PRINCIPLES AND PROSPECTS OF
THE LIFE COURSE PARADIGM
In recent years, the scientific interest of
many demographers has shifted from
studying “demographic regimes” and
large-scale processes to analysing longitudinal
micro data in the form of “life
courses”. By some of its advocates, the
“life course approach” is heralded as a
new paradigm capable of reinvigorating
the study of populations. Since, ultimately,
demography deals with the fates
and choices of individuals, the micro
level is optimally suited to study chains of
causation. Thus, according to the
supporters of the life course, demography
no longer has to make do with the
second-best option of aggregate data on
the meso or macro level. Moreover, no
longer is demography limited to studying
snapshots of births, marriages and deaths.
Instead, it will be able to analyse
processes of family formation and wellbeing
that span life times and can even be
followed across generations. New strategies
of data collecting and data sharing, as
well as new statistical techniques have
opened up vistas of research that will
reshape demography’s landscape.
Has this vision appeal for historical
demographers as well? Are life courses—
basically standardized biographies—really
different from the traditional tool of
family reconstitution? Are the new techniques
capable of changing scholarly practices
in historical demography? To what
extent is the life course adopted in historical
demography, what results have been
reaped and what kind of obstacles can be
perceived? In this paper, I will discuss the
life course approach in terms of its theoretical
claims, its methodological practice,
and the kind of results it tends to
produce. In the next section, I will introduce
the approach and discuss basic
concepts and principles. Then, I will
discuss whether its application in the field
of historical demography indeed can be
seen as a paradigmatic shift. A short
survey of recent literature will serve to
illustrate the current practice of life course
demography. Does this research live up to
the expectations of life course theorists? If
not, does this result from problems with
the theory itself, from the current state of
statistical methods, or from the inadequacy
of historical data? In the remainder
of the paper I will discuss the potential
pitfalls and try to come up with solutions
and recommendations. If the life course
approach is indeed potentially capable of
both rejuvenating historical demography
and linking this field to ongoing scientific
debates, what needs to be done?
203
ANNALES DE DÉMOGRAPHIE HISTORIQUE 2007 n° 1 p. 203 à 230
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE
PARADIGM
by JAN KOK*
* The author wishes to thank the participants of the workshop “Historical demography: Past Accomplishments,
Present debates, and Future Developments” (Nijmegen, June 14th2006) as well as Kees
Mandemakers and Lex Heerma van Voss for comments on an earlier version of this paper.
09-kok.qxp 16/07/2007 16:24 Page 203
WHAT IS THE LIFE COURSE
APPROACH?
In the social sciences, the systematic
study of life courses has been inspired by
the life history techniques that were developed
by W. I. Thomas and F. Znaniecki,
who studied the lives of immigrant
Polish peasants in the United States
(Thomas and Znaniecki, 1918-1920).
In the Chicago-school of sociology these
techniques were popular, especially in
the study of deviant behaviour. In the
1950's and 1960's, when the social
sciences fell under the spell of
functionalism, the analysis of biographies
lost its attraction. From the
1970's, however, dissatisfaction rose
with main-stream sociology's a-historical
study of societal structures on the
basis of aggregated data. Renewed interest
in historical developments and in the
complex interaction between individual
lives and social processes led to a revival
of the life history approach. However,
this did not entail a rejection of quantitative
analysis. On the contrary, new
statistical techniques have been developed
which enable life patterns of large
numbers of people to be analysed.
The life course approach is essentially
a heuristic device to study the interaction
between individual lives and social
change. It is a way of conceptualizing
lives within the contexts of families,
society and historical time. The life
course can be defined as the sequence of
positions of a particular person in the
course of time. A position can be either
marital status, parenthood, employment,
residence in a particular location
etcetera. A life course analysis studies
the frequencies and the timing of
changes in positions, generally of groups
such as birth cohorts. These changes are
called events or transitions. Every life
course is characterized by a sequence
and combination of transitions, such as
leaving home, finding work, finding a
partner and becoming a parent. Such
sequences of roles or statuses are named
trajectories, whereas the time between
transitions is known as duration. To a
large extent people follow in their life
normative patterns regarding the proper
behaviour at a given age and the proper
sequence of transitions (e.g. education
followed by starting employment
followed by marriage followed by
parenthood). Such “standard” patterns
are known as social pathways or cultural
scripts (Elder et al., 2003; Liefbroer and
Dykstra, 2000). These—often genderspecific—pathways
are not only shaped
by ethical prescriptions and cultural
preferences but have also been institutionalised
by the regulations of the
modern welfare state (education, oldage
pensions etc.) (Mayer, 1991; NaveHerz,
1991). However, they change
continually as the shifts in aggregate life
course patterns of subsequent birth
cohorts testify (Liefbroer and Dykstra,
2000). Finally, turning points denote a
substantial change in the direction of
one’s life, and can be determined either
subjectively or objectively (Hareven,
2000, 153, 329).
According to its advocates, the life
course “provides a framework for studying
phenomena at the nexus of social
pathways, developmental trajectories,
and social change” (Elder et al., 2003,
10). Thus, its most ambitious claim is
that it can unearth the interactions and
intersections between the micro level of
individuals and the macro levels of
culture, economy, social policy etcetera.
How is this claim substantiated? To
understand how life course analysis
204
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 204
operates, five principles are essential
(Elder et al., 2003, 1) Life-Span development.
The principle of life-span Development
claims that we can only understand
choices and behaviour by taking
into account experiences in earlier stages
of life. The life course is a cumulative
process and should therefore be studied
as a whole. 2) Agency. Life course theory
acknowledges that people determine—
within given constraints and opportunities—their
own life course. In other
words, we cannot fully understand
behaviour without taking account of the
scope, the effectiveness and the direction
of individual life plans. Thus, in
considering agency, we need to distinguish
between long-range life plans
(Hareven, 1982), short-term tactical
reconsiderations and the kind of decision
making geared at sheer survival. 3)
Time and place. This principle simply
states that individuals and birth cohorts
are influenced strongly by historical
context and specific location. Due to
historical change, every birth cohort has
a unique set of constraints and opportunities
that shapes the courses of its lives.
Life course analysis is devoted to studying
the interaction between life courses
and demographic, economic, institutional
and cultural changes. 4) Timing.
The impact of individual experiences
and historical events on the subsequent
life course depends to a large extent on
their timing. That is, at what specific
age does a particular transition (or
combination of transitions) occur?
Much attention goes to the interaction
of different forms of time: the time of
the individual (age), of the family (stage
in the family cycle) and historical time
(economic cycles, social changes). 5)
Linked lives. This principle emphasizes
that life courses are interdependent with
others. This is particularly salient in the
context of families. Transitions of
certain family members such as leaving
home have immediate ramifications for
the others. To some extent, the temporal
associations between transitions of
family members result from, and thus
reveal, coordination within families or
households. In the words of Modell and
Hareven: “The examination of timing
can, at least, provide insights into internal
processes of family decision making”.
(Modell and Hareven, 1978; see
also Hareven, 2000).
It is clear from these principles that
the life course approach is highly holistic
in its aspiration to grasp the behaviour
of individuals within their personal
networks and within their specific place,
historical time and society. The strong
emphasis laid on time, path dependency
and contextuality qualifies the life
course approach as an essentially historical
method. Thus, in theory, it should
be highly relevant for family history and
historical demography.
A NEW PARADIGM FOR
HISTORICAL DEMOGRAPHY?
According to various scholars, the life
course approach is not merely an addition
to the historical demographer’s toolbox,
it signifies nothing less than a
change of paradigm. In principle, it is
capable of superseding the dominant
paradigm of “pure state” demography
(Courgeau, 2000; Courgeau and
Lelièvre, 1999; Willekens, 1999).
Demographers commonly study only
one phenomenon at the time in “homogeneous
populations”, that is populations
whose relevant characteristics are
supposed to remain constant in the
period of observation. For instance, one
205
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 205
studies the frequency of marriage of rural
men in a particular region in a particular
period. The assumptions underlying this
approach are often taken for granted.
Thus, people who are removed from the
population, e.g. through migration, are
supposed to have the same behaviour
towards marriage as people still remaining
or entering through migration. Also,
alternative (competing) events, in this case
cohabitation, are not taken into account.
Complex processes such as partner selection
and family formation are often
simplified in order to study them in their
“pure state”. Finally, to assure homogeneity,
researchers often revert to creating
small subpopulations (e.g. sons of
wealthy farmers) that do not allow for
significance tests. In reality, people
constantly move from one “subpopulation”
to another, they shift from urban to
rural places, they change occupational
positions, wealth and status groups and
even alter their religious denomination. It
is unrealistic to assume that the behaviour
of individuals in a new group immediately
converges to the group average
and that their previous experiences are no
longer relevant. The life course approach
is devised precisely to deal with the ongoing
impact of past experiences (principle
of Life-Span Development) in constantly
changing contexts (Principle of Time and
Place). Moreover, event history analysis,
the statistical technique most appropriate
to the life course method, can deal with
both the problem of selection by loss
from observation and the reality of
heterogeneity in any population,
however small (see below).
Another problem that traditionally
besets historical demography is the risk
of “ecological fallacy”. This means that
inferences about behavioral propensities
of individuals are based solely upon
aggregate statistics collected for the
group to which those individuals belong
(Robinson, 1950; Courgeau and
Baccaïni, 1998). For instance, hypotheses
with respect to the effect of religion
on fertility are often based on associations
of (marital) fertility and religious
composition of the population measured
at the municipal or regional level
(Haan, 2005). In theory, however, the
correlation of e.g. catholicism with high
fertilty may be explained by a third, as
yet unknown, factor.
To some extent, historical demographers
have already made the “micro”
turn long ago, by creating family reconstitutions
at the level of one or several
villages. However, these reconstitutions
were often based solely on the records of
baptism, marriages and burials, thus
having little information on unmarried
persons and childless couples. In addition,
the prevailing techniques did not
allow for integrating the experiences of
people who either entered or left observation
through migration. In general,
neither destinies nor personal networks
beyond the confines of the village were
taken into account (for some exceptions
see Reay, 1996; Noordam, 1986). The
life course approach aspires to overcome
all these limitations.
