally, the picture we have presented has been essentially static: we have not

addressed the important issue of how household structures are evolving (Billari et al

2002). We are unable to answer this question definitively with the cross-sectional

data we have at our disposal; however, we may make inferences based on evidence

drawn from elsewhere. As far as attitudes are concerned, there is some evidence

that these are converging across Europe, with smaller between-country differences

among younger people than their parents. Rosina and Fabroni (2004) find that

although marriage remains the predominant route out of the parental home in Italy,

cohabitation is becoming more prevalent and is set to increase further. Billari (2005)

investigates trends in a range of indicators, concluding that there is limited evidence

of general convergence in attitudes and behaviours, but that on a number of

indicators, convergence is apparent. A further clue towards the evolution of living

patterns lies in the fact that incomes in the new Eastern European Member States

are growing faster than those in the “old” EU-15 (Van Kerm, 2009). To the extent that

behaviour is driven by economic factors – for example, to the extent that intergenerational

co-residence is driven by economic constraints – this suggests that

again, we may observe a degree of convergence in living arrangements between the

countries of the European Union.

Family structure and children’s living conditions. A comparative overview of 23 countries

2011-06-08T00:11:00.000-07:00

Abstract

Partly due to differences in household resources, children in single-mother households tend to report lower own resources compared with children living with two original parents, e.g. in terms of health, educational performance and educational attainment. The paper uses large-scale cross-national data of 23 countries to analyse the implications of living with a single mother. Three central areas of children’s living conditions are focussed: social support, health, and economic resources. The aim of the study is to assess: 1) whether there are differences between children who live with a single mother and children who live with two original parents within these areas; and 2) whether the association between family type and the resources varies systematically across countries. The data come from the international WHO study Health Behaviour in School-aged Children (HBSC) of 2001/02. Surveys have been conducted with nationally representative samples of school pupils of 11, 13, and 15 years of age in countries in Europe and North America. After relevant selections, total n=92,564. The main results are that children in single-mother households have lower social support from parents, less good health, and less economic resources than children living with two original parents. Findings are consistent for all included indicators and the general tendency is consistent across countries. There is no clear pattern in how the association with family type varies between countries, e.g., a low poverty rate among children in single-mother households relative to children in two-parent households is not associated with small family type differences in children’s resources.

Keywords Children; family structure; single-mother household; cross-national

Introduction

Divorce is a common phenomenon in Europe as well as North America. Large numbers of children have experienced parental separation and most of these continue to live with their mother, implying that single-mother households constitute a frequent family type (Andersson 2002a). The shares of children living with lone mothers have increased in a wide range of Western countries (Beaujot and Liu 2002). The present paper does not focus on consequences of divorce per se, but on implications of family structure for children, more specifically on characteristics associated with living with a single mother compared with living with two original parents. Three central areas of children’s living conditions are analysed: social support, health, and economic resources.

Although there is a large body of work on the implications of family structure for children, crossnational studies are scarce, and in particular studies that include a broader set of indicators which are based on reports from children themselves. A recent report on single parenthood and child well-being in the OECD includes a meta-analysis and a literature review (Chapple 2009), but a majority of the previous cross-national studies focus on family type differences in academic achievement (e.g., Schiller, Khmelkov and Wang 2002; Pong, Dronkers and Hampden-Thompson 2003; Hampden- Thompson and Pong 2005; Marks 2006; Garib, Martin Garcia and Dronkers 2007; de Lange, Dronkers and Wolbers, forthcoming).

The present paper uses large-scale cross-national data of 23 countries to assess, firstly, whether a range of living conditions of children here and now – as reported by the children themselves – differs between those who live with both original parents and those who live with a single mother, and secondly, whether the association between family structure and children’s living conditions varies systematically across countries. The living conditions which are focussed – social support, health, and economic resources – are three central aspects of children’s welfare, which are also intertwined.

Children in families suffering from economic stress have more psychosomatic complaints (Östberg, Alfvén and Hjern 2006) and less good social relations (Bolger et al 1995; Olsson 2007) than children in families with greater economic resources. There is also a connection between social relations and health. Strained social relations with and lack of social support from parents and peers are linked with more psychological and physical complaints (Due, Lynch, Holstein and Modvig 2003; Brolin Låftman and Östberg 2006) and ease of communication with parents is associated with better self-rated health (Pedersen, Granado Alcón and Moreno Rodriguez 2004).

Differences in resources between two-parent and single-mother households

The theoretical point of departure is that there are differences in resources between two-parent and single-mother households, which have consequences for children’s living conditions.1

Firstly, economic resources differ between the two household types. Single-mother households have lower income than two-parent households (Raschke 1987; Gähler 1998; Bradbury and Jäntti 1999; Beaujot and Liu 2002). The economic status of the household has been shown to be connected with a wide range of child outcomes, such as school attainment and achievement, internalising and externalising problems, the child’s own economy and material resources, social relations, leisure activities, health and well-being (see reviews by Voydanoff 1990; White and Rogers 2000, see also Jonsson and Östberg 2004).

Secondly, single-mother households have on average lower social class and lower education compared with two-parent households. This is due to the fact that there is only one, instead of two, adults to derive highest parental education and household social class from (Jonsson and Gähler 1997) and the non-resident father often has the highest class position.2 Because there is a strong association between parents’ and children’s educational level it is not surprising to find that children in single-mother households have lower educational achievement and attainment compared with children in intact families (for the US, e.g. Sandefur and Wells 1999; Biblarz and Gottainer 2000; for Sweden, Jonsson and Gähler 1997; Ringbäck-Weitoft, Hjern and Rosén 2004; for Germany, Mahler and Winkelmann 2004). Furthermore, social class is associated with a number of child outcomes including health related behaviours such as smoking and physical exercise (see e.g. Jonsson and Östberg 2001) as well as health (for a review of the Swedish case, see Bremberg 2002).

Thirdly, there is a difference in social resources between single-mother households and two original parent families. Single mothers can be expected to both provide and receive less social support than two parents who live together. Single mothers are likely to suffer from greater time stress than individuals with a partner with whom they can share household and other tasks, and parents’ time constraints imply smaller possibilities of monitoring and socialising children (Astone and McLanahan 1 In the present paper, I focus only on single-mother households vs. two original-parent households. Even though reconstituted families are an increasingly common phenomenon, living with both original parents or with a single parent are the two most common family types for children in all countries studied (yet with large cross country variation in absolute percentages) (Pedersen et al 2004, Figure 2.7). Children who live with single fathers are excluded as they constitute a very small group for which it would not be possible to obtain robust results with the data used here.

2 It is however not evident that children living with a single mother should automatically be categorised

according to the mother’s social class, as they may be in contact also with the non-residential father, and perhaps even live with him for part of the time. Obviously the range of “father contact” may be wide both within and between countries.

1991; Sandefur and Wells 1999; Carlson and Corcoran 2001), and of providing social support to the

children (Dornbusch et al 1985). There are also differences in access to social support between

married and single mothers (Gähler 1998). Single mothers have smaller social networks than mothers

in two-parent families (Gunnarsson and Cochran 1999). The existence of social networks is positively

associated with the individual’s health (see e.g. Cohen and Syme 1985). It also seems likely that

parents’ social networks are beneficial for the children in various ways. For instance, Coleman (1988)

argues that intergenerational closure, i.e. the fact that the parents know the parents of the children’s

friends, has a positive effect on school achievement because parents who know each other monitor

each others’ children.

Fourthly, single mothers have lower psychological well-being than married mothers and fathers (for

reviews, see Raschke 1987; Kitson and Morgan 1990). This may be due to stress from the divorce, or

from economic hardship and limited social resources (Carlson and Corcoran 2001). Single mothers

also have less good self-rated health and higher mortality than mothers with a partner (Fritzell,

Ringbäck Weitoft, Fritzell and Burström 2007). Parents’ depression has been shown to have negative

effects on outcomes related to children’s adjustment problems and depression (see review by Downey

and Coyne 1990).

To summarise, it is likely that children in single-mother households have less resources than children

living with two original parents, in particular due to single-mother households’ smaller financial

resources, lower average educational level and lower social class, greater time constraints, more

limited access to social support, and lower psychological well-being. However, it deserves to be

stressed that it does not follow that single-parent households have any intrinsic characteristics that are

detrimental to children’s well-being. Instead, the differences in household resources have implications

for the children. As pointed out clearly by Wu, Hou and Schimmele (2008), it is not family structure

per se, but the day-to-day familial circumstances that matter. And circumstances such as economic

deprivation, ineffective parenting, parent absence, and exposure to parental conflict are more prevalent

in single-headed households and reconstituted families than in married original-parent households

(Wu et al 2008).

Cross-national variability in effects of family structure on children’s living conditions?

There are reasons to believe that differences in resources between two-parent and single-mother

households exist in most if not all countries. However, these differences are likely to vary due a

number of factors, including cross-national differences in absolute divorce rates and in the selection

into divorce, differences in stigmatization, and variations in economic circumstances, which is closely

linked to differences in social policies.

The absolute divorce rates differ widely across countries in Europe and North America (Andersson

2002b; Kalmijn 2007), meaning that children’s experience of family dissolution varies substantially

across countries (Andersson 2002a), as well as the amount of children who live in single-mother

households (see Figure 1). Even though parental separation or divorce is one main reason behind

single motherhood, there are also other factors, which differ cross-nationally: the shares of single

mothers who never married or cohabited, who separated or divorced, and who are widowed in fact

vary considerably between countries (see Chapple 2009, Table 4).

[FIGURE 1 ABOUT HERE]

The rate of single-mother households in a country may be of importance for children’s outcomes. In

their conceptual framework of why the effect of family type on academic achievement would differ

across countries, Pong et al (2003) as well as Garib et al (2007) point at the importance of the share of

single mothers in a country. They refer to the “family conflict hypothesis” (Pong et al 2003, p. 684),

according to which the selection of families that end up in divorce differs depending on norms and on

how common divorce is: in countries with low divorce rates, only high-conflict marriages will end up

in divorce, while in countries where divorce is common, also the “moderately dissatisfied” end up in

divorce. According to this hypothesis, it follows that “in nations where divorce rates are high, children

from single-parent homes only have the negative consequences of the parental divorce (greater

psychological stress) and fewer of the positive ones (less serious parental divorce) than do their

counterparts in countries where divorces are relatively rare, who have both the negative and positive

consequences of parental divorce.” (Garib et al 2007, p. 31). Both Pong et al (2003) and Garib et al

(2007) find empirically that the family type performance gap is greater in countries where singleparent

households are more prevalent.