Several statistical techniques are particularly
suited for the life course approach.
These techniques put time—individual
age as well as chronological time—into
the centre of the analysis. Of special
importance is event history analysis in
which the dependent variable is the
length of time that elapses until a transition
from one state to the next occurs,
such as non-married to married or
employed to unemployed. The condition
for this method is that the exact date of
the transitions are known. The rates of
206
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 206
occurrence of these transitions can be
calculated as hazard rates in life tables that
incorporate the information on persons
who leave the observation (e.g. through
death or migration). “Proportional
hazard” models combine the strengths of
life tables and regression analysis. The
parameter estimates of the models assess
the risks of a particular transition, and the
ratio of two risks is called the relative risk.
The statistical significance of individual
coefficients can be determined by t tests,
whereas tests for the overall significance of
a particular model can be made using
likelihood ratio statistics (Ritschard and
Oris, 2005; Allison, 1984; Yamaguchi,
1991; Blossfeld and Rohwer, 1995). In
the past decades, these models have been
applied frequently in historical demography,
in particular in the context of life
course research (Alter, 1988; Alter, 1998).
To sum up, the life course approach
and its methodological translation in
event history analysis promises a shift
from the study of simplified single
events to the study of processes. To paraphrase
Hobcraft (2006): historical
demography can now move from the
study of births to the process of parenthood,
from marriages and divorces to
partnership, from deaths to the process
of well-being, from migrations to the
life-long shifts of positions and, finally,
from household structures to networks
of kin and friends. To what extent does
current research in historical demography
live up to this promise?
EXPLORATIONS OF LIFE COURSE
HISTORY
Within the scope of this paper, only a
brief and highly incomplete survey can
be made of the life course approach’s
inroads in historical demography.
Not surprisingly, life course theory
and (statistical) methodology have been
applied most fruitfully in the field of
family history that seems to have undergone
a true revitalization. Based on the
principle of linked lives, short and long
term effects of the family situation on
the life course have been studied.
Following the principle of life span
development, much attention has been
given to the impact of working and residential
experiences of adolescents for
their choices in later life: choices regarding
partner selection, geographical and
social mobility.
Traditional demographic history was
focussed –through the concept of the
“family cycle”–on the study of “normal”
family formation and “normal” reproduction
(Hareven, 1978). It tended to
ignore or relegate to the margins the
experiences of unwed mothers and their
children, of foundlings and orphans, of
cohabiting couples and of permanent
celibates. In the “new” family history,
these issues are studied afresh. With the
aid of event history analysis, researchers
tackle, for instance, the question of the
surmortality of illegitimate children.
Why was single motherhood so detrimental
for the survival of their children?
Has this to do with their social isolation
(e.g. as recent immigrants), their work
and incapacity to breastfeed their children,
or did the stigma of illegitimacy
induce infanticide (Van Poppel, Kok and
Kruse, 1997)? What do the residential
histories of the children (e.g. their being
taken care of by the grandparents) reveal
about illegitimacy as such (Blaikie, 1993;
Brändström, 1996)? Even less is known
about the backgrounds (legitimate or illegitimate)
and fate of foundlings (e.g.
Kertzer, Koball and White, 1997; Kertzer
and Sigle, 1998; Brunet, 2002; Levene,
207
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 207
2003). In the early nineteenth century,
child abandonment was very common in
several European countries but its implications
still have to be translated in terms
of the hypotheses on Western European
marriage patterns and reproduction.
Permanent celibacy, although well integrated
in Hajnalian theory, has hardly
been studied with respect to specific
family backgrounds, the relationship
with delayed marriage (Engelen and Kok,
2003), and residential and career choices
of the celibates themselves (Dorsman and
Stavenuiter, 1993; Capron and Oris,
2000 ; Alter and Capron, 2004).
The topic of the circulation of working
adolescents among households has
also been relatively understudied in
traditional historical demography,
although this circulation was highly
relevant for the functioning of nuclear
and stem family systems and was intrinsic
to the delayed marriage found in
large parts of western Europe (Hajnal,
1982). This lack of interest can be
explained from the fact that, apart from
the typical snapshots of households
found in censuses, mobile adolescents
are not visible in records of baptisms,
marriage and burials. In addition, their
circular mobility made no impression
on regional net migrations flows.
However, there are many pertinent
questions about this facet of European
history. How did the mobility of individual
youths relate to both their
parental household composition and
that of their employers? Were they
expelled or retained at home during
economically bad periods? Is the background
of the parents (e.g. their migration
history) relevant for the destinations
of their children? Is the experience
of servant girls in urban households
translated in their later geographical
mobility and partner choice? Recently, a
whole new literature on the subject of
leaving home and servant mobility has
sprung up, often directly working with
notions from the life course paradigm
(e.g. Van Poppel, Oris and Lee, 2004;
Fauve-Chamoux, 2004; Bras, 2004;
Dribe and Lundh, 2005).
Family dynamics form the core of
demographic history inspired by the life
course paradigm. By linking current
household composition to mortality in
different stages of life, patterns of internal
redistribution of food submerge, as well
as differential protection in times of
economic stress, the power relations
between generations and sexes et cetera.
This type of research is particularly fruitful
when families systems can be
compared. What does differential
mortality tell us about gender equality,
generational hierarchy and protection in
nuclear, stem and complex households.
How vulnerable are members of different
family systems in times of severe
economic stress? In the study Life Under
Pressure (Bengtsson, Campbell and Lee,
2004) these issues are tackled with event
history techniques that enable analysis of
the impact of fluctuations in food prices.
The study has revealed that gender and
age discrimination affected mortality in
the West as much as in the East of the
Euasian continent. In addition, local
variation proved to be more important
for mortality than expected, which
suggests that local traditions and systems
of care were of great importance.
A related topic concerns the fate of
widows and orphaned children. The
“nuclear hardship” hypothesis suggests
that in nuclear family systems the
“victims” were thrown back upon collective
care arrangements, such as charity
(Laslett, 1988). Did widows and
208
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 208
orphans fare better in non-nuclear
systems, such as the stem families of
southern Europe and Japan, or the
multiple-family households in China? A
recent volume edited by Derosas and
Oris (2002) addresses this question. For
Italy, Breschi and Manfredini (2002)
show that widows were, indeed, better
off when they lived in a complex household.
On the other hand, the ties
between parents and children as well as
relations with more distant kin were
stronger and more effective than Laslett
has supposed. In eastern Belgium, the
effects of the father’s death on the
mother’s survival were reduced greatly
when adolescent children were present
in the household. In the agrarian village
of Sart, the presence of sons was of
crucial importance, but in the industrial
town of Tilleur it was the daughters who
stayed to help their mother (Alter et al.,
2002). Also, kin living in the community
significantly decreased the widow’s
risk of dying. As in Italy, complex
households in China offered protection
to widows. The more adult men were
living in the household, the better her
chances of survival. However, the value
put on male offspring implied that a
widow would be treated much better if
she had a son. On the other hand, sons
without a mother had a rough time
when their aunts were living in the same
household. Apparently, aunts managed
to divert family resources to their own
children. This finding puts the patriarchal
nature of the Chinese family into
perspective. As Campbell and Lee put it:
“Fathers could not care for their sons, in
the face of competing mothers.”
(Campbell and Lee, 2002, 322).
Presently, the impact of the composition
of the parental household on the
subsequent life course is being explored
in fields as diverse as marriage (e.g.
Neven, 2005), social mobility (e.g. Van
Poppel, De Jong and Liefbroer, 1998;
Van Bavel, 2005) and mortality in later
life (e.g. Bengtsson and Lindstrom,
2000). Also, a whole new research fields
is emerging devoted to studying sibling
differentials in life destinies (e.g.
Barrera-Gonzalez, 1992; Adams and
Kasakoff, 1992; Oris et al., 2007; Bras
and Neven, 2007). Under what conditions
do destinies differ strongly
between one child and the other?
Clearly, inheritance practices related to
different family types are particularly
relevant for this kind of questions (e.g.
Fauve-Chamoux and Arrizabalaga,
2005; Augustins, 2002). How did
marriage, reproduction, occupational
career and residential trajectories vary
between heirs and non-heirs (Florey and
Guest, 1988; Bonnain, 1996)? On the
other hand, what kind of assistance did
siblings offer one another during the life
course, eg. by cohabiting (Perrier, 2000;
Oris, Ritschard and Ryczkowska, 2005)?
There is still much we need to know
about how families in the past actually
functioned. However exciting the new
research is turning out, the emphasis is
still very strong on household composition,
whereas much less is known on the
quality and character of relations
between family members and on family
resources in terms of economic and
social capital (networks) (Uhlenberg
and Mueller, 2003, 127). There is much
to be gained from a closer cooperation
with qualitative family history, a point
we will return to shortly.
The life course approach is clearly able
to enrich family history with a host of
new or previously disregarded research
topics. But is it also able to direct historical
demography towards issues that
209
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 209
rank high in current debates of social
scientists and social policy makers? I
refer here to concerns with ageing,
migration and integration, and demographic
transition.
Ageing
In social gerontology, the life course
approach is very much en vogue. Special
courses, journals and even research institutes
are devoted to study the process of
ageing with life course concepts. Clearly,
these concepts are helpful to reveal how
the current situation of the elderly is
related to past choices, to changes in
their social networks and in the
surrounding society (e.g. Settersten,
2002). Historical demographers can add
a long-term perspective to gerontology
by studying the trajectories, living
arrangements and health of old people
in past times. What was the role of children,
siblings and other kin in caring for
the elderly in the past? How did the
elderly preserve their independence and
at what costs? Did they benefit from
their savings in the past? How do life
course patterns of the aged vary between
regions with different family systems
and with collective care arrangements?
Surprisingly, historical demographers have
not flocked to this field, although interest is
growing (Bulder, 1993; Bourdelais, 1999;
Neven, 2002; Thane, 2003; Van Poppel
and Liefbroer, 2005).
Migration
Of all demographic variables, migration
is the “Cinderella”, the least known
in both population studies and historical
demography (Oris, 2003, 189). The
focus of migration studies has been on
long-distance and international migration
flows. The traditional approach was
to explain these flows from a comparison
of (expected) employment and income in
the sending and receiving areas. This
“push and pull” type of explanation did
not account for the intense circulation
between rural regions or the temporary
moves to cities. Moreover, migration
theorists have increasingly become aware
that individual acts of migration engender
new moves. This principle of “cumulative
causation” is becoming one of the
dominant approaches in migration studies
(Massey et al., 1993). People who
have migrated once are more likely than
others to do so in the future. Also, people
who are related to persons with migration
experience, may receive information or
practical help that lowers their own costs
of moving. These “chains” also operates
at the level of communities: villages in
which many people have left are likely to
lose even more. Thus, migration decisions
of individuals are related to their
past experience, the experience of others,
and the changes in their local context.
Clearly, the life course principles of Life
Span Development, Linked Lives, and
Time and Place are highly applicable to
this type of migration studies. We still
know very little about the moves people
made during their lives and how these
moves were related to other life course
transitions. Who were the typical
“repeaters” and what explains their
behaviour? Gradually, a new historical
research field emerges that analyses individual
migration trajectories including
intralocal moves (e.g. Pooley and Turnbull,
1998; Bras, 2002; Kok, Mandemakers
and Wals, 2005). Apart from insights in
migration decisions and in the actual
trajectories, we also need to know more
about the nature of (successful) integration
in receiving communities, preferably
in the second and third generation as
210
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 210
well. Presently, the barriers to this type of
research in historical demography are
being overcome, by creating databases of
immigrants and their descendants, by
linking across censuses, by using automated
lists of steamboat passengers,
etcetera. Examples are studies by Gjerde
(1985) who traced a group of Norwegians
both before and after their migration
to America, and studies by the
research team of Lucassen which traced
German and Italian migrants to the
Netherlands and their descendants
(Lesger, Lucassen and Schrover, 2002;
Lucassen, 2004, 2005). An inspiring
example for this kind of research has been
the study of (three) generations of internal
migrants to Turin by Gribaudi
(1987).
Demographic transition.
Finally, how is the life course approach
performing in what is often perceived as
the “core business” of historical demography:
the study of the demographic transition?
In recent decades, the classical
unified theory of the fertility decline as an
integral part of economic modernization
has been shattered. More or less to its own
surprise, the Princeton European Fertility
Project concluded that economic change
did not necessarily imply demographic
adaptation in the form of smaller families
(Coale and Watkins, 1986). Conversely,
birth control occasionally spread rapidly
in areas where economic transformation
was delayed or absent. Instead, patterns of
diffusion of innovative behaviour seemed
more appropriate to understand how and
why fertility declined. However, by and
large, the idea of diffusion has remained a
“description in search of theory” (McNicoll,
1992, 406). If religion, language or
regional dialects are to function as barriers
or mediators of innovation, how does this
work? Recently, new studies have begun
to question the outcomes of the Princeton
Project by reasserting the impact of
economic changes, such as employment
opportunities for women, on the rate of the
fertility decline (e.g. Brown and Guinnane,
2001, 2003). Even more problematic for
transition theory is the recent attack on the
notion of “natural fertility”. As conventional
wisdom will have it, in “natural
fertility” populations interference with
childbearing within marriage was “outside
the calculus of conscious choice” (Coale,
1973). However, recent studies have
demonstrated the existence of conscious
spacing in historical populations, which
even appears to have been parity specific
(Anderton and Bean, 1985; David and
Mroz, 1987; Van Bavel, 2003, 2004; Van
Bavel and Kok, 2005; Bengtsson and
Dribe, 2006). Thus, the advance of stopping
may simply have been a change of
strategy in a long tradition of birth
control. In the light of these conflicting
and confusing findings that have undermined
transition theory, Szoltysek (2007)
suggests that historical demography is
currently in a state of “epistemological
crisis”. He paraphrases Szreter (1996) that
instead of one, unilinear process we shall
have to look for “many geographically
and chronologically dispa-rate processes
occurring in distinct contexts and for
different reasons”. Unappealing as this
may be in terms of Grand Theory, such
an approach will be worthwhile, in particular
when combined with life course
concepts and methods. For instance,
using life course data from the Belgium
city of Leuven (XIXth century) Van Bavel
has analysed the advent of stopping
behaviour in ethnically mixed neighbourhoods,
thus testing the diffusion hypothesis
(Van Bavel, 2004b). Likewise, Matthys
211
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 211
and Vanhaute (2006) traced the fertility
history of rural girls who had lived, as
servants, with urban families, and might
have picked up notions of birth control.
Life course data also makes it possible to
study intergenerational mechanisms in
fertility control, an as yet largely unexplored
area (Kok and van Bavel, 2005).
For instance, by analysing careers of children
from families of different size, Van
Bavel (2005) investigated the notion that
by reducing the size of their family, the
parents could invest in the opportunities
of their children. The life course approach
is particularly suited to analyze the experiences
and contexts that may induce individual
couples to apply some form of
birth control.
In family history, but also in the more
“demographic” fields of ageing, migration
and fertility, the life course
approach is fruitful in engendering
innovative, in-depth analysis. But will
this alter the face of historical demography?
Will it attract a new generation of
practitioners and will the results be integrated
in new historical syntheses? In
the following section, a number of
looming problems will be discussed and
tackled.
REVIEWING PITFALLS AND
PROSPECTS
In an earlier review of the life course
approach, I suggested that—give the
relative newness of this methodology—
it is too early to expect new syntheses of
demographic (or social) history based
on life courses (Kok, 2000). Notwithstanding
the innovative research
discussed in the previous paragraph, I
am beginning to suspect that we are not
moving at all in the direction of a new,
coherent vision on demographic history.
To make a stronger impact on historiography,
historical life course demography
needs more focus, more appealing and
relevant results, and more “critical mass”
in the sense of more practitioners. To
realize this, we need to ask ourselves
what exactly is impeding the advance of
the life course in historical demography.
The obstacles can be related to the theoretical
principles of the approach, to the
statistical methods employed, to problems
with data and, finally, to the
research practices in our field. Below, I
identify and discuss these potential
problems in the form of questions:
Can the principle of Agency really be
applied to individual life courses in the
past?
A person’s planning capacity and
belief in his /her effective control is
recognized as a crucial element of the
life course (Gecas, 2003). However, the
concepts and semantics of the life course
approach actually seem to exclude
“agency”. Skolnick (1983) points out
how the life course “is represented as a
socially determined trajectory through
time, shaped by age norms, economic
constraints, institutionalised patterns.
The very word “transition” suggest an
impersonal, agentless movement”. She
states that this approach leads to an
“oversocialized” image of man. The
more or less arbitrary selection of transitions
to denote core events of one’s life
has irritated some scholars (“whose life
course is it, anyway? ”) (Skolnick, 1983,
see also Watkins, 1980). If the concept
of agency is already difficult to integrate
in life course sociology, the problem
seems insurmountable in an historical
context. Given the impossibility of
interviews and the lack of other personal
212
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 212
information, how are we to incorporate
self-perception, capacities and life plans
of individuals in the analysis? And, what
is actually an individual in the context
of the strong collectivities (families,
villages) that characterized many past—
and even present—societies?
Often, decisions about education,
work, migratory moves and even
marriage of individual children were
taken at the household level. Thus,
understanding individual life courses
means understanding the operative principles
and mechanisms at the level of
families and households. To deal with
such collective decision-making, the
concept of “family strategies” is
frequently employed (e.g. Moen and
Wethington, 1992), although lately its
popularity seems to be waning. Indeed,
the concept has been attacked from various
quarters. One problem is that the
goals underlying family or household
strategies (or individual ones, for that
matter) are often simply derived from
economic theories of “utility maximization”.
Maximization (e.g. of income) is
applied to households where, supposedly,
the division of tasks and even life choices
(education, employment) is organized by
balancing (opportunity) costs and benefits.
This “harmonic” model has been
criticized for neglecting the unequal positions
of the sexes and generations within
families, leading to exploitation of some
members for the benefit of others. It has
underestimated the role of culture in
setting non-economic goals such as
adherence to religiously inspired moral
standards or seeking status and
respectability. In addition, experimental
psychologists have demonstrated that
people are generally much more focussed
on averting risks than on maximizing
gains (Tversky and Kahneman, 1992;
Kahneman, 2002; Hobcraft, 2006, 169;
McNicoll, 1992, 405). Historians have
doubted the usefulness of the notion of
strategy—at least in the sense of planning
for the long term—because of the uncertainty,
unpredictability and incompleteness
of information that characterized the
past (e.g. Viazzo and Lynch, 2002). Last
but not least, the “strategy” approach is
berated because, basically, it operates by
inferring motives of individuals from
their behaviour, with all the implications
of circular reasoning (Engelen, 2002).
Indeed, the concept of strategy may
have overemphasized certain aspects
(deliberateness, planfulness) of agency at
the expense of other aspects. Therefore, a
more nuanced and less “strategic” view on
agency is called for. In his seminal article
on fertility behaviour, Leibenstein (1981)
argued that maximization models are
clearly incapable of clarifying decisionmaking.
The practice of acquiring information,
balancing costs and benefits of
alternatives, deciding and evaluating
courses of action is simply too cumbersome.
Given the costs of changing behaviour
(in terms of information, time and
stress to make and monitor decisions),
“nonoptimal passivity” is often the best
solution. In many cases, it is more rational
to follow routines or to delay decisions (in
the future, more and better information
may be available). Leibenstein coins the
term “inert area” to denote traditional,
routine behaviour that is only changed
when strong external pressures render
active decision-making unavoidable.