The social selection into divorce also varies across countries, as suggested already by Goode (1962).

He argued that where divorce is difficult and costly, it is a phenomenon recurrent mostly in the higher

social strata. Goode’s argument receives empirical support still today, at least partially. In their

analysis of the effect of women’s education on marriage disruption in Italy, Germany, and Sweden,

Blossfeld et al (1995) conclude that the effect of education is strongest in Italy and weakest in Sweden

with Germany in between, meaning that in Italy divorce is most common among highly educated

women, whereas in Sweden women of all social strata divorce. Härkönen and Dronkers (2006)

conduct a similar analysis of 17 countries, and find a positive association between women’s education

and divorce in France, Greece, Italy, Poland and Spain, a negative association in Austria, Lithuania

and the USA, and no association in Estonia, Finland, Germany, Hungary, Latvia, Sweden,

Switzerland, Flanders and Norway.

Kalmijn and Uunk (2007) find empirical support to the hypothesis that there is a stigma associated

with divorce in regions where divorce is less tolerated, as compared with regions where it is more

tolerated. They conclude that there is variability in stigmatization between countries in Europe but also

between regions within countries (Kalmijn and Uunk 2007). It seems possible that children of

divorced parents who live in locations where there is a stigma attached to it suffer more than those in

countries and regions where it is very common and tolerated.

Finally, there is considerable cross-national variability in the economic circumstances of single-mother

households (Bradbury and Jäntti 1999). This is due to differences in welfare systems, and, which is

related, to differences in single mothers’ labour market participation rates (Uunk 2004; Iversen,

Rosenbluth and Soskice 2005). At least two sets of factors may be at work here, namely, what Uunk

(2004) labels income-related welfare provisions and employment-related welfare provisions. Incomerelated

welfare provisions aim at directly affecting the post-divorce economic situation through

measures such as general social welfare and special allowances for single-parent families.

Employment-related welfare provisions aim at improving women’s post-divorce labour market

position, e.g. through employment programs for the inactive or for re-entry to the labour market, and

through public childcare provision (Uunk 2004).

There is great cross-national variability in the generosity of social expenditure, which according to

Beaujot and Liu (2002) is of importance for the economy of single-mother households. The share of

social expenditure that is targeted for children or families also varies between countries (Beaujot and

Liu 2002).

The association between single mothers’ labour market participation and children’s outcomes is

complex and it is not evident whether the association should be expected to be positive or negative.

High rates of single mothers’ labour market participation imply a better financial situation of these

households. At the same time, it means that the single mother is away from home and therefore has

less time for the children. Yet again, several studies indicate that parents’ unemployment has negative

implications for children (for a review, see Madge 1983). For instance, a negative association between

parental unemployment and children’s psychological well-being has been found (Hagquist and Starrin

1994).

Yet another factor which may play a role for differences in the economic situation of single mothers is

the fact that there are cross-national differences in the obligations of non-resident fathers to financially

support their children (Kunz, Villeneuve and Garfinkel 2001).

Consequently, poverty rates of children in both single-mother and two-parent households vary

substantially across countries. Obviously, there is a myriad of ways in which to define “poverty”. A

basic distinction is the one between absolute and relative poverty. Measures of relative poverty take

into account where in the income distribution an individual or a household is found. In contrast,

measures of absolute poverty are defined by a fixed amount of money which is said to be sufficient for

a certain standard of living. Here, figures of relative poverty rates from the Luxemburg Income Study

(LIS) are used, and presented in Figure 2. The bars illustrate the shares of children who live in poor

households – defined with a poverty threshold of 50 per cent, i.e. households with an income of less

than 50 per cent of the median income. In all included countries, it is much more common for children

in single-mother households to live in poverty, compared with children in two-parent families. The

United States have the greatest share of poor children in single-mother households (49 per cent),

followed by Ireland (47 per cent) and the UK (45 per cent). Yet, what interests us more is the relative

difference in poverty between family types, i.e. the gap between poor children in single-mother

households and in two-parent families. The bars in Figure 3 illustrate the ratio between the percentage

of poor children in single-mother households in a country and the percentage of poor children in twoparent

households in the same country. The relative family type difference is largest in the Czech

Republic, where children in single-mother households about eleven times as likely to be poor,

followed by Germany, where the risk is about eight times as high. What is noteworthy is that some

countries with low child poverty rates such as Sweden and Norway have large family type differences.

Conversely, the US, for instance, has rather high poverty rates for children in both household types

and thereby not a particularly high poverty ratio.

[FIGURES 2 AND 3 ABOUT HERE]

The present study aims to assess whether the associations between family type and the child outcomes

vary systematically across countries. Despite the complexity of single motherhood and the crossnational

variations in the characteristics linked with it, it seems reasonable to expect that the family

gap in child outcomes is associated with family type differences in economic conditions, i.e., that in

countries with small family type differences in economic conditions, the family type gap in child

outcomes is smaller, and conversely, that in countries with large family type differences in economic

circumstances, the family type gap in child outcomes is larger. Consequently, the country-wise

analyses of the family type differences in child outcomes will be sorted according to the relative

difference in poverty as shown in Figure 3.

Data

The data used are derived from the international WHO study Health Behaviour in School-aged

Children (HBSC) (see www.hbsc.org) of 2001/02. Cross-sectional surveys with children of 11, 13, and

15 years of age have been conducted regularly since 1983/84 (Roberts et al 2007). The 2001/02 survey

included 36 countries and regions of Western and Eastern Europe as well as Israel, Canada, and the

US. The data are nationally representative, with a few exceptions: regional samples were selected in

Germany, and separate studies cover the Flemish- and French-speaking regions of Belgium, as well as

England, Scotland and Wales of the United Kingdom. The sampling technique was cluster sampling

through school classes. The sample is approximately 1,500 students in each of the three age groups,

i.e. a total of approximately 4,500 from each country. The number of respondents in each country

ended up being around 4,000 and the total number of respondents for all 36 countries was 162,306.3

Questionnaires were delivered to schools, administered by teachers, and returned to the national

research institute in charge when completed (Roberts et al 2004). The Luxemburg Income Study

provide data on relative poverty rates of children in single-mother and two-parent households for 23 of

the countries included in the HBSC, so these were chosen in order to get a country-level “background

variable” for all cases. Furthermore, I only included children who live with two original parents or

with their mother only. Cases with a missing value on any of the variables used were excluded (9.5 per

cent). Thus, after selections, n=92,564. For details on the number of cases in each country, see Table

A1 in the Appendix.

Method

The method used is logistic regression. The effect of family structure on each separate outcome is

analysed country-wise, with controls for gender and age. The figures shown in the tables are odds

ratios with corresponding p-values, showing the over- or underrisk for children living with a single

mother compared with children living with both original parents. To correct for the fact that

respondents are clustered in school classes, I have estimated robust standard errors using Stata’s

‘cluster’ command.4

To assess country differences in the effect of family type, I have analysed the entire sample with

dummy variables for each country, and included interactions between each country and family type. In

3 For details on number of respondents in each country, by gender and by age group, see Roberts et al 2004,

Table 2.

4 In the analyses, I used the cluster command for school class, with some exceptions: in the analyses of Belgium,

Estonia, and Greece, I clustered on school, because information on school class was not available for all cases. In

the analyses of Germany, I did not use the cluster command since for a large number of cases, information on

both school class and on school was missing.

these analyses, I have estimated robust standard errors to account for the fact that respondents are

clustered in countries.5

Variables

Family type is defined from the pupil’s responses to a question with whom he/she lives. The options

were ”Mother”, ”Father”, ”Stepmother (or father’s girlfriend)”, ”Stepfather (or mother’s boyfriend)”,

”Grandmother”, ”Grandfather”, ”I live in a foster’s home or children’s home”, ”Someone or

somewhere else: please write it down”. Those who ticked ”Mother” and ”Father” are defined as living

with two original parents, and those who ticked ”Mother” but not ”Father” or ”Stepfather” are defined

as living with the mother only. The question was posed for ”your main or your only home”, but also

for a second home if there was one. I only included information for the main home, and I only

included young persons who stated that they lived in the main household ”All the time” or ”Most of

the time”.

Social support from mother is derived from the question ”How easy is it for you to talk to the

following persons about things that really bother you?”, with response options ”Very easy”, ”Easy”,

”Difficult”, ”Very difficult” and ”Don’t have or see this person”. Those who found it ”Very easy” or

”Easy” to talk with their mother are defined as having social support from their mother.

Social support from father is derived from the same question as above. Those who found it ”Very

easy” or ”Easy” to talk with their father are defined as having social support from their father. Those

who replied “Don’t have or see this person” were excluded from the analysis.6

Health complaints is constructed from the question: ”In the last 6 months, how often have you had the

following….?” The complaints listed are: Headache, stomach-ache, back ache, feeling low, irritability

5 Even though a number of previous studies have used multilevel analysis on the HBSC data, I prefer not to do

this due to the relatively small number of countries. Grouping countries into welfare state typologies, classified

according to various criteria, is also common in cross-national research. Yet, this is not necessarily ideal for

every purpose. For instance, there may be empirical exceptions to the expected pattern as well as within-regime

type variation (see e.g. Uunk 2004). Lumping countries together according to a given typology may hide

similarities and differences between countries. Given the exploratory and descriptive objective of this paper, it

seems appropriate to start by analysing countries separately, and, as a second step, look for country clusters or

patterns in the obtained findings. More importantly, social policy and economy are not the only possible factors

which affect the living conditions of children in single-mother households. Thus, it is not evident that welfare

regime types are an ideal way of categorising the countries in this study. Furthermore, the n in each country is

large enough to analyse separately.

6 The reason for this is that in some countries, such as Italy and Austria, a substantial share of single mothers are

in fact single due to widowhood (Chapple 2009, Table 4). When excluding children whose father is deceased

(but also, due to the wording of the response option, those who are not in contact with their father), the variable

becomes more comparable across countries. Consequently, the number of cases is somewhat smaller in the

analyses of this item: n=88,184.

or bad temper, feeling nervous, difficulties in getting to sleep, feeling dizzy. For each complaint, the

respondent could choose between ”About every day”, ”More than once a week”, ”About every week”,

”About every month”, ”Rarely or never”. Those who had two or more complaints, more than weekly

were defined as having health complaints. This is the definition used in HBSC reports (Torsheim,

Välimaa and Danielson 2004).