When does “passive agency” give way to
“active agency”? This may occur, for
instance, in a period of severe economic
depression. In fact, the aim of the Eurasia
project is precisely to study how family
behaviour changes through acute
economic stress. Thus, in their compara-
213
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 213
tive study on mortality, Bengtsson,
Campbell and Lee (2004) have shown
how “traditional” principles of (food)
redistribution within families may be
reversed during periods of crisis. Thus,
the challenge for historical demographers
lies in hypothesizing situations and identifying
periods when passive agency is no
longer adequate.
This is not equivalent to saying that,
given the right economic impetus,
“Culture” (tradition, religion) will give
way to the Agency of rational man.
Several authors have argued that culture
and agency cannot be separated and that
they are in constant interaction. In their
view, agency should be interpreted as
“reflexive monitoring and rationalization
of a continuous flow of conduct”
(Carter, 1995, 61; see also Giddens,
1979; Bengtsson, Campbell and Lee,
2004). This means that people
constantly evaluate, reformulate and
reject their original plans. In fact, the
life course approach that incorporates
past experiences at each decisionmaking
moment (principle of Life Span
Development) is deemed ideally suited
for this dynamic concept of agency
(Greenhalgh, 1995, 22). For instance, in
their study of migration based on a large
number of British life histories, Pooley
and Turnbull (1998) have shown that
many individual trajectories can be
interpreted as combinations of efforts at
long-term improvement with shortterm
adaptations necessitated by adverse
circumstances. Likewise, goals that are
highly inspired by the dominant culture
are, more or less, “negotiable” in daily
decision-making (Carter, 1995).
Culture is not seen at “outside” behavioural
practices. Choices of people are
evaluated by themselves and others and
eventually alter the range of (moral)
choices that make up “culture”. Or, in
the words of Hammel (1990), culture
can be seen as an “ongoing commentary
on behaviour” and actors “move in an
intensely evaluative cloud of commentary,
in the presence of which they select
behavioural alternatives, balancing and
choosing those they think will minimize
moral risk and maximize moral gain”
(Hammel, 1990, 467). Hammel advises
that, in order to understand this continuous
feedback between behaviour and
commentary, we need to (compare)
small groups of actors well situated
within their networks.
These nuanced views on agency can
help us making sense of historical life
courses. In my view, not knowing individual
people’s motives and goals should
not refrain us from hypothesizing about
the most likely goals of people in specific
places and periods (Kok, 2002). We
need to place people carefully within
their historical and local contexts, which
implies we know their general cultural
preferences and behavioural routines.
We then construct a series of hypotheses
about the most likely goals of individuals
and families, given their resources
and their options. By following their
behavioural choices across the life
course, keeping an eye both on past
individual experiences and on the interactions
with historical influences, we
can test under what circumstances these
hypothesized goals were realized and
when they were modified. How to make
meaningful hypotheses? Clearly, even
with a large collection of individual life
courses, we cannot dispense with qualitative
information (Kertzer, 1997). On
the contrary, we need to make a strong
effort to integrate qualitative information
on social groups that are matched
as closely as possible to our cohorts. This
214
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 214
information includes general mentality
(such as openness to outside views),
adherence to religious prescriptions,
knowledge of alternative options, frames
of references for evaluating living standards,
nature of relation with parents
and siblings, attitudes towards risks and
risk-taking et cetera. For instance, in her
study on life courses of servants, Bras
(2002, 2004) made use of letters, diaries
and interviews that helped her to elaborate
hypotheses on the social and
geographical trajectories. Likewise, Van
Poppel and Röling (2003) combined
detailed knowledge on the attitude of
late nineteenth century medical profession
towards “Neo-Malthusian” birth
control with the fertility histories of
general practitioners themselves.
Do event history models focus too much
on single events and too little on processes?
Event history models are designed to
analyse (the timing of) transitions from
one “state” (e.g. unmarried) to another
(married). Frequently, objections are
raised against the selection and definition
of these transitions. The selection is
criticized for being arbitrary or for being
normatively associated with middleclass
and western family concepts. The
transitions often simplify complex and
protracted processes. For instance,
marriage is merely the final moment in a
drawn-out process of courtship and
partner selection. Finally, transitions are
too often depicted as one-time events
whereas they are often reversible. For
instance, leaving home is a typically
repeated event (Hogan, 1991, 1).
Perhaps even more problematic is that
regression models do not add very much
to the explanation of particular transitions.
Too often, researchers are satisfied
with having shown the independent
impact of variables on the hazards of a
particular event. Thus, the approach is
more statistical than analytical (McNicoll,
1991). This concern reflects the warning
of Elder and Pellerin (1998, 293-294):
“History makes its impress on a cohort
one individual at a time. Without an
understanding of the process by which
that impress is made and a direct effort
to model this process, studies may result
only in tables of inexplicable correlations.
”
Studying and explaining processes
instead of single events requires more
dynamic models that those currently
used in (historical) demography. For
instance, the family background of individual
transitions is often incorporated
in the models in the form of a
synchronic snap-shot of household
composition. More elaborate models are
needed that capture the constant change
in family constellation (Uhlenberg and
Mueller, 2003, 126, 141-142). Indeed,
these dynamics will have to be extended
to kin relations beyond the household as
well (Hobcraft, 2006, for an example in
historical demography, see Kesztenbaum,
2005). More elaborate models
(that include several time-dependent
covariates) are needed to study processes
and interactions between those
processes.
Apart from adding more dynamic
information to the covariates, event
history analysis also needs to take
account of behavioural alternatives.
Indeed, one may leave one’s position
(e.g. unmarried) for several others
(cohabitation or marriage). When
studying migration, the destinations
may be similarly conceptualised as
“competing risks” (urban or rural, long
or short distance et cetera). So-called
215
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 215
competing risk analysis can be done
with event history models but also with
multinomial logistic regression (for
examples in historical demography see
respectively Dribe and Stanfors, 2002
and Engelen and Kok, 2003).
In addition, there is a strong need for
research techniques that really unearth the
pathways of life courses. What are typical
and non-typical (the latter can be seen as
innovative or deviant) sequences of transitions?
What factors explain the choice of a
trajectory? In “multi-state demography”
the researchers combine transitions
within various domains of life (residence,
fertility, partnership) to make meaningful
potential trajectories (Courgeau, 2000).
For instance, the “pathways” to the first
child include living at the parental home
(and having a child), leaving home
followed by living alone (and having a
child), leaving home followed by cohabitation
(and having a child), leaving home
followed by marriage (and having a child)
et cetera (Matsuo and Willekens, 2003).
Recently, new techniques allow the
researcher to determine what sequences
are more or less typical. These “datamining”
techniques also provide “association
rules”, that is they indicate under
what conditions a particular sequence is
likely to occur (Ritschard and Oris, 2005,
306-307). Pathways can also be revealed
automatically—and visually attractive—
with “induction trees” (decision trees
induced from data). These trees are
created by statistical programs that show
what background characteristics (e.g.
one’s birthplace, fathers’ occupation) are
most revealing—and in what order—for
a particular life course destination (for
detailed examples, see Ritschard and Oris,
2005, 308-314 and Billari et al., 2006 ).
Current statistical practice is not given
the life course approach its full due, in
the sense of the principle of Life Span
Development. However, more elaborate
analytical and descriptive methods are
being developed that will be able to
reveal processes and trajectories.
Are event history models really integrating
the principle of Time and Place?
The risk inherent in the life course
approach is that too much emphasis is
laid on personal traits and characteristics
and not enough on the social settings,
the options and constraints, in which
individuals operate. Thus, Bourdieu’s
(2000) warning against the “biographical
illusion” may be extended to the life
course approach. Micro-level studies
avoid the problem of “ecological fallacy”
but may subsequently fall into the trap
of “atomic fallacy” (Courgeau, 2000).
Clearly, the principle of Time and Place
urges the researcher to study the interactions
of the life course with the temporal
and local contexts. Has this been done
consistently and convincingly?
The temporal context can be
processed in event history analysis in
several ways. First, time series (for
instance, on food prices) can be
included as time-dependent covariates.
Second, differences between periods can
be disclosed by adding dummy variables
indicating time periods or by contrasting
event history models of subsequent
birth cohorts (e.g. Van Bavel, 2003,
2004a). However, it remains difficult to
translate those finding to “aggregate”
trends. For instance, what can the shifts
in the explanatory power of covariates in
models of fertility behaviour of subsequent
birth cohorts tell us about the
timing and pace of the fertility decline?
Even more difficult is to take account of
macro-level changes in economy and
216
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 216
society. Clearly, the rise of the welfare
state, the shifts in labour markets and
cultural changes such as secularisation
all had profound impacts on the social
pathways, but how to disentangle these
effects, apart from a superficial comparison
of time periods? One way to deal
with this may be to initiate more
comparative studies of life course in
societies that differ in, for instance, their
systems of social security. Also, studies
need to be designed that span really long
periods. For instance, the impact of a
cultural process such as individualization
on life courses might be studied by
looking at how collective identities
(region, religion, social class) gradually
became less important in determining
life choices (Matthijs et al., 2005).
People live their life in multiple
contexts: their household and wider
family, their village or city, their region
and nation. Increasingly, the global
context can be added to this list. All these
contexts contribute in various ways to the
patterning of life courses. In event history
models, researchers generally capture the
effects of the context by adding information
on, for instance, the households (e.g.
household size, socio-economic position)
and localities (e.g. number of inhabitants,
occupational structure) (for an example,
see Kok, 2004). Such “contextual” models
have not proven very satisfactory for various
reasons. Firstly, the developments in
the context (see previous paragraph) is
seldom integrated let alone its effects on
individual life course. So far, adding
context has not contributed to a more
processual approach: contextual models
have “provided only a skeletal view of the
life course” (Elder et al., 2003). Secondly,
the interactions between contexts themselves
are seldom taken into account. For
instance: the effect of school on individual
achievements may depend in part on the
kind of neighbourhood in which the
school is located. Thus, not only are the
contexts of school and neighbourhood
relevant, but also their interactions (Elder
et al., 2003, 15). Lastly, an important
problem with micro level data is that the
outcomes for individuals are seldom independent.