Less than good self-reported health is derived from the question ”Would you say your health is…”,

with response options ”Excellent”, ”Good”, ”Fair”, and ”Poor”. Those who replied that they had fair

or poor health were defined as having less than good self-reported health.

Computer in household is defined from the question ”How many computers does your family own?”

and the response options are: ”None”, ”One”, ”Two”, ”More than two”. I dichotomised the variable

into those who do not have a computer and those who have at least one computer.

Not well off comes from the question ”How well off do you think your family is?”. The response

options were ”Very well off”, ”Quite well off”, ”Average”, ”Not very well off” and ”Not at all well

off”. The three latter alternatives were defined as regarding one’s family as not well off.

Gender and age at the time of the interview are included as control variables.

Results

The analytical strategy is to conduct a set of country-wise analyses for each outcome variable. Logistic

regression analyses were run with family type as the independent variable and with gender and age as

control variables. Each graph presents the results for one dependent variable: the bars show the odds

ratios and p-values are derived from robust standard errors. Raw percentages by family type are shown

below the graph. Results are presented country-wise in the same order as in Figure 3, i.e. according to

the relative difference in poverty between children in single-mother households and children living

with two original parents. It is expected that family type differences in the child outcomes are larger

the greater the family type differences in economic conditions, i.e., the further to the right on the x axis

of the graph.

To assess whether there are differences between countries, for each dependent variable I conducted

analyses of the whole sample, i.e. all cases in all countries, and included interactions between country

dummy variables and the single-mother household dummy variable. Interaction terms were included

one at the time and the US was chosen as reference category. Countries which are significantly

different from the US are marked with an inequal sign (_).

The first two graphs present findings for children’s experience of social support from their parents,

measured by an indicator of ease of communication with one’s mother and father, respectively. As to

ease to talk with one’s mother (Figure 4), in eleven countries there is a statistically significant

association with family type showing that children in single-mother households less often find it easy

to talk with their mother, compared with children living with two original parents. In the remaining 12

countries, there are no statistically significant associations, even though the direction of the association

is similar in all cases but one (Denmark). In the next graph, showing ease of communication with

father (Figure 5), children in single-mother households report less social support from the father in 17

of the 23 included countries.7

[FIGURES 4 AND 5 ABOUT HERE]

There is also consistency across countries when looking at health outcomes. Both health complaints

(Figure 6) and less than good self-reported health (Figure 7) are more common among children in

single-mother households than in intact families in a range of countries.

[FIGURES 6 AND 7 ABOUT HERE]

The association between economic resources and family type is very clear across countries. In all but

four countries, children in single-mother households are more likely than children in intact families to

report that they do not have a computer (Figure 8). What is noteworthy is the large cross-national

differences in absolute numbers (see percentages below the graph) and how this affects the odds

ratios: for instance in Russia 51 per cent of children in intact families report to have a computer, and

38 per cent of children in single-mother households. In the Netherlands, the percentages are 99-95,

which in fact renders a much lower odds ratio than the one for Russia. This figure makes clear that one

needs to look at both the absolute and the relative family type differences to get a fuller picture.

When it comes to the description of one’s household as ”not well off”, children in single-mother

households tend to report this more often than their peers living with two original parents in all

included countries (Figure 9).

[FIGURES 8 AND 9 ABOUT HERE]

7 As noted in the Variables section, respondents who replied “Don’t have or see this person” about the father

were excluded from this analysis. However, I also conducted analyses where I included these cases and coded

them as not having social support from father. With this alternative variable, family type differences were very

clear for all countries and larger than those presented in Figure 5 (p<0.001 for all countries).

To assess whether there is a pattern in the association between family structure and the included

dependent variables, a summary of the statistically significant family type differences for each

outcome, by country, is presented in Table 1. Again, countries are sorted by family type differences in

economic conditions, with the smallest family type difference in poverty at the top (Russia) and the

largest at the bottom (the Czech Republic). It was expected that the family type gap in child outcomes

would be smaller in countries with small family type differences in poverty, and larger in countries

with large family type differences in poverty. However, no such evident pattern is shown. In all

countries, there is at least one statistically significant difference by family type. Some countries, such

as Israel, Spain, and Greece, report a family type difference on only one outcome, while other

countries – Russia, Belgium, Switzerland, the USA, the UK, the Netherlands and France – report

family type differences on all the studied outcomes. There seems to be a vague pattern indicating that

countries with small shares of children in single-parent households tend to report fewer differences,

and countries with large shares of children in single-mother households report more differences.

However, nor this pattern holds fully, as for instance Netherlands with a relatively small share of

children in single-mother households reports statistically significant family type differences on all

included indicators.

[TABLE 1 ABOUT HERE]

When analysing the whole sample and testing for statistically significant differences between countries

(with the US as reference category), a number of such differences are found and marked with inequal

signs in the graphs. No clear pattern is found, although there is a weak tendency that (some) family

type differences are significantly smaller in countries where family type differences in poverty are

small, such as Russia, Poland, and Estonia, and larger in countries where family type differences in

poverty are larger, such as the Netherlands, France and Sweden. This tendency is however far too

weak to be used to drawing any conclusions about the variation in family type ‘effects’ between

countries.

Concluding discussion

The present study has analysed the relative difference between children living with a single mother

and with two original parents on a number of living conditions indicators in 23 countries, based on

reports from the children themselves. Children in single-mother households have lower social support

from parents, less good health, and smaller economic resources than children living with two original

parents. Findings are consistent for all indicators. However, the absolute family type differences are

overall relatively modest. This reflects the previously observed finding that there is a small but

consistent gap in well-being between children in divorced and in two original-parent families (Amato

and Keith 1991; Amato 2000; Pryor and Rodgers 2001; Chapple 2009).

Interestingly, the general tendency that children in single-mother households report less welfare is

rather consistent across all countries. Even though far from all differences are statistically significant,

it is striking that the bars in the graphs in nearly all cases point in the expected direction. Thus, the

associations between family structure and various living conditions are systematic in a wide range of

countries. There is no clear pattern in how the family type difference in poverty is associated with

family type differences in the child outcomes. However, the countries with the fewest statistically

significant family type differences – Israel, Spain, and Greece – all have small shares of children

living in single-mother households (see Figure 1). Furthermore, the USA and the UK, with the largest

shares of children in single-mother households, report statistically significant differences by family

type on all included items. This finding is in line with the results of Pong et al (2003) and Garib et al

(2007): that children of single-mother households in countries with small shares of single-mother

households come out better compared with their peers in countries where single-mother households

are more common. Yet, the empirical pattern is not fully consistent. For Italy, with one of the lowest

divorce rates in Europe, we find statistically significant family type differences on several outcomes.

Conversely, countries with rather small shares of children in single-mother households such as

Switzerland and the Netherlands (see Figure 1) report family type differences on all included

indicators.

Several previous studies on the HBSC data have found clear cross-national differences in the

socioeconomic gradient in children’s health. Torsheim et al (2006) come to the conclusion that

children in countries with a high dispersion of material resources have poorer self-rated health

compared with children in countries with greater equality in material resources, also when adjusted for

the individual family level of affluence. Zambon et al (2006) categorise the countries into different

welfare regimes and find that the association between socioeconomic position and health is weaker in

social democratic and conservative regime types and stronger in liberal and other regime types. How

come, then, that there are no systematic country differences in the effect of family type in the

outcomes studied here – at least not following a pattern of family type differences in poverty? One

reason may be that single-motherhood is complex and that also other characteristics associated with it,

apart from economic hardship, have consequences for children. Single mothers have on average lower

educational level and social class than two-parent families, greater time constraints, less social support

and less good psychological health – characteristics which may all contribute to the family type gap in

child well-being. And these characteristics are not necessarily systematically related to social policy

and other macro-level variables.

In fact, several cross-national studies of family structure and child outcomes do not find that family

type differences are smaller in countries with a strong welfare state. In a recent meta-analysis and

literature review of single parenthood and child outcomes in OECD countries, Chapple (2009)

concludes that the effect sizes differ between countries but that there is no systematic link to

differences in policies. The average effect size is somewhat larger for the US and for the Nordic

countries compared with other OECD countries (Chapple 2009). From their cross-national analyses of

family type and mathematics achievement, Schiller et al (2002) conclude that the negative effect of

non-traditional family types is greater in more affluent countries. Marks (2006) finds similar results in

his analyses of reading and mathematics scores: effects of family type are not weakest in countries

with generous welfare systems, but relatively strong family type effects are found in Belgium, Finland,

Norway and Sweden. Furthermore, in their overview of studies of family transitions and child

outcomes, Pryor and Rodgers (2001) conclude that there is a small but consistent gap between children

in intact and separated families in a range of outcomes, and that the effect sizes are similar across

geographical locations (though only the UK, the US and Australia are included in their overview).

Schiller et al (2002) discuss their results in terms of an increasing importance of parental time and

attention, and suggest that this may be lacking especially in non-traditional families in more

industrialised nations. They also argue that geographical mobility is higher in these countries and that

extended families are smaller, which implies lower access to social support from extended family and

community networks (Schiller et al 2002).

Despite the fact that we do not find a systematic cross-national pattern of family type differences, is it

possible to draw any conclusions from the findings with regard to social policy? It should be noted

that the analyses focus on the relative difference in living conditions between children in singlemother

households and children living with both original parents. The relative difference is to a great

extent rather similar across countries. That the absolute living conditions may, and do, vary across

countries becomes particularly evident in our measure of economic resources, whether or not there is a

computer in the household. Also here, a gap between family types is depicted in most countries, but

just as noteworthy is that the absolute shares of children with a computer in their household differ

substantially, in particular between Eastern and Western Europe. A parallel can be drawn to Bradbury

and Jäntti’s (1999) comparison of absolute and relative child poverty levels: they show that in Russia, child poverty is around 25 per cent when defined by a relative poverty line, but that 98 per cent of children are poor when the US official poverty line is used (i.e. a measure of absolute poverty) (Bradbury and Jäntti 1999, Table 3.3). Hence, just because there is no systematic cross-national pattern in the family type difference in children’s living conditions, one cannot draw the conclusion that social policy or macroeconomic conditions make no contribution at all. A high absolute average level of living conditions, with a small family type gap, is obviously to prefer than a low average level of living conditions, also with a small family type gap.

Finally, although the family type differences in children’s living conditions are relatively small, they should not be neglected by policy-makers. From a general equality as well as from a child rights perspective, all children should have equal chances. One of the main principles of the UN Convention on the Rights of the Child (CRC) is non-discrimination, i.e. that all children have equal rights and thus that these rights shall be equally accessible for all children (Article 2 of the CRC). The systematic and cross-nationally consistent findings that children in single-mother households fare less well compared with children living with two original parents hence counter the non-discrimination claim of Article 2 of the CRC. What is noteworthy is that this goes also for countries such as the Nordic ones, which have explicitly “child-friendly” policies.

Family Structure Changes and Children’s Health, Behavior, and Educational Outcomes

2011-06-07T23:40:00.000-07:00

1 Introduction

More and more children do not grow up in traditional nuclear families. Instead, they grow up in single-parent households or in families with a step-parent. Hence, it is important to improve our understanding of the impact of "shocks" in family structure due to parental relationship dissolution on children. To give an example of the severity of this problem, we can look at a western country like Denmark. In 1980, almost 83% of all Danish children in the ages 0 to 17 lived with both of their biological parents, but this number steadily decreased to 73% in 2005 as shown in Figure 1. That is, more than 25% of all Danish children did not live with both of their biological parents in 2005. In many other western countries the divorced population is increasing as well, and for these countries we expect to see a similar picture as that depicted in Figure 1 for Denmark.

The topic of family structure changes has been studied extensively in the psychology and sociology literature and recently also in the economics literature. International studies mainly suggest a negative relationship between non-nuclear family structures and child outcomes. There are two potential explanations for this. First, families that split up may possess characteristics that are di¤erent (and worse) than what is seen in nuclear families, i.e. non-nuclear families are a selected group of families.1 Another explanation is that there may be negative causal e¤ects of separation.2 To analyze the e¤ect of a change in family structure on child outcomes properly, detailed data are required. Many international studies only have access to datasets where family structure is observed once during a child.s life and this "value" is then extrapolated to the entire childhood.3 This approach leads to imprecise results and is known as the "window problem" (Wolfe et al. (1996)). Danish register data makes it possible to avoid the window problem and to make quantitative economic analyses which can shed light on the selection and causation explanations. However, until now, register data from Denmark has not been exploited systematically to analyse family structure and its relation to child outcomes.

Thus, this study is an empirical study of the relation between family structure changes and several di¤erent child outcomes using an administrative register dataset consisting of the entire population of Danish children born from January to May 1985. These children are followed on a yearly basis until 2005, i.e. until they are 20 years old. The study is partially descriptive, but it also attempts to identify causal e¤ects of family structure changes. A di¤erences-in-di¤erences analysis is used for the latter purpose for the health outcomes.

Though descriptive in its nature, this study contains multiple contribu-

tions to the Danish and international literature. One contribution is to use

detailed data for the quantitative economic analyses to avoid the window

problem and obtain more precisely identi.ed results. Some other studies

are also using panel data, but they do not have a population sample of chil-

dren available for their analyses. Using a population sample implies a lot of

observations and thus more precision in the empirical analysis. Another con-

tribution is to illuminate new aspects of family structure by investigating a

variety of short- and long-term child outcomes. In particular, I focus on the

relation between family structure and child outcomes such as health (hos-

pitalizations), behavior (crime), and education (high school enrolment and

high school GPA). A third contribution is to propose new questions to this

area such as the importance of the timing of family dissolution, the short-

and long-term impact of multiple family structure changes, and the time

spent in a single-parent household. This study can be used as a reference

point for further studies.

To preview the results, in the descriptive analysis I .nd a negative rela-

tion between all three types of child outcomes and family structure changes.

The results further indicate that the number of family structure changes

might be more important for child outcomes than the years spent in a single-

parent household. The di¤erences-in-di¤erences analysis continues to show

a negative e¤ect on health from a family structure change, thus indicating a

negative causal e¤ect on children of experiencing a family structure change.

The remainder of the paper proceeds as follows. In Section 2, a literature

review is presented and in Section 3 the estimation strategy is explained.

Section 4 gives an introduction to the datasets used in the analysis, and

the empirical results are presented in Section 5 along with a sensitivity

analysis. Finally, the results are summarized and a brief discussion of policy

implications is presented in Section 6.

2 Literature Review

Human capital (HC) theory provides an explanation of family structure.s

(causal) e¤ect on child outcomes (Becker (1991) and Becker (1993)). The

intuition is that households are time- and money-constrained. With fewer

adults in the household there is less time and money to invest in children and

this may reduce child outcomes through decreased parent-child interaction

and fewer goods. Becker.s theories of the family suggest that two adults can

pool their resources and specialize in tasks within and outside the household,

thereby increasing total market and household production.

However, HC theory cannot explain why children growing up in non-

nuclear families seem to have lower outcomes even with a step-parent present

(Gennetian (2005)). Possible explanations include asymmetric informa-

tion and monitoring (Weiss and Willis (1985)) or di¤erent bargaining power

among biological and step-parents (Lundberg and Pollak (1996)). The latter

explanations imply that biological mothers may act to ensure that a child liv-

ing with a stepfather is no worse o¤ than biological children of both adults in

the household. If the stepfather allocates less resources to a stepchild, but

the mother thereafter redistributes resources to make all children equally

well o¤, in total these children will be worse o¤ than children living in a

nuclear family. This holds despite there being two adults in the household.

Recent theoretical studies on child development and child outcomes sug-

gest that skills beget skills and capabilities today foster future capabilities

through self-productivity and dynamic complementarity. That is, higher

stocks of skills in one period create higher stocks of skills in the next period,

and stocks of skills acquired in one period make investments in future pe-

riods more productive (Cunha and Heckman (2007) and Heckman (2008)).

Thus, early investments in children are often most fruitful and even more so

if they are followed up by later investments, but this depends on the exact

technology of skill formation. Cunha and Heckman (2007) demonstrate that

some periods may be more e¤ective in producing certain skills (sensitive pe-

riods), or it may be the case that only one period is e¤ective in producing

a certain skill (critical period). With this theory in mind, the timing of

family structure changes might be extremely important for child outcomes.

An early family structure change might a¤ect children more negatively than

if they experience a family structure change in their teens. I return to this

question in the hypotheses and the empirical analysis, in particular by look-

ing directly at family structure changes in di¤erent age brackets. Finally,

the sociological and psychological literatures have also developed theories

of the relation between family structure and children.s outcomes. One ex-

ample is loss of parental control theories which focus on trauma or shocks

experienced during childhood, e.g. changing family structure. I will refer to

this theory as the stress theory. Ginther and Pollak (2004) summarize the

latter and related theories in more detail.

In general, the empirical literature cannot agree on whether selection

or causation (or maybe both) explain how children.s outcomes are related

to family structure. The selection theory claims that families that split

up may possess characteristics that are di¤erent (and worse) than what is

seen in nuclear families, i.e. non-nuclear families are a selected group of

families. Children from non-nuclear families will thus have worse outcomes

than children from nuclear families both before and after a family structure

change. Studies pointing in this direction are Björklund and Sundström

(2006), Björklund et al. (2007), and Piketty (2003). On the other hand,

Ermisch and Francesconi (2001) and Steele et al. (2009) argue among others

for causal e¤ects of separation, i.e. negative e¤ects on child outcomes caused

by a change in family structure. The potential causal e¤ect of separation

may be visible even before the separation formally occurs, especially if the

parents argue or .ght a lot before the actual separation (Amato (2000)).

Furthermore, a practical issue pointed out in Manski et al. (1992) em-

phasizes that assumptions about the actual process generating family struc-

ture and child outcomes are important for the estimation strategy. Strong

assumptions lead to more precisely estimated e¤ects, but at the cost of

maybe being less realistic. Thus, we need to consider this when choosing

the estimation strategy.

The main problem in empirical analyses of family structure is speci.-

cation of a counterfactual to divorce and separations (Ginther and Pollak

(2004)). "No divorce" is not a reasonable counterfactual since parents can-

not be forced to live together. This problem of identifying the e¤ect of

divorce or separation can be solved by using an instrument, e.g. a quasi-

experiment such as a reform in the rules of divorce. It has been done in the

studies by Gruber (2004) and Francesconi et al. (2005), but unfortunately

it is di¢ cult to .nd suitable reforms and thus good instruments. Further-

more, using divorce laws as an exogenous cause of divorce does not solve the

problem that changes in divorce regimes may directly a¤ect the nature of

intrafamily bargaining, and this may lead to di¤erent implications for the

children (Stevenson and Wolfers (2006)). Alternatively, one can consider the

death of one parent as an exogenous shock to family structure as in Corak

(2001) and Francesconi et al. (2005). The problem with using this .exoge-

nous.shock is, that it might not be completely exogenous if, for example,

the parent.s death is the result of several years of sickness. Then it might

be perfectly expected and the e¤ect on the child of losing a parent might be

reduced due to .preparation.for the event. If the parent.s death is due to

e.g. a tra¢ c accident, the loss is truly an exogenous and unexpected event.

However, loosing a parent permanently due to parental death may not be

comparable to loosing a parent due to divorce. In the latter case the parent

is still alive and may still participate actively in the child.s daily life.

In this current study (as in most of the existing studies in the litera-

ture), the problem of not having a reasonable counterfactual is ignored and

the results of the empirical analysis in Section 5 thus have to be interpreted

with some caution. For reasons of comparison, I will carry out simple cross-

section estimation strategies, but also do a di¤erences-in-di¤erences (D-i-D)

analysis as done by Sanz-de Galdeano and Vuri (2007) to get closer to the

causal e¤ects. The D-i-D analysis makes it possible to control for common-

alities during the time periods for both children that do experience family

structure changes and those that do not. Moreover, .xed characteristics of

the same child is also di¤erenced away in this panel D-i-D. Thus, this leads

to results far beyond the simple cross-section analysis and thereby gives

some indication of the causal e¤ects of family structure changes.