Observations are generally clustered:
people come from the same families,
live in the same village et cetera. In
particular for members of the same
family, group effects are strong (for
instance, through death clustering (Alter
et al., 2001)). If we really want to find out
how contextual factors have an impact on
individual life course, we need to control
not only for hypothesized contextual variables,
but for “unobserved” group effects
(shared heterogeneity) as well.This can be
done by adding “random effects” to linear
models that allow each group its own
coefficients (or only its own intercepts).
Such models are known as multilevel (or
mixed effects or hierarchical) models (e.g.
Courgeau, 2000; Snijders, 2002;
Ritschard and Oris, 2005).
Adding those random effects tends to
change not so much the direction but
the strength and significance of the
observed contextual variables. So far,
multilevel models still need to be integrated
in most software for event history
analysis.
Is the combination of family history, life
course methodology and advanced
statistics beyond the reach of individual
researchers?
Historical demographers have always
had to combine numerous skills. Obviously,
they had to be well versed in
history, sociology and economy, but
they also needed basic knowledge on
217
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 217
fields as diverse as family law, folk
customs, religion, biology and last but
not least, statistics. Interdisciplinarity
was and still is the hallmark of historical
demography (Van der Woude, 1979).
Presently, life course (historical) demography
poses even higher demands. As I
pointed out above, understanding life
courses implies, ideally, a combination
of quantitative and qualitative methods
(Furstenberg, 2003, 667). Interviews,
(auto) biographies, letters and other
sources can help immensely in refining
hypotheses about people who have only
left traces in administrative records.
However, qualitative data needs to be
approached and interpreted with
caution (Clausen, 1998). In the preceding
paragraphs, I have also discussed the
need for (even) more elaborate statistical
models, whereas the current techniques
already pose serious problems, at least
for the average historian. Clearly, it is
unrealistic to expect scholars to combine
themselves all the skills required for this
branch of historical demography
(Kertzer, 1997, 844). It is even more
unrealistic given the short span of
academic courses (including PhD
appointments) and the monodisciplinarity
of most curricula. In other words,
the “learning curve” for historical
demography has become steeper
whereas the “rewards”, in the form of
prestige and career opportunities, have
become less visible and less certain.
If we cannot expect individual
researchers to acquire all the necessary
skills for life course demography, we will
have to change our research practices.The
solution will have to come from interdisciplinary
research teams that are able to
tackle the complexities of life course
research (Elder et al., 2003, 15). Ideally,
such a research team consists of historians,
life course sociologists, data experts and
statisticians. Typically, this combination
of skills is not found in any faculty or
department, often not even within the
same university. Building multidisciplinary
teams requires interuniversity, sometimes
even international, cooperation.
Recently, the opportunities for new forms
of scientific cooperation have expanded
enormously.
Already, international teams have taken
up the challenge posed to historians by
the recent development of East Asia.
Economic historians are reviewing the
long-term economic potential of China
and wonder to what extent China’s backwardness
of the XIXth and XXth centuries
had been caused by internal of external
factors (Pomeranz, 2000). In a similar
vein, demographic historians question the
“divergence” between European demography
(with its supposedly rational family
formation) and the demography of the
rest of the world, in particular China
(Bengtsson, Campbell and Lee, 2004;
Engelen and Wolf, 2005; Chuang, Engelen
and Wolf, 2006). Without the internet,
such interdisciplinary and international
teamwork is hardly conceivable.
However, there are clear limits to the
building of “virtual” research teams. They
appear to be most effective when organized
along the “traditional” line of
frequent face-to-face-contacts (Matsat,
2006). Electronic communication stimulates
the team’s research by enhancing
discussion, text-reviewing etcetera.
However, it does not solve the bottleneck
of data.
Is building life course datasets too time-
consuming?
Notwithstanding the gains made by
computerization, creating datasets with
218
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 218
reconstructed life courses is still enormously
time-consuming. Creating such
datasets with a sufficient number (that
is, allowing for significance tests) of life
courses is not within reach of individual
Ph D students. Generally, such databases
have been created by research
teams within faculty departments who
tended to concentrate on particular
cities or regions. Also, their reconstructions
were often limited to the rather
selective subset of individuals who
remained in that particular city or
region. This means that the migration
experience still needs to be integrated in
the life course. Moreover, it is difficult
to generalize from these disparate local
and regional collections of life courses.
It seems to me there are two ways to
overcome this dearth of life course data.
The first is data-sharing between
researchers and the second is investing
in public-use datasets.
Data-sharing as a new research practice
is frequently advocated in the
context of collaboratories, (virtual)
research teams that have pooled their
data. A collaboratory can be seen as “a
laboratory without walls, where scientists
are connected to each other, to
instruments, and to data independent of
time and location” (Finholt, 2002, 73).
Collaboratories are rather popular in the
beta sciences, in particular in those
fields where the use of costly instruments
(e.g. telescopes) can be shared.
However, many collaborative projects
have already collapsed. In practice, it
proves very difficult to reconcile the
more immediate needs of individual
researchers with potential long-terms
benefits of collective work. Why should
one relinquish the dataset one has spent
years to create to others who may profit
disproportionate to their own imput?
The success of a collaboratory depends
on its capacity to deal with the problems
of free-riding, trust and instability
(Matsat, 2006; Wouters and Schröder,
2003).
How to avoid that some participants
benefit from data-sharing and joint
research without contributing themselves
and thus undermine the project as
a whole? How to avoid doubts about the
intentions of partners who may “scoop”
one’s results? In other words, how to
engender the trust without which a
collaboratory cannot function? Finally,
how to avoid the risk that the team falls
apart before reaching its goals? Although
we still need to discover the “best practices”
for collaboratories in the humanities
and social sciences, it seems evident
that certain guidelines should be
followed. First of all, the benefits have to
be recognized by all parties involved:
that is, their research will be enhanced
in terms of quality (e.g. more analytic
depth by adding comparisons), quantity
(more publications) and more prestige
by joining a team. The project needs to
focus on a small number of research
issues that will be tackled by a restricted
number of peers that are already well
acquainted (through their research
network, conferences et cetera). The
participants agree on renouncing exclusive
property rights of their data and
devise procedures to ensure the quality
of the pooled data (standardizing, peerreviewing,
quality checks). They also
agree on the targets of the projects: what
publications are foreseen, who will be
first author for what paper et cetera.
Finally, it seems advisable to agree on
the duration of the collaboratory.
However, what is to be done with the
combined dataset when the project has
ended? It does not make sense to split it
219
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 219
up again in the privately owned original
components. Probably the best course of
action is to make the dataset public in
the form of a website or to deposit it at a
data archive. In that case, everyone can
benefit from it. However, the research
team can and will no longer guarantee
quality control of the data (De Moor
and Van Zanden, 2007).
Guaranteed quality control is precisely
the hallmark of public-use datasets that
are being created in various countries,
such as Canada, Sweden and The
Netherlands (Hall et al., 2000). Indeed,
these datasets have already been used for
a host of innovative research projects,
including some comparative ones.
Historical microdatasets have been
employed in researches ranging from
epidemiology and genetics to demography
and sociology. Still, the use of these
datasets is often not as extensive as has
been anticipated originally. One probable
reason for this is the fact that the
data are stored in a “general” format that
needs further processing before it can be
used for any specific research question.
Furthermore, the complexity of the
datastructure still requires (extensive)
visits of the datacentre and/or the costs
of producing ready made datasets by
programmers. Not surprisingly, the bulk
of publications based on public use
datasets are generally produced by a
limited number of already highly
trained experts. It seems advisable that
managers of public use datasets create
interfaces that stimulate access to the
data as well as produce simplified
datasets that can be used in training
courses and will answer simple research
questions. Without efforts in this direction,
it is doubtful that the number of
researchers active in this field can be
expanded.
Will the life course approach attract a
new generation of researchers to historical
demography?
Judging from the number of sessions
in, for instance, the American and European
Social Science History Conferences
and the number of publications in international
journals, historical demographers
form a vibrant, successful and
tightly-knit research community.
However, an age profile of this community
would reveal an insufficient influx of
young researchers. Clearly, young
researchers need to be assisted and motivated
by proper training, by reducing
their costs by giving them access to
public use or shared datasets, and by
increasing their benefits by giving them
opportunity to join multidisciplinary
teams and to co-author articles. However,
motivation will have to come, first and
foremost, from the research topics.
It is likely that historians interested in
the subjects of household, family and
gender will find their way in increasing
numbers to the life course approach. This
is testified by the successes of the “new
family history” which is largely based on
life courses and concomitant statistical
techniques. However, family history as
such is—and is likely to remain—a relatively
small subfield within history at
large. Historical demography has always
recruited most of its followers from the
social sciences. The problem of attracting
new scholars to demography lies in part
with demography itself. Currently, social
scientists (demographers among them)
complain that the field is lagging behind
in theory-building, is focussing too much
on quantification, and is loosing touch
with developments in other areas of the
social sciences (Greenhalgh, 1995). As
McNicoll remarked, rather cynically:
220
JAN KOK
09-kok.qxp 16/07/2007 16:24 Page 220
“The bankclerkly and backroom activities
that now make up most of population
studies […] are increasingly
divorced from any large, cumulative
social scientific enterprise […] and […]
they seem ill-suited for treating the kinds
of population-linked issues that may
soon appear on the public policy agenda.”
(McNicoll, 1991, 400).
Can historical demography avoid the
image of an arcane subdiscipline within
history or, even worse, of a “backroom”
activity within social science? In my
opinion, a number of topics can and
should be addressed that will increase
interest from young scholars and policy
makers alike. We have already discussed
ageing, surely one of the most pressing
population issues in western societies.