3 Empirical Model

In the empirical analysis I attempt to test the stress theory, i.e. whether

children are traumatized in the short and the long run by shocks in fam-

ily structure during childhood. I investigate how the timing and sum of

family structure shocks a¤ect children to get information on negative im-

pacts (stress) accompanying a change in family structure. Furthermore, by

comparing the importance of years in a single-parent household to the im-

portance of the number of family structure changes experienced during the

entire childhood, I give some evidence on the validity of Beckers human cap-

ital theory. The main hypotheses analysed in the empirical study are the

following:

Hypothesis 1 A family structure change is a (negative) shock for the child

and thus has negative implications for the child.s outcomes both in

the short and long run.

Hypothesis 2 The number (and timing) of family structure changes a¤ect

child outcomes by stressing the child.

Estimates are based on a linear version of the education production func-

tion model (Todd and Wolpin (2003)). Assume that achievement for child

i, Ti, can be expressed as a linear function of the explanatory variables

Tija = _0 + _1Dija+_2Fij (a) + j + _ija; (1)

where Tija is the outcome for child i from family j at age a (education,

behavior (criminal activity), or health); Dija is a family structure change

(measured as year-to-year changes); F is a vector with child and family

background information (child.s birthweight and gender, ethnicity, number

of siblings at age 10, age at .rst separation, parental work experience, annual

wage income, and level of education); j is a family .xed e¤ect, e.g. norms

and values in the family; and, .nally, _ija is an error term.

Child outcomes depend on the child.s unobservable endowed ability which

I proxy by birthweight. This is a common way of dealing with unobserved

endowed ability in the literature (e.g. Behrman et al. (1994)). To get

consistent estimates, I include information on both contemporaneous and

historical inputs. Relevant background information such as labor market

information for the "social" parents (the parents living in the household) is

included in the analysis, as well as information on the child, siblings, child.s

age at the time of the .rst shock, etc.

As outcome measures I use education, measured as high school enrolment

and high school grade point average. Further, behavioral measures such as criminal activity are used, and .nally health outcomes (hospitalized in a

year, number of sickdays) are investigated. By studying this wide variety

of outcomes, di¤erent aspects of the e¤ect of family structure changes are

illuminated.

I also apply a di¤erences-in-di¤erences model for the health outcomes in

which I assume there are only two time periods, 1 and 2. Family structure

changes are now expressed as

D(i; t) =

1 if divorce or separation in period t

0 if no divorce or separation in period t

Changing the notation slightly, we get the following expression for child

achievement as a function of family structure changes:

T (i; t) = _0 (t) + _1D(i; t)+ (i) + v (i; t) ; (2)

where _0 (t) is a time-speci.c component; _1 is the impact of parental di-

vorce or separation; v (i; t) is a serially uncorrelated transitory component;

and (i) is an individual-speci.c component representing unobserved pre-

disruption characteristics, e.g. stress associated with an unhappy family life.

D(i; t) is probably correlated with (i), i.e. we cannot identify the e¤ect of

parental divorce or family dissolution from OLS estimates of this equation.

Di¤erencing Equation (2) with respect to t and adding covariates implies:

T (i; 2) 􀀀 T (i; 1) = _0 + _1D(i; 2)+_2X (i) + (v (i; 2) 􀀀 v (i; 1)) ; (3)

where _0 = _0 (2) 􀀀 _0 (1), _2 = _2 (2) 􀀀 _2 (1) and X (i) is a vector of

observed characteristics assumed uncorrelated with v (i; t).

The model in Equation (3) is identi.ed if the conditional restriction is

ful.lled

P (D(i; 2) = 1 j X (i) ; v (i; t)) = P (D(i; 2) = 1 j X (i))

for t = 1; 2;

i.e. there is identi.cation if changes in the outcome variable over time would

have been exactly the same in both treatment and control groups in the

absence of divorce or separation (parallel trend assumption) at least when

including the covariates.

The D-i-D approach follows Sanz-de Galdeano and Vuri (2007). The

control group consists of those children that do not experience any family

structure change during childhood, and the treatment group is the children

that have experienced a family structure change from period 1 to period 2.

The dataset has to be set up in a certain way for this to work. First, in period

1, none of the children must have experienced a family structure change, i.e.

children who do not have (typically) their father in the household during the

.rst time period of their life have to be deleted from the sample. Children

most often stay with their mother in the year after a family structure change,

but few, especially older children, choose to stay with their father the year

after a family structure change. Figure 2 shows the distribution for Denmark

in 2008. For families that break up, about 14.5% of the children live with

the father the following year, whereas more than 80% of the children live

with their mother. Some of the oldest children choose not to live with either

parent the following year.

Secondly, the D-i-D strategy is only possible for the health outcomes us-

ing my datasource since D-i-D requires outcome measures that are observed

over time along with the changes in family structure. The educational out-

comes are not observed until the child is quite old, and thus the e¤ect of

Figure 2: Family type for children who lived with both mother and father the

previous year, but not in the current year, Denmark 2008. Source: Statistics

Denmark

early family structure changes cannot be analysed on the educational out-

comes. Observations on criminal activity start when the child is 15 years old

and thus also quite late. Health outcomes are available from the child is 6

years old and therefore health outcomes are preferred for the D-i-D analysis.

Due to the richness of the data, having yearly observations, it is possible to

de.ne the two time periods in the D-i-D analysis in di¤erent ways. I use

ages 0 to 11 as period 1 and ages 12 to 20 as period 2 in the main speci.ca-

tion. To test the robustness of the results and the importance of the timing

of family structure changes, I also repeat the analysis using ages 0 to 7 as

period 1 and 8 to 20 as period 2.

When empirically analyzing the e¤ect of family structure on child out-

comes, there are some challenges related to .nding the best speci.cation of

the empirical model. The problem arises because family structure is clearly

not random and thus simple OLS or probit parameter estimates may be

biased. OLS is nevertheless used in most of the existing literature and will

therefore also be used here for comparison. In addition to OLS and probit

estimation of Equation 1, I use D-i-D analysis to di¤erence away unobserved

individual-speci.c components. These might be components that can be re-

lated directly to family dissolution decisions and thus would bias a simple

cross-section analysis. In other words, the D-i-D analysis allows for the

possibility that parental divorce or separation is correlated with unobserved

family characteristics that may in.uence child outcomes. A family or sib-

lings .xed e¤ects analysis is employed in several studies in the literature,

but the advantage of D-i-D compared to a family .xed e¤ects analysis is

that also children without siblings are analyzed in the D-i-D framework.

Analyzing only families with siblings and only in cases where the siblings

have experienced di¤erent family structures reduces the external validity of

the study as pointed out by Mo¢ tt (2005). Thus, a D-i-D analysis is easier

to generalize to larger groups of the population.

4 Data

Estimations are based on an administrative register dataset consisting of

the entire population of Danish children born from January to May 1985.4

About 50,000 children were born per year in Denmark in the mid-1980.s.

Due to seasonality in childbirths, I have about 36,000 unique children in

the dataset. The data source contains yearly, rich information on family

background along with other information on children, parents, and siblings.

This rich datasource implies more reliable estimates of the e¤ect of family

structure changes on child outcomes than what is seen in most existing

studies. Family structure changes are observed until the child reaches the

age of 18. After this point in time, the child is registered as an adult whether

or not the child still lives with his/her parents.

In the Danish tax and income registers created by Statistics Denmark,

the children and their parents are followed on a yearly basis from the year

of birth to 2005 if they have not left the country or died. Parental .gures

are identi.ed each year based on the household characteristics of that year.

If the child lives with both biological parents, parental information concerns

the biological parents. If, on the other hand, the child lives with one biolog-

ical parent and one step-parent, then the parental information collected is

for these individuals. I will distinguish between biological and non-biological

parents by referring to either .biological.or .social parents..

At this point it is useful to de.ne precisely what is meant by a family

structure change. Throughout this study, family structure changes will be

de.ned as follows unless explictly noted:

De.nition 1 A family structure change is any change in the compo-

sition of biological or social parents that the child experiences. It does not

have to be a legally binding change, such as divorce or remarriage.

Thus, any observed change in the number or identity of the (biological or

social) parents in the household from one year to the next is considered a

family structure change. This can be a reduction in the number of .parents.,

but also an increase in the number of .parents..

The registers provide information on the parents. and children.s mar-

ital status, residence, education, income, labor market activities, health,

criminal activity, etc. This includes information on the number of hospital-

izations and days in hospital, whether the individual has been convicted of a

crime, the children.s highest completed level of education, whether they are

currently enrolled in education, and their grade point average if they have

completed high school. While most Danes complete high school at age 20,

some Danish children take an optional 10th grade before enrolling in high

school and they may therefore still be enrolled in high school at the age of

20. It is also very popular among the Danish youth to take a sabbatical year

between high school and college. Thus, higher educational goals may not be

clear in 2005 when the children are 20 years old. When investigating educa-

tional outcomes, I therefore focus on high school enrolment and high school

completion, which in short will be referred to as high school enrolment.

Furthermore, information on criminal activity and health is available

in the data, which enables analyses of both short and long-term outcomes,

cognitive and non-cognitive. I measure children.s health in two ways; an

indicator for being hospitalized during the year in question and the number

of days in the hospital. In some cases I aggregate this information into an

indicator for ever being hospitalized as this provides less noisy information.

Health information is available from 1991 on. Thus, health outcomes can be

investigated on a yearly basis beginning when the child is 6 years old. For

criminal activity I use a dummy for being convicted in the year in question,

but in some cases I also aggregate the information into an indicator for ever

being convicted. The age of criminal responsibility is 15 in Denmark so I

cannot investigate this behavioral outcome before the child is 15 years old.

Educational, health, and behavioral outcomes are important outcomes both

for the children themselves and more generally for society. These variables

are therefore the focus of this study.

In Denmark, GPAs from high school range from 0 to 13 with 8 as the middle grade. 6 is equivalent to passing the exam. However, by looking at the number of observations it is clear that we only have a measure of GPA for those who actually have completed high school by age 20. Thus, it seems reasonable to use high

school enrolment as the main outcome variable for education instead of the

GPA from high school. It is also clear from the table that only 8% of the

children are hospitalized during 2005, whereas 47% of the children have ever

been hospitalized. When hospitalized, they spend 3.6 days in hospital on

average. The most common reason for being hospitalized is problems during

a pregnancy (for girls). 7% of the children are convicted of a crime during

2005, but 19% have ever been convicted. However, the gender distribution

is very uneven as only about 2% of girls get a conviction in 2005.

After a separation, most of the children live with their mother. The

children that live with their father are usually in the group of 11- to 16- year-

olds, i.e. the oldest children. On average, children growing up in non-nuclear

families experience their .rst family structure change at age 6. Finally,

fathers have a higher annual wage income than mothers, and fathers.work

experience is also higher. Most of the children have one sibling by the age

of 10 and the gender composition of the sample is almost 50/50.

In Table 2 the distribution of family structure changes for the sample

is shown. The highest number of (observed) family structure changes for

a child is 135, but it is not common to experience such a high number of

family structure changes. Even when collecting 8 and more family structure

changes into one single category, only 1% of the sample is present here. The

most common number of family structure changes is 0, i.e. the child grows

Table 2: Total family structure changes during childhood by 2005.

up in a nuclear family. 27% of the children have experienced 1 or 2 family

structure changes during childhood, implying that they have experienced

the parents separate and potentially also experienced a step-parent moving

into the household afterwards. In some cases, the children with 2 family

structure changes experience one parent moving out of the household and

then the same parent returns some years later. A descriptive analysis (avail-

able from the author upon request) shows that these families are di¤erent

than .normal.nuclear families. Thus, even though it is the same biological

parent who is moving back into the household, this is counted as 2 family

structure changes.6

Focusing on descriptive evidence, Figures 3 to 5 show three di¤erent

outcomes plotted against the child.s age and separately by the .nal number

of family structure changes the child experiences during childhood. The

pattern from the .gures is clear: the more family structure changes, the

worse the children.s outcomes. This can of course be a selection e¤ect as the

families with most family structure changes might be di¤erent than a typical

nuclear family. This will be investigated later. But .rst it can be seen in

Figure 3: No high school enrolment by the age of the child and the .nal

number of family structure changes during childhood. 1985-cohort.

Figure 4: Convictions by the age of the child and the .nal number of family

structure changes during childhood. 1985-cohort.

Figure 5: Hospitalizations by the age of the child and the .nal number of

family structure changes during childhood. 1985-cohort.

Figure 3 that more than 60% of children with three or more family structure

changes do not enrol in high school compared to about 40% of the children

in nuclear families. For convictions the picture is just as clear as shown in

Figure 4. At the age of 20, more than 12% of the children experiencing three

or more family structure changes have been convicted whereas less than 6%

of the children in nuclear families have been convicted. When splitting the

sample by gender, the numbers are even more disturbing. More than 20% of

the boys with three or more family structure changes have been convicted in

the year they turn 20 whereas less than 10% of the boys in nuclear families

have been convicted. A similar di¤erence is observed for girls, but at much

lower levels. For hospitalizations the picture is slightly more blurred. There

is not much di¤erence in the probability of hospitalization between those

experiencing one or two family structure changes, but there is still quite a

big di¤erence between children from nuclear families and children from the

families with most family structure changes. This also holds for boys and

girls separately, but with girls having more hospitalizations at the higher

ages and vice versa for the youngest ages. The .gures for boys and girls

separately are available from the author upon request.7

5 Estimation Results

In the empirical analysis I .rst look at cross-section estimations as is com-

mon in the literature. When investigating hospitalizations, convictions, and

no high school enrolment using a probit estimation procedure, I generally

.nd highly signi.cantly estimated marginal e¤ects as shown in Table 3. All

the estimates point to a negative relation between single-parent households

(or family structure changes) and children.s outcomes. Furthermore, using

OLS, the relation between GPA and years in a single-parent household and

the total number of family structure changes during childhood is analyzed.

Again I .nd highly signi.cantly estimated coe¢ cients as shown in Table 3

and all estimates point to a negative relation between single-parent house-

holds (or family structure changes) and children.s GPA after high school.

These results con.rm what is generally found in other empirical studies

using data from both the US and Sweden, for example Björklund et al.

(2007). Relatively few children live in single-father households compared to

single-mother households and the marginal e¤ects for this group are thus

less precisely estimated.

Interpreting the coe¢ cients in Table 3 as causal e¤ects, the most interesting observation is the relatively larger e¤ect family structure changes have

as compared to the e¤ect of years in a single-parent household. For all out-

comes, the .e¤ect.of a(nother) family structure change is about four times

higher than the .e¤ect.of living a(nother) year in a single-parent household

during the childhood. As expected, the .e¤ect. of experiencing a family

structure change (0/1) is higher than the .e¤ect. of the number of family

structure changes, as the .rst variable captures the .total e¤ect.of experi-

encing family structure changes. Thus, these results suggest that instability

of households has a greater impact on children.s outcomes than potentially

a lower household income and fewer adults in the household to take care of

the children. In other words, this .nding provides some evidence against

the mechanism suggested by Becker (Becker (1991) and Becker (1993)).

It is also interesting to note that the e¤ect of family structure changes on

children.s outcomes is linear for all outcomes, but high school GPA. When

including dummies for 1 family structure change, 2 changes, 3 changes, and 4

or more changes and comparing to children with no changes in the analyses,

it turns out that the marginal e¤ects are linear as shown in Figure 6. Includ-

Figure 6: Marginal e¤ects for ever hospitalized, ever convicted, no high

school enrolment, and high school GPA by the number of family structure

changes.

ing squared or cubic terms of the total number of family structure changes

in the di¤erent speci.cation also reveals the same picture; the coe¢ cients

for the squared and cubic terms are insigni.cant, whereas the coe¢ cient for

the linear term is highly statistically signi.cant. The marginal e¤ects for

high school GPA is not as clearly linear as for the other outcomes, but this

may be explained by the fact that the children with a high school GPA is

a smaller and positively selected sample. Thus, they are not representative

for the full sample and I will therefore not rely heavily on the potential

non-linear pattern shown by this outcome measure. Instead, I will continue

to include a variable for the total number of family structure changes in the

estimations as this measure seems to be linear for the main sample.

Now focusing only on years in a single-mother household as in most

of the literature and again investigating cognitive and non-cognitive child

outcomes in 2005 by OLS and probit, we see in Table 4 that for the di¤erent

Table 4: OLS and probit estimation results for the e¤ect of years in a single-

mother household at di¤erent ages on various child outcomes in 2005.

outcome variables the .e¤ects.di¤er by the age of the child. This con.rms

what has also been seen previously in the literature. The age groups are

de.ned based on the type of care the child receives besides parental care.

For example, children in the ages 0-2 are eligible for daycare, children in the

ages 3-5 usually attend preschool, and children aged 6 and older are in school.

For hospitalizations and convictions, living in a single-mother household in

the .rst three age groups seems to be important. The marginal e¤ects are

generally higher for the younger age groups. For GPA, the period 3-5 years

of age seems to be most important, whereas for high school enrolment all

age groups are important, but the marginal e¤ects are again higher for the

younger age groups. Thus, living in a single-mother household in the earliest

childhood seems to have the most detrimental e¤ect on most outcomes.

Families that split up may be a selected group. Thus, many studies

include covariates in their cross-sectional estimations to control for this se-

lection. If the signi.cance of the coe¢ cient for family structure change dis-

appears after including the covariates, this is interpreted as evidence for se-

lection being the explanation of negative e¤ects of family structure changes.

That is, if the coe¢ cient turns insigni.cant, there is no evidence of a causal

e¤ect of family structure changes on children.s outcomes. I have also in-

cluded covariates8 in all the estimations shown in Tables 3 and 4, but the

signi.cance of the coe¢ cient for family structure change does not disappear.

However, the size of the coe¢ cient decreases somewhat.9 Since there is no

evidence for selection being the main explanation for the negative relation

between child outcomes and family structure changes, I decided to show the

most raw numbers in the tables, i.e. without the covariates included.

Another way of investigating the selection and causation explanations

is to carry out a di¤erences-in-di¤erences (D-i-D) analysis of the health

outcomes. The parameter of interest is the coe¢ cient to the interaction

between treatment (experiencing parental divorce or separation) and the

age dummy (aged 12 or older). The comparison age group consists of those

between the ages of 6 and 11.10 Any statistically signi.cant coe¢ cients on

this interaction term are interpreted as a causal e¤ect of family structure

changes. The D-i-D analysis is possible solely for the health outcomes, since

this is the type of outcome for which there is information for most years

(from the child is aged 6 and on). The two outcomes that are depicted in

Table 5 are .rst a variable counting the total number of years a child has

been hospitalized in each period and, secondly, the total number of days

spent in hospital by period. Results are shown both with and without a list

of covariates that includes the child.s birthweight, gender, ethnicity, number

of siblings at age 10, parental work experience, parental annual wage income,

and parents.level of education.

The parameter of interest shows a statistically signi.cant positive e¤ect

Table 5: Di¤erences-in-di¤erences estimates of the e¤ect of experiencing a

Family structure change at age 12 or older on children’s health outcomes.

On child outcomes. This can be interpreted as the causal effect of experiencing a family structure change after age 12. Thus, children who experience

that their parents divorce or separate when they are 12 years old (or older)

have worse health outcomes (more often hospitalized, more days in hospi-

tal) than children that do not experience a family structure change during

their childhood. The results do not change remarkably after including a

wide array of covariates which again indicates that selection is not the main

explanation for the e¤ect of family structure changes on children.s health.

That is, children.s lower outcomes in families that incur family structure

changes are not caused by remarkably worse characteristics of these families

compared to nuclear families. This is somewhat surprising and does not

support the .ndings in many of the studies in the literature. However, it

does support the .nding in Francesconi et al. (2005).

Table 6: Di¤erences-in-di¤erences estimates of the e¤ect of experiencing a

family structure change at age 8 or older on children.s health outcomes.