Life course studies are crucial for studying
shifts in intergenerational relations,
in the development of age norms, in the
transformation and social implications
of welfare systems et cetera. Another
important subject is (international)
migration and the integration of ethnic
minorities. As we pointed out, life
course research designs, in particular
those spanning several generations, can
yield important insights in the family
mechanisms of (return) migration and
social mobility. Finally, the past decades
have witnessed a virtual revolution in
family formation, at least in Western
Europe. Families are no longer formed
exclusively through marriage and they
are much less stable than in the past.
What are the causes and implications of
increased rates of cohabitation and
divorce? Why has fertility fallen below
replacement level and what will be the
consequence? What are the long-term
effects on individuals from living in
different (subsequent and simultaneous)
household settings? Clearly, many of the
questions surrounding the “second
demographic transition” are already
being addressed by demographers (Van
der Kaa, 1991). However, the origins of
this dramatic shift in behaviour are
understudied. Also, demographic
behaviour—in particular individual
variation and change over time—is
seldom placed in the contexts of groups
(social or religious), of time and place.
In other words, a contextualized and
historicized analysis of the second
demographic transition should be put
high on the research agenda.
CONCLUSIONS
The analysis of standardized biographies
is not merely a new research technique
for historical demography. Indeed,
the life course approach can be seen as a
new paradigm because demographic
behaviour can now be studied in its true
complexity, that is, involving past experiences,
alternative options, multiple
processes and ongoing interaction with
place and time. Already, historical life
course studies have invigorated the field
of family history. This field will be
strengthened even further by integrating
qualitative and quantitative techniques.
However, various problems may
obstruct the advance of the life course
approach. First, although event history
analysis has proven a highly valuable tool
for showing the relative strength of
factors involved in demographic behaviour,
it has still fallen short of elucidating
life courses as processes. More elaborate
models will have to incorporate the
impact of past experiences, of behavioural
alternatives and of interactions
among processes and contexts. Secondly,
the learning curve for historical demography
threatens to become too steep, in
221
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
09-kok.qxp 16/07/2007 16:24 Page 221
particular when such advanced techniques
need to be mastered. More
(inter)disciplinary team work might solve
this problem. Third, the life course
approach is highly demanding in terms
of data. These demands can hardly be
met by ordinary faculty-based research
teams, let alone by individual researchers.
Research (and training) practices in this
field will have to shift towards data-sharing
within collaboratories and towards
exploiting public use datasets. Finally, in
order to attract scholars from various
fields, research need to be focused on
topics such as population ageing, below
replacements fertility, migration and
ethnic integration. Datasets of life
courses will prove an invaluable tool for
the study of our day and age.
Jan KOK
Virtual Knowledge Studio for the
Humanities and Social Sciences
Cruquiusweg 31
1019 AT Amsterdam
jan.kok@vks.knaw.nl
222
JAN KOK
BIBLIOGRAPHICAL REFERENCES
ALLISON, P. D. (1984), Event History Analysis.
Regression for Longitudinal Data,
Beverly Hills etc., Sage Publications.
ALTER, G. (1988), Family Life and the
Female Life Course. The Women of Verviers,
Belgium, 1849-1880, Madison, University
of Wisconsin Press.
ALTER, G. (1998), « L’Event History Analysis
en démographie historique. Difficultés et
perspectives », Annales de Démographie
Historique, 2, 25-35.
ALTER, G., ORIS, M., BROSTRÖM, G.
(2001), “The Family and Mortality: a
Case Study from Rural Belgium”, Annales
de Démographie Historique (1), 11-31.
ALTER, G., CAPRON, C. (2004), “Leavers and
Stayers in the Belgian Ardennes”, 117-141,
in The Road to Independence. Leaving Home
in Western and Eastern Societies, XVIth-XXth
Centuries, edited by F. van Poppel, M. Oris
and J. Lee, Bern, Peter Lang.
ALTER, G. et al. (2002), “When Dad Died:
Household Economy and Family Culture
in Nineteenth-Century East-Belgium”,
389-420, in When Dad Died. Individuals
and Families Coping with Family Stress in
Past Societies, edited by R. Derosas and M.
Oris, Bern, Peter Lang.
ANDERTON, D. L., BEAN, L. L. (1985),
“Birth Spacing and Fertility Limitation: A
Behavioral Analysis of a NineteenthCentury
Frontier Population”, Demography,
22 (2), 169-183.
AUGUSTINS, G. (2002), “The Perpetuation of
Families and the Molding of Personal
Destinies”, 322-348, in Family Life in the
Long Nineteenth Century, 1789-1913. The
History of the European Family, vol. 2, edited
by D. I. Kertzer and M. Barbagli, New
Haven and London, Yale University Press.
BENGTSSON, T., LINDSTROM, M. (2000),
“Childhood Misery and Disease in Later
Life: the Effects on Mortality in Old Age
of Hazards Experienced in Early Life,
Southern Sweden, 1760-1894”, Population
Studies, 54, 263-277.
BENGTSSON, T., CAMPBELL, C., LEE, J. Z.
(eds.) (2004), Life under Pressure. Mortality
and Living Standards in Europe and
Asia, 1700-1900, Boston MA, MIT Press.
BENGTSSON, T., DRIBE, M. (2006), “Deliberate
Control in a Natural Fertility Population:
Southern Sweden 1766-1864, Demography,
43 (4),727-746.
BILLARI, F., FÜRNKRANZ, J., PRSKAWETZ, A.
(2006), “Timing, Sequencing, and Quantum
09-kok.qxp 16/07/2007 16:24 Page 222
223
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
of Life Course Events: A Machine Learning
Approach”, European Journal of Population,
22, 37-65.
BLAIKIE, A. (1993), Illegitimacy, Sex, and
Society: Northeast Scotland, 1750-1900,
Oxford, Oxford University Press.
BLOSSFELD, H.-P., ROWHER, G. (1995),
Techniques of Event History Modeling: New
Approaches to Causal Analysis, Mahwah
New Jersey, Erlbaum.
BONNAIN, R. (1996), “ Houses, Heirs and
Non-Heirs in the Adour Valley: Social
and Geographical Mobility in the Nineteenth-Century
France”, History of the
Family. An International Quarterly, 1, (3),
273-295.
BOURDELAIS, P. (1999), “Demographic Aging:
a Notion to Revisit”, History of the Family.
An International Quarterly, 4, 31-50.
BOURDIEU, P. (2000), “ The Biographical
Illusion ”, 299-305, in Identity: a Reader,
edited by P. Du Gay, J. Evans and P.
Redman, London, Sage.
BRÄNDSTRÖM, A. (1996), “ Life Histories of
Single Parents and Illegitimate Infants in
Nineteenth Century Sweden ”, History of
the Family. An International Quarterly 2,
205-226.
BRAS, H. (2002), Zeeuwse meiden. Dienen in
de levensloop van vrouwen, ca. 1850-1950,
Amsterdam, Aksant.
BRAS, H. (2004), “Social Change, Service
and Youth in the Lives of Rural-Born
Dutch Women, 1840-1940, Continuity
and Change, 19 (2), 241-264.
BRAS, H., NEVEN, M. (2007), “The Effects of
Siblings on the Migration of Women in
Two Rural Areas of Belgium and the
Netherlands, 1829-1940”, Population Studies,
61 (1), 1-19.
BRESCHI, M., MANFREDINI, M. (2002),
“Parental Loss and Kin Networks: Demographic
Consequences on Children in an
Italian Village”, 381-400, in When Dad
Died. Individuals and Families Coping
with Family Stress in Past Societies, edited
by R. Derosas and M. Oris, Bern, Peter
Lang.
BROWN, J. C., GUINNANE, T. W. (2001),
The Fertility Transition in Bavaria. Yale
Economic Growth Center Discussion
Paper No. 821.
BROWN, J. C., GUINNANE, T. W. (2003),
Two Statistical Problems in the Princeton
Project on the European Fertility Transition.
Yale University Economic Growth
Center Discussion Paper No. 869.
BRUNET, G., BIDEAU, A., GERBE, M.-N.
(2002) « Vers une insertion sociale ? Le
mariage des enfants abandonnés à Lyon
au XIXe siècle », Annales de Démographie
Historique, 2, 161-174.
BULDER, E. (1993), The Social Economics of
Old Age. Strategies to Maintain Income in
Later Life in the Netherlands (1880-1940),
Amsterdam, Thesis.
CAMPBELL, C., LEE, J. (2002), “Widowhood
and Orphanhood in Late Imperial
Liaoning, 1789-1909”, 313-334, in
When Dad Died. Individuals and Families
Coping with Family Stress in Past
Societies, edited by R. Derosas and M.
Oris, Bern, Peter Lang.
CAPRON, C., ORIS, M. (2000), « Ruptures de
cohabitation entre parents et enfants dans les
villes et campagnes du pays de Liège au XIXe
siècle », 229-269, in De l’usage des seuils.
Structures par âges et âges de la vie, edited by
A. Bideau, P. Bourdelais, J. Légaré , Paris,
Société de Démographie Historique.
CARTER, A. (1995), “Agency and Fertility: For
an ethnography of practice”, 55-85, in Situating
Fertility. Anthropology and Demographic
Inquiry, edited by S. Greenhalgh,
New York, Cambridge University Press.
CHUANG, Y.-C., ENGELEN, T., WOLF, A. P.
(eds). (2006), Positive or Preventive. Fertility
Developments in Taiwan and the
Netherlands, 1850-1950, Amsterdam,
Aksant.
CLAUSEN, J. A. (1998), “Life Reviews and Life
Stories”, 189-212, in Methods of Life Course
Research: Qualitatitive and Quantitative
09-kok.qxp 16/07/2007 16:24 Page 223
224
JAN KOK
Approaches, edited by J. Z. Giele en G.H.
Elder, Thousand Oaks, SAGE.
COALE, A. J. (1973), “The Demographic
Transition Reconsidered”, 53-72, in International
Population Conference, vol 1, Liège,
IUSSP.
COURGEAU, D. (2000), New approaches and
methodological innovations in the study of
partnership and fertility behaviour, paper
for Flagship conference Partnership and
fertility – a revolution, Brussels 2000.