For testing the robustness of the results I have carried out a series of

sensitivity tests. First, I have excluded all children whose parents are immi-

grants, i.e. the children were born in Denmark in the relevant time period

to biological parents not born in Denmark. The reason for excluding these

children from the sample is that these families may have a di¤erent culture

in particular with respect to whether or not it is acceptable to divorce or

separate. However, excluding the children born by immigrants does not

change the main results.11

As shown in Table 6, I have also tried to change the de.nition of the age

brackets for the time periods. The descriptive evidence in Table 4 suggested,

that the timing of family structure changes might be important for most of

the outcomes, in particular for the health outcomes. Thus, I have now

de.ned time period 1 as children aged 6 to 7 and time period 2 as children

aged 8 to 20. By changing the time periods, I only exclude individuals that

experience parental divorce or separation before they turn 8. Moreover, I

can assess the e¤ect of much earlier family structure changes, i.e. when the

children have just started in school. By comparing Tables 5 and 6, we see

that the parameter of interest is still signi.cant and positive. The size of

the coe¢ cient is even bigger in Table 6 than in Table 5 implying that earlier

family structure changes might be more detrimental to child outcomes than

later family structure changes. It might also be the case that what we see in

the table is that the result of moving the age bracket 4 years down makes it

possible to experience even more family structure changes during childhood.

This might be causing the .stronger.e¤ect on child outcomes.

6 Conclusion

More and more children do not grow up in traditional nuclear families. In-

stead they grow up in single-parent households or in families with a step-

parent. Hence, this paper aims at improving our understanding of the im-

pact of "shocks" in family structure due to parental relationship dissolution

on children. Today, in some countries, more than 40% of all children do not

grow up in a household with both of their biological parents.

Using a Danish administrative register dataset, I .nd in the empirical

analysis that children who have experienced family structure changes during

childhood seem to have worse educational outcomes and a higher propensity

of being hospitalized and convicted of a crime. The children in the dataset

experience up to 13 family structure changes during childhood. More family

structure changes imply worse outcomes and might actually be more impor-

tant than the number of years a child has spent in a single-parent household.

Also the age at which the family structure change occurs seems to be im-

portant as separations at younger ages are more negatively related to the

outcomes. Finally, the D-i-D analyses show that there seems to be a causal

e¤ect of family structure changes on health outcomes. A causal e¤ect on

child outcomes is also found by Francesconi et al. (2005) using German data.

However, the results are in contrast to studies using Swedish and U.S. data,

which .nd only a selection e¤ect of family structure changes (for example

Björklund and Sundström (2006) and Björklund et al. (2007)).

Particularly for studies using Danish or Scandinavian data, this study

is thought to be a useful reference study for further research on the topic

of family structure changes. Many of the analyses in other studies in the

literature are replicated and extended in this study and I use both short-

and long-run outcomes as well as cognitive and non-cognitive outcomes.

This study, like most of the studies in the literature, ignores the problem of

having a reasonable counterfactual to separation or divorce. Thus, in future

research I expect to look more into families that are already separated to

avoid the problem of the .missing.counterfactual.

The lack of a clear counterfactual makes it di¢ cult to give policy recommendations. The study clearly shows that family structure changes cause negative e¤ects on children.s outcomes, but despite this .nding it is not necessarily a better option for the child that the parents do not separate. However, for ethical reasons, it is not possible to investigate this question further as we cannot force families to either split up or remain together.

Family Structure, Institutions, and Economic Growth

2011-06-07T23:33:00.000-07:00

There is a vast amount of literature that considers the importance of the family as an institution. Little attention, however, has been given to the impact of family structures on institutions, their dynamics, and the ability to change them. This limits our ability to understand distinct institutional developments B and hence growth B in the past and the present. This paper illustrates this by highlighting the importance of the European family structure in one of the most

fundamental institutional changes in history.

This change has been the emergence of economic and political corporations in late medieval Europe.1 >Corporations= are defined here, consistent with their historical meaning, as intentionally created, voluntary, interest-based, and self-governed permanent associations. Guilds, fraternities, universities, communes and city-states are some of the corporations that dominated Europe in the past. Businesses and professional associations, business corporations, universities, consumer groups, republics and democracies are some of the corporations in modern economies.

Providing institutions through corporations is a novelty. Historically, the institutions that secured one=s life and property and mitigated problems of cooperation and conflicts were initially provided by large kinship groups. Subsequently, such institutions that rely on this family structure were complemented or replaced by those provided by self-interested rulers. Corporation-based institutions can substitute for those provided by both kinship groups and self-interested rulers.

When they substitute kinship-based institutions, corporations are complementary to a different

family structure, namely, the nuclear family structure. For an individual, corporations reduce the

1 This paper draws on Greif (2006, chapters 8, 9, and 12). benefits from belonging to a kinship group while a nuclear family increases the benefits from being a member of a corporation.

In tracing the origins of the European corporations this paper focuses on their complementarity with the nuclear family structure. It presents why large kinship groups (i.e., clans, lineages and tribes) declined in medieval Europe (section 1) and why the resulting nuclear family structure, along with such factors as individualism and the interest of the Church, were fundamental to the rise of corporations (section 2). European economic growth in the late medieval period was based on an unprecedented institutional complex of corporations and nuclear families (section 3). Subsequent European history suggests that this institutional complex was conducive to long-run growth and growth-enhancing institutions. Yet we know surprisingly little regarding the reasons for this or why it is so difficult to transplant this complex to other societies.

The paper concludes by reflecting on these issues (section 4).

I. The Evolution of Family Structures in Europe

The conquest of the Western Roman Empire by Germanic tribes during the medieval period probably strengthen the importance of kinship groups in Europe. Yet, the actions of the Church caused the nuclear family B constituting of husband and wife, children, and sometimes a handful of close relatives B to dominate Europe by the late medieval period.

The medieval church instituted marriage laws and practices that undermined large kinship groups.2 From as early as the fourth century, it discouraged practices that enlarged the family, such as adoption, polygamy, concubinage, divorce, and remarriage. It severely prohibited marriages among individuals of the same blood (consanguineous marriages), which had constituted a means to create and maintain kinship groups throughout history. The church also curtailed parents=abilities to retain kinship ties through arranged marriages by prohibiting unions in which the bride didn=t explicitly agree to the union.

2 This dogma was self-serving by increasing, for example, the likelihood that an individual=s bequest would be given to the church. Mitterauer and Sieder (1982); Goody 1983; Ekelund et. al. (1996). European family structures did not evolve monotonically toward the nuclear family nor was their evolution geographically and socially uniform (Herlihy 1969; Greif 2006, chapter 8).

However, by the late medieval period the nuclear family was dominate. Even among the Germanic tribes, by the eighth century the term family denoted one=s immediate family, and shortly afterwards tribes were no longer institutionally relevant. (Guichard and Cuvillier, 1996.)

Thirteenth-century English court rolls reflect that even cousins were as likely to be in the presence of non-kin as with each other. (Razi 1993).

The practices the church advocated, such as monogamy, are still the norm in Europe.

Consanguineous marriages in contemporary Europe account for less than one percent of the total number of marriages. In contrast, the percentage of such marriages in Muslim, Middle Eastern

countries, where we also have particularly good data, is much higher - between twenty to fifty percent. (Bittles 1994.) Among the anthropologically defined 356 contemporary societies of Euro-Asia and Africa, there is a large and significant negative correlation between Christianization (for at least 500 years) and the absence of clans and lineages; the level of commercialization, class stratification, and state formation are insignificant. (Korotayev 2003.)

II. Incorporation as a Function of Family Structure

The decline of large kinship groups in Europe transpired during a period in which the state was also disintegrating there. At the same time, individualist cultural beliefs hindered the rise of social networks for diffusion of information and lowered expectations for collective, informal

enforcement. (Greif 1994.) To solve problems of coordination and enforcement, a new solution was required and the medieval economic and political corporations constituted this solution.

These corporations were intentionally created, voluntary, interest-based, and self-governed permanent associations. Participation was voluntary in the sense that one had to be

induced to be a member and to achieve that corporations had to cater to the interests of their

members. Corporations were self-governed in the sense that their members specified the rules that regulated their activities and they often did so without being established by a state. Power was shared and the corporation=s leaders were held accountable for their actions. Corporations were permanent in the sense that whether one was established for an indefinite or a finite length of time, its perpetuation didn=t depend on the participation of any particular member. The corporations had often acted as having a legal personality, their property, rights, and obligations of corporations were distinct from those of the members. The lack of effective states, however, implied that they often could not be sue and be sued in a state court.

Although the weakness of the state and the nuclear family structure were important to the

rise of corporations, this rise was not determined by these factors. A multiplicity of outcomes

could have prevailed given the rules of the game spanned by these factors. Where power was

asymmetrically distributed, as was the case in Eastern Europe, local rulers established themselves.

Where power was more equally distributed, anarchy and the strengthening of large kinship family

structures sometime were the result. More often, in such cases, the European political culture and

legal heritage made corporations a focal point. Political organization among equal and free

individuals governed by their laws under leaders who were first among equals is central to the

Greek, Roman, and Germanic political traditions. Similarly, the Roman, Germanic, and Christian

legal heritage provided precedents to corporate bodies (Kuhn 1912; Berman 1985).

The actions of the late medieval Church were rendered corporations focal point. During the

eleventh century, the Church incorporated itself to deflect threats to its independence and property

from secular rulers and its own agents. To this end, the papacy formulated a corporation law that

drew on Roman, Germanic and Christian legal principles. It rejected the view that royal

permission was required to establish a corporation, asserted that any corporation had taxation

rights, legislative and judicial jurisdiction over its members, and placed the corporation=s agents

under the authority of the corporation=s members. (Berman 1983). This codification bestowed

legitimacy on self-organized and self-governed corporations in general, created a related moral

code, and provided a legal code for coordinating expectations.

III. Europe Inc. and Late-Medieval Growth

By the late medieval period, economic and political corporations dominated Europe. We

still lack a micro-analysis of their internal governance but we know how they fostered growth.

They promoted growth by securing private property rights (from the grabbing hand of the state,

pirates and each other), providing public goods, supporting markets, fostering innovation and

training, and mobilizing armies to advance their economic interests. Indeed, from circa 1050 to

1348, Europe experienced its longest, post-Roman period of economic growth.

Corporations, such as monasteries, fraternities and mutual-insurance guilds, provided

social safety nets against famine, unemployment and disability. The majority of the population

belonged to such corporations, at least in England (Richardson 2005). By providing social safety

nets, corporations enabled individuals to better align their economic decisions with their

preferences. Individuals could make such economic decisions as those about marriage,

childbearing, investment, migration, and risks without interference by members of a larger kinship group.