COURGEAU, D., BACCAÏNI, B. (1991), “Multilevel
Analysis in the Social Sciences”, Population:
An English Selection, 39-71.
COURGEAU, D., LELIÈVRE, E. (1997),
“Changing Paradigm in Demography”,
Population: An English Selection, 9, 1-10.
DAVID, P. A., MROZ, T. A. (1989),
“Evidence of Fertility Regulation Among
Rural French Villagers, 1749-1789. A
Sequential Econometric Model of BirthSpacing
Behavior”, European Journal of
Population, 5, 1-26 and 173-206.
DE MOOR, T.,VAN ZANDEN, J. L. (2007),
“Do ut des. Collaboratories as a ‘New’
Method for Scholarly Communication and
Cooperation for Global and World
History”, History and Computing.
DEROSAS, R., ORIS, M. (2002), When Dad
Died. Individuals and Families Coping with
Family Stress in Past Societies, Bern, Peter
Lang.
DORSMAN, J., STAVENUITER, M. (1993),
Nooit gehuwd, maar niet alleen. Vrijgezelle
vrouwen uit de arbeidende klasse in de
tweede helft van de negentiende eeuw,
Hilversum, Verloren.
DRIBE, M., STANFORS, M. (2002), Leaving
home in Post-War Sweden. A Micro-level
Analysis of the Determinants of Leaving the
Parental Home in Three Birth Cohorts.
Lund papers in Economic History 81.
DRIBE, M., LUNDH, C. (2005), “Determinants
of Servant Migration in NineteenthCentury
Sweden”, Continuity and Change,
20 (1), 53-91.
ELDER, G. J. Jr, KIRKPATRICK JOHNSON, M.,
CROSNOE, R. (2003), “The Emergence
and Development of Life Course
Theory”, 3-19, in Handbook of the Life
Course, edited by J. T. Mortimer and M. J.
Shanahan, New York, Plenum.
ELDER, G. H., PELLERIN, L. A. (1998),
“Linking History and Human Lives”,
264-294, in Methods of Life Course
Research: Qualitatitive and Quantitative
Approaches, edited by J. Z. Giele and G.
H. Elder, Thousand Oaks, SAGE.
ENGELEN, T. (2002), “Labour Strategies of
Families: A Critical Assessment of an
Appealing Concept”, International Review
of Social History, 47 (3), 453-464.
ENGELEN, T., KOK, J. (2003), “Permanent
Celibacy and Late Marriage in the
Netherlands, 1890-1960”, Population, 58
(1), 67-96.
ENGELEN, T., WOLF, A. P. (eds.) (2005),
Marriage and the Family in Eurasia.
Perspectives on the Hajnal hypothesis,
Amsterdam, Aksant.
FAUVE-CHAMOUX, A. (ed.) (2004), Domestic
Service and the Formation of European
Identity. Understanding the Globalization
of Domestic Work, XVIth-XXIst Centuries,
Bern, Peter Lang.
FAUVE-CHAMOUX, A., ARRIZABALAGA, M.-P.
(eds.) (2005), Family Transmission in
Eurasian Perspective, special issue of
History of the Family. An International
Quarterly, 10 (3).
FINHOLT, T. A. (2002), “Collaboratories”,
Annual Review of Information Science and
Technology, 36, 73-107.
FURSTENBERG, F. (2003), “Reflection on the
Future of the Life Course ”, 661-670, in
Handbook of the Life Course, edited by J. T.
Mortimer and M. J. Shanahan, New York,
Kluwer Academic/Plenum Publishers.
GECAS, V. (2003), “Self-Agency and the life
course”, 369-388, in Handbook of the Life
Course, edited by J. T. Mortimer and M. J.
Shanahan, New York, Kluwer Academic/Plenum
Publishers.
09-kok.qxp 16/07/2007 16:24 Page 224
225
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
GIDDENS, A. (1979), Central Problems in
Social Theory: Action, Structure and
Contradiction in Social Analysis, London,
Macmillan.
GJERDE, J. (1985), From Peasants to Farmers.
The Migration from Balestrand, Norway, to
the Upper Middle West, Cambridge,
Cambridge University Press.
GREENHALGH, S. (1995), “Anthropology
Theorizes Reproduction: Integrating
Practice, Political Economic, and Feminist
Perspectives”, 3-28, in Situating
Fertility. Anthropology and Demographic
Inquir, edited by S. Greenhalgh, New
York, Cambridge University Press.
GRIBAUDI, M. (1987), Itinéraires ouvriers.
Espaces et groupes sociaux à Turin au début
du XXe siècle, Paris, EHESS.
HAAN, M.D. (2005), “Studying the Impact
of Religion on Fertility in NineteenthCentury
Canada. The use of Direct
Measures and Proxy Variables”, Social
Science History, 29 (3), 373-411.
HAJNAL, J. (1982), “Two Kinds of Pre-Industrial
Household Formation Systems”, Population
and Development Review, 8, 449-494.
HALL, P. K., MCCAA, R., THORVALDSEN, G.
(eds.) (2000), Handbook of International
Historical Microdata for Population Research,
Minneapolis, Minnesota Population
Center.
HAMMEL, E. A. (1990), “A Theory of
Culture for Demography”, Population and
Development Review, 16 (3), 455-485.
HAREVEN, T. K. (1978), “Cycles, Courses and
Cohorts: Reflections on Theoretical and
Methodological Approaches to the Historical
Study of Family Development”, Journal
of Social History, 12, 97-109.
HAREVEN, T. K. (1982), Family Time and
Industrial Time: the Relationship between
the Family and Work in a New England
Industrial Community, New York,
Cambridge University Press.
HAREVEN, T. K. (2000), Families, History,
and Social Change, Life Course & CrossCultural
perspective, Boulder/Oxford,
Westview Press.
HOBCRAFT, J. N. (2006), “The ABC of
Demographic Behaviour: How the Interplays
of Alleles, Brains and Contexts over
the Life Course Should Shape Research
Aimed at Understanding Population
Processes”, Population Studies, 60 (2), 153-
187.
HOGAN, D. P. (1991), “Reintroducing
Culture in Life Course Research”,
Contemporary Sociology, 20 (1), 1-4.
KAHNEMAN, D. (2002), Maps of bounded
rationality. A perspective on intuitive judgement
and choice. Nobel prize lecture.
KERTZER, D. I. (1997), “Qualitative and
Quantitative Approaches to Historical
Demography”, Population and Development
Review, 23 (4), 839-846.
KERTZER, D. I., KOBALL, H., WHITE, M. J.
(1997), “Growing up as an Abandoned
Child in Nineteenth-Century Italy”,
History of the Family. An International
Quarterly, 2, 211-228.
KERTZER, D. I., SIGLE, W. (1998), “The
Marriage of Female Foundlings in Nineteenth-Century
Italy”, Continuity and
Change, 13, 201-220.
KESZTENBAUM, L. (2005), Places of live
events as bequestable wealth. Familial
territory and migration in France, 19th
and 20th centuries, paper presented at the
IUSSP seminar New history of Kinship
30 september–2 october 2005, Paris,
France.
KOK, J. (2000), “Transities en trajecten. De
levensloopbenadering in de sociale geschiedenis”,
Tijdschrift voor Sociale Geschiedenis
26, 309-329.
KOK, J. (2002), “The Challenge of Strategy:
A Comment”, International Review of
Social History, 47 (3), 465-485.
KOK, J. (2004), “Choices and Constraints in
the Migration of Families. Central
Netherlands 1850-1940”, History of the
Family. An International Quarterly, 9 (2),
137-158.
09-kok.qxp 16/07/2007 16:24 Page 225
226
JAN KOK
KOK, J., MANDEMAKERS, K., WALS, H.
(2005), “City Nomads. Changing Residence
as a Coping Strategy, Amsterdam
1890-1940”, Social Science History, 29,
15-43.
KOK, J., VAN BAVEL, J. (2005), The intergenerational
transmission of fertility in
rural Holland, 1850-1940, paper
presented at 30th annual meeting of the
Social Science History Association,
November 4, 2005 Portland USA.
LASLETT, P. (1988), “Family, Kinship, and
Collectivity as Systems of Support in
Preindustrial Europe”, Continuity and
Change, 3 (2), 153-175.
LEIBENSTEIN, H. (1981), “Economic Decision
Theory and Human Fertility Behavior:
A Speculative Essay”, Population and
Development Review, 7 (3), 381-400.
LESGER, C., LUCASSEN, L., SCHROVER, M.
(2002), “Is there life outside the migrant
network? German immigrants in XIXth
century Netherlands and the need for a
more balanced migration typology”,
Annales de Démographie Historique, (2),
29-50.
LEVENE, A. (2003), “The Origins of the
Children of the London Foundling
Hospital, 1741-60: a Reconsideration”,
Continuity and Change, 18 (2), 201-235.
LIEFBROER, A. C., DYKSTRA, P. A. (2000),
Levenslopen in verandering. Een studie naar
ontwikkelingen in de levenslopen van
Nederlanders geboren tussen 1900 en 1970,
The Hague,WRR.
LUCASSEN, L. (2004), “De selectiviteit van
blijvers: een reconstructie van de sociale
positie van Duitse migranten in Rotterdam
(1870-1885)”, Tijdschrift voor Sociale
en Economische Geschiedenis, 1 (2), 92-115
LUCASSEN, L. (2005), “Huwelijken van
Duitse migranten in Nederland (1860-
1940). De rol van herkomst, religie,
beroep en sekse”, Tijdschrift voor Sociale en
Economische Geschiedenis, 2, (2), 54-80.
MASSEY, D., ARANGO, J., HUGO, G. et al.
(1993), “Theories of International
Migration: A Review and Appraisal”,
Population and Development Review, 19
(3), 431-466.
MATSAT, U. (2006), Knowledge Management
in a Virtual Organization: Are embedded
online communities of practice more
successful than exclusively virtual ones?
Report Netherlands Organization for
Scientific Research program Society and the
Electronic Highway.
MATSUO, H., WILLEKENS, F. (2003), Event
histories in the Netherlands Fertility and
Family Survey 1998. A Technical report.
Population Research Centre. Research
report 03-1 (Groningen PRC).
MATTHIJS, K., KOK, J., ENGELEN, T. et al.
(2005), A turn of fate. Life courses in the
low countries between 1850 and 1940.
Paper presented at 30th annual meeting of
the Social Science History Association,
November 3-6, 2005 Portland USA.
MATTHYS, C., VANHAUTE, E. (2006), “A
‘silent class’ and a ‘quiet revolution’.
Female domestics and fertility decline in
Flanders”, 1-23, in Were Women Present at
the Fertility Decline, edited by A. Janssens,
Bern, Peter Lang Publishers.
MAYER, K. U., TUMA, N. B. (eds.) (1990),
Event history analysis and life course research,
Madison Wisc., University of Wisconsin
Press.
MAYER, K. U. (1991), “Life courses in the
welfare state”, 171-186, in The Life Course
and Social Change: Comparative Perspectives,
edited by W.R. Heinz, Weinheim,
Deutsche Studienverlag.
MCNICOLL, G. (1992), “The Agenda of
Population Studies: A Commentary and
Complaint”, Population and Development
Review, 18 (3), 399-420.
MODELL, J., HAREVEN, T. K. (1978), “Transitions:
Patterns of Timing”, 245-269, in
Transitions: The Family and the Life Course
in Historical Perspective, edited by T.K.
Hareven, New York: Academic Press.
MOEN, P., WETHINGTON, E. (1992), “The
Concept of Family Adaptive Strategies”,
Annual Review of Sociology, 18, 233-251.
09-kok.qxp 16/07/2007 16:24 Page 226
227
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
NAVE-HERZ, R. (1991), “Historical Tendencies
in the Development of Family Experiences
and Labour Force Participation
Patterns of Women”, 130-143, in The Life
Course and Social Change: Comparative
Perspectives, edited by W.R. Heinz, Weinheim,
Deutsche Studienverlag 1991.
NEVEN, M. (2002), “The Influence of the
Wider Kin Group on Individual LifeCourse
Transitions in Herve (Belgium)”,
Continuity and Change, 17 (3), 405-535.
NEVEN, M. (2005), « Établir ses enfants
dans une région rurale en déclin. Sélection
et égalitarisme dans le Pays de Herve
durant la seconde moitié du XIXe siècle”,
167-199, in Le mariage dans l'est de la
Wallonie, edited by P. Servais and G. Alter,
Louvain-la-Neuve, Academia.
NOORDAM, D. J. (1986), Leven in Maasland.
Een hoogontwikkelde plattelandssamenleving
in de 18e en het begin van de 19e
eeuw, Hilversum,Verloren.
ORIS, M. (2003), “The History of Migration
as a Chapter in the History of the European
Rural Family: An Overview”, History of the
Family. An International Quaterly, 8 (2),
187-215.
ORIS, M., BRUNET, G., WIDMER, E.,
BIDEAU, A. (eds.) (2007), Les fratries. Une
démographie sociale de la germanité, Bern,
Peter Lang.
ORIS, M., RITSCHARD, G., RYCZKOWSKA, G.
(2005), “Siblings in a (Neo-)Malthusian
Town. From Cross-sectional to Longitudinal
Perspectives”, Historical Social
Research, 30 (3), 171-194.
PERRIER, S. (2000), “Coresidence of Siblings,
Half-Siblings, and Step-siblings in Ancien
Régime France”, History of the Family. An
International Quarterly, 5 (3), 299-314.
POMERANZ, K. (2000), The Great Divergence:
China, Europe, and the Making of
the Modern World Economy, Princeton NJ,
Princeton University Press.
POOLEY, C., TURNBULL, J. (1998), Migration
and Mobility in Britain since the XVIIIth
century, London, UCL Press.
REAY, B. (1996), Microhistories: Demography,
Society and Culture in Rural England
1800-1939, Cambridge, Cambridge
University Press.
RITSCHARD, G., ORIS, M. (2005), “Life
Course Data in Demography and Social
Sciences: Statistical and Data Mining
Approaches”, 289-320, in Towards an Interdisciplinary
Perspective on the Life Course,
edited by P. Ghisletta, J.-M. le Goff, R.
Levy et al., Advances in Life Course
Research vol. 10, Amsterdam, Elsevier.
ROBINSON, W. S. (1950), “Ecological Correlations
and the Behavior of Individuals”,
American Sociological Review, 15, 351–357.
ROSENTAL, P.-A. (1999), Les sentiers invisibles.
Espace, familles et migrations dans la
France du XIXe siècle, Paris, EHESS.
SETTERSTEN, R. A. (ed.) (2002), Invitation
to the Life Course: Toward New Understandings
of Later Life, Amityville,
Baywood Publishing.
SKOLNICK, A. (1983), “Looking at Lives, or
‘Whose life course is it, anyway?”’,
Contemporary Sociology, 12 (4), 386-388.
SNIJDERS, T. A. B (2002), Multilevel analysis:
an introduction to basic and advanced multilevel
modeling, London, Sage Publications.
SZOLTYSEK, M. (2007), “Science without
Laws? Model Buliding, Micro Histories
and the Fate of the Theory of Fertility
Decline”, Historical Social Research, vol.
32, 10-41.
SZRETER, S. (1996), Fertility, Class and
Gender in Britain 1860-1940, Cambridge,
Cambridge University Press.
THANE, P. (2003), “Social Histories of Old
Age and Aging”, Journal of Social History,
36, 93-107.
THOMAS, W. I., ZNANIECKI, F. (1918-1920),
The Polish Peasant in Europe and America,
2 vols, Boston, Richard G. Badger.
TVERSKY, A., KAHNEMAN, D. (1992),
“Advances in Prospect Theory. Cumulative
Representation of Uncertainty”, Journal
of Risk and Uncertainty, 5, 297-323.
09-kok.qxp 16/07/2007 16:24 Page 227
228
JAN KOK
UHLENBERG, P., MUELLER, M. (2003),
“Family Context and Individual Wellbeing.
Patterns and Mechanisms in Life
Course Perspective”, 123-148, in Handbook
of the Life Course, edited by J.T.
Mortimer and M.J. Shanahan, New York,
Kluwer Academic/Plenum Publishers.
VAN BAVEL, J. (2003), “Birth Spacing as a
Family Strategy. Evidence from XIXth
Century Leuven, Belgium”, The History of
the Family: An International Quarterly, 8
(4), 585-604.
VAN BAVEL, J. (2004a), “Deliberate Birth Spacing
before the Fertility Transition in Europe:
Evidence from Nineteenth-Century
Belgium”, Population Studies, 58 (1), 95-107.
VAN BAVEL, J. (2004b), “Diffusion Effects
in the European Fertility Transition:
Historical Evidence from within a Belgian
Town (1846-1910)”, European Journal of
Population, 20 (1), 63-85.
VAN BAVEL, J. (2005), “The Effect of Fertility
Limitation on Intergenerational Social
Mobility: the Quality-Quantity Trade-off
during the Demographic Transition”,
Journal of Biosocial Sciences, 1-17.
VAN BAVEL, J., KOK, J. (2005), Are deliberate
birth spacing effects in fact statistical and
bio-demographic artifacts? A critical study
with data from the Netherlands, paper
presented at the XXV IUSSP International
Population Conference, Tours 23 July.
VAN DE KAA, D. (1991), “Emerging Issues
in Demographic Research for Contemporary
Europe”, Population Studies, 45
Supplement, 189-206.
VAN DER LINDEN, M. (2002), “Introduction”,
1-23, in Rebellious families. Household
Strategies and Collective Action in the Nineteenth
and Twentieth Centuries, edited by J.
Kok, Oxford/New York, Berghahn Books.
VAN POPPEL, KOK, J., KRUSE, E. (1995),
« La cause principale des différences c'est
le crime : essai de explication de la
surmortalité des enfants illégitimes à La
Haye au milieu du XIXe siècle », Annales de
Démographie Historique, 241-275.
VAN POPPEL, F., DE JONG, J., LIEFBROER, A.
(1998), “The Effects of Paternal Mortality
on Son’s Social Mobility: a Nineteenth
Century Example”, Historical methods, 31
(3), 101-112.
VAN POPPEL, F., RÖLING, H. (2003), “Physicians
and Fertility Control in the Netherlands”,
Journal of Interdisciplinary History,
34 (2), 155-185.
VAN POPPEL, F., ORIS, M., LEE, J. Z. (eds.)
(2004), The Road to Independence. Leaving
Home in Western and Eastern Societies
XVIth-XXth Centuries, Bern, Peter Lang.
VAN POPPEL, F, LIEFBROER, A. C. (2005),
“Living Conditions during Childhood
and Survival in Later Life—Study Design
and First Results”, Historical Social
Research, 30 (3), 265-285.
VAN DER WOUDE, A. M. (1979), “De historische
demografie in de ontwikkeling van de
geschiedwetenschap”, 157-199, in Theoretische
en methodologische aspecten van de
economische en sociale geschiedenis 1, edited
by P. A. M. Geurts and F. A. M. Messing,
The Hague, Martinus Nijhoff.
VIAZZO, P. P., LYNCH, K. A. (2002),
“Anthropology, Family History, and the
Concept of Strategy”, International
Review of Social History, 47 (3), 423-452.
WATKINS, S. C. (1980), “On Measuring
Transitions and Turning Points”, Historical
Methods, 12, 181-186.
WILLEKENS, F. (1999), Theoretical and Technical
orientations toward Longitudinal
research in the Social Sciences, paper
presented at the workshop on longitudinal
research in Social Sciences: A Canadian
Focus, Windermere Manor, London,
Ontario, October 25-27.
WOUTERS, P., SCHRÖDER, P. (eds.). (2003),
Promise and Practice in Datasharing,
Amsterdam, NIWI