Craft guilds regulated production, training, and the protection of brand names. Guilds,

universities and monastic orders developed and distributed scientific and technological knowledge

(Mokyr 1990; Epstein 1998). Corporations, particularly merchant guilds and communes, provided

the institutional foundations of markets. They protected property rights at home and abroad,

secured brand names, and provided contract enforcement in exchange. Corporations such as

city-states, the German Hansa and military orders mustered armies to advance their economy.

(Greif, Milgrom, and Weingast 1994; Greif 2004; Greif, 2006, chapters 8, 10; Gonzalez 2004;

Richardson 2004a,b.)

Many late medieval corporations, as is well recognized regarding the Italian city-states

were political in the sense that they had independent legal, administrative, and military capabilities.

Yet, most European cities west of the Baltic Sea in the north and the Adriatic Sea in the south were

also political corporations known as communes. In England alone there were about 500 hundred

of them (Beresford and Finberg 1973). Political corporations also prevailed among west European

peasants (Reynolds, 1984, ch. 4-5; Brenner 1987).

Because these corporations often preceded the state in Europe, they provided these nascent

states with indispensable services such as tax collection, administration of the law, and mustering

armies. Self-interested rulers were therefore constrained from adopting policies hindering these

corporations= economic interest. (Greif 2005.) These corporations= importance is reflected in the

European rulers= responses to the increasing burden of financing warfare. By the thirteenth century,

all European principalities had representative bodies to approve taxation and in all of them

communes were represented (Ertman 1997). Their military and economic strength and their

administrative importance probably rendered their concession to taxation indispensable.

IV Corporations, Nuclear Families and Development

To the modern period, corporations and nuclear families constituted a distinguishing

feature of the particularly European institutional foundations of markets, polities and knowledge.3

Europe=s economic ascendancy began after nuclear families and corporations began to dominate

its society, economy and polity and since then growth prevailed particularly when and where

corporation-based institutions reigned. Why?

Relative to the kinship family structure, the nuclear family has both growth-enhancing and

growth-retarding properties. For example, while the nuclear family enhances flexibility in

economic decision-making, kinship groups can better provide social safety nets than the nuclear

family. European corporations, however, enabled non-kin to benefit from safety nets while

retaining the nuclear family structure. In the contemporary west, the state B a corporation B fulfills this role.

We know, as discussed above, how the European corporations fostered growth, but we

know surprisingly little about why the interests of the corporations were aligned with growth. After all, corporations aimed at benefiting their members rather than pursuing growth. History is rich with examples of even initially growth-enhancing economic corporations that became means for rent seeking, redistribution, or control by the state. Some political corporations degenerated into oligarchies and conflicts within others facilitated the rise of despotism. Contemporary developing countries provide many examples of growth-inhibiting corporations.

Yet, comparing historical and contemporary experiences suggests several reasons why

European corporations overall effect was to foster growth. In Europe, corporations preceded the

rise of relatively effective rulers who had B due to the nature of world economy B to draw on

domestic resources while facing inter-state military competition. Rulers’ dependence on domestic corporations sometimes constrained their arbitrary powers. At the same time, rulers often constrained growth-inhibiting inter-corporation military competition forcing corporations to compete economically. The overall beneficial impact of European corporations may well be due to the luck of having locations, such as the Dutch Republic and England, where these conditions persisted. (Greif 2005.)

In contrast, developmental policy by the end of colonization emphasized first establishing

3 Greif (1994, 2006); Kuran (2005); Jabar and Dawod (2003); Hamilton (1991) an effective state with an administrative and military monopoly. International capital markets and international corporations made rulers less dependent on the cooperation and resources of their citizens. Domestic corporations were therefore more likely to become instruments for redistribution and patronage.

The lack of a complementary nuclear family structure seems to have been a source for the limited success of fostering growth by transplanting corporations, including the western modern state, in developing countries. Indeed, in the past and present, large kinship groups resisted the expansion of a corporative state that threatened to undermine them and their power. (Greif 2006,

chapter 8.)

Last and not least, centuries of corporations and nuclear families in Europe seem to have created cultural heritage which is absent from developing countries. Corporations and nuclear families embody and reinforce a culture of self-governance, cooperation among non-kin, the legitimacy of majority rule, respect for minority rights, the accountability of leaders, and individualism. The finding in cross-country regressions that history matters to the stability of the democratic B corporative B form of government and growth (Persson and Tabellini 2005), may well reflect the cultural heritage of the late medieval corporations.

Reference

Beresford, M. and Finberg, H.P.R. 1973. English Medieval Boroughs: A Handlist. Newton Abbott: David and Charles.

Berman, Harold J. 1983. Law and Revolution. The Formation of the Western Legal Tradition.

Cambridge: Harvard University Press.

Bittles, Alan H. 1994. AThe Role and Significance of Consanguinity as a Demographic Variable.@ Population and Development Review 20 (3): 561-84.

Brenner, Robert. Agrarian Class Structure and Economic Development in Pre-Industrial Europe.

In T.H. Aston and C.H.E. Philpin (eds), The Brenner Debate. Agrarian Class Structure and

Economic Development in Pre-Industrial Europe. Cambridge: Cambridge University Press. Pp.

10-63.

Ekelund, Robert B., Jr., Robert F. Hébert, Robert D. Tollison, Gary M. Anderson, and Audrey B.

Davidson. 1996. Sacred Trust: The Medieval Church as an Economic Firm. New York: Oxford

University Press.

Mitterauer, Michael, and Reinhard Sieder. 1982. The European Family. Oxford: Basil Blackwell.

Epstein, Steven. R. 1998. ACraft Guilds, Apprenticeship and Technological Change in

Preindustrial Europe.@ Journal of Economic History 53 (4): 684-713.

Ertman, Thomas. 1997. Birth of the Leviathan. Cambridge: Cambridge University Press.

Gonzalez de Lara, Yadira. 2004. AThe State as an Enforcer in Early Venetian Trade: A Historical Institutional Analysis.@ Memo, University of Alicante.

Goody, J. 1983. The Development of the Family and Marriage in Europe. Cambridge: Cambridge University Press.

Greif, Avner. 1994. "Cultural Beliefs and the Organization of Society: A Historical and

Theoretical Reflection on Collectivist and Individualist Societies." The Journal of Political

Economy, vol. 102, No. 5 (October): 912-50.

Greif, Avner. 2004. Impersonal Exchange without Impartial Law: The Community Responsibility System. Chicago Journal of International Law, Vol. 5(1): 109-138.

Greif, Avner. 2005. Commitment, Coercion, and Markets: The Nature and Dynamics of

Institutions Supporting Exchange.@ In the Handbook for New Institutional Economics. Edited by Claude Menard and Mary M. Shirley. Norwell MA: Kluwer Academic Publishers. Chapter 28.

Greif, Avner. 2006. Institutions and the Path to Economic Modernity: Lessons from Medieval

Trade. Cambridge: Cambridge University Press.

Greif, Avner, Paul Milgrom and Barry Weingast. 1994. "Coordination, Commitment and

Enforcement: The Case of the Merchant Gild." The Journal of Political Economy, vol. 102, No.

4 (August): 745-776.

Guichard, Pierre and Cuvillier, Jean-Pierre. 1996. Barbarian Europe in Ed. André Burguiére et. al.

A History of the Family, Vol. I. Cambridge: Polity Press. Pp. 318-378.

Hamilton, Gary G. 1991. The Organizational Foundations of Western and Chinese Commerce: A

Historical and Comparative Analysis. In Gary G. Hamilton (ed.), Business Networks and

Economic Development in East and Southeast Asia, 48-65. Hong Kong: University of Hong Kong, Centre of Asian Studies.

Herlihy, David. 1969. Family Solidarity in Medieval Italian History. In David Herlihy, Robert S.

Lopez, and Vsevolod Slessarev (eds). Economy, Society and Government in Medieval Italy. Kent, Ohio: Kent University Press.

Jabar, Faleh A. and Hosham Dawod. 2003. Tribes and Power. Nationalism and Ethnicity in the

Middle East. London: Saqi.

Korotayev, A.V., 2003. Unilineal Descent Organization and Deep Christianization: a

Cross-cultural Comparison. Cross-cultural Research 37 (1): 133-157.

Kuhn, Arthur K. 1912. Comparative Study of the Law of Corporations. Columbia in the Social

Sciences, no. 123. New York: AMS Press.

Kuran, Timur. 2005. 2005. Why the Islamic Middle East Did Not Generate an Indigenous

Corporate Law. Memo, University of Southern California.

Mokyr, Joel. 1990. The Lever of Riches. Oxford: Oxford University Press.

Persson, Torsten and Guido Tabellini. 2005. Democratic Capital: the Nexus of Political and

Economic Change. Memo.

Razi Z.. 1993. The Myth of the Immutable English Family. Past & Present 140(Aug), Pp. 3-44

Reynolds, Susan. 1984. Kingdoms and Communities in Western Europe, 900-1300. Oxfrod:

Clarendon Press.

Richardson, Gary. 2004b. Brand Names before the Industrial Revolution. Working paper. UC

Irvine.

Richardson, Gary. 2005. The Prudent Village: Risk Pooling Institutions in Medieval English

Agriculture. Journal of Economic History, 65(2) (June): 386-413

Richardson, Gary. 2004a. Guilds, Laws, and Markets for Manufactured Merchandise in

Late-medieval England. Explorations in Economic History 41(1): 1-25

Welcome to Family Lab International

2011-06-03T13:48:00.000-07:00

Thanks for your interest! We envision a world of Family. We are proud to get you here. Please comeback soon
This blog is in developing process, meanwhile you may visit other affiliated blogs:

<Family Lab Bangladesh> www.familylab.blogspot.com

<Family Lab International> http://familylabinternational.blogspot.com/

<Family Based Capitalization (FBC)> http://fbcbd.blogspot.com/

<SMEs International> http://smecentre.blogspot.com/

<Familinomics International> http://familinomicsinternational.blogspot.com/

<Entrepreneurs Resource Centre> http://homeofentrepreneurs.blogspot.com/

<Home Enterprise Resource Centre> http://hecentre.blogspot.com/

<Family Enterprise Resource Centre> http://fecentre.blogspot.com/

<Mizanism: Extended Horizon for Open Thoughts> http://mizanism.blogspot.com/

<Inclusive Finance-Finance for ALL> http://familifinance.blogspot.com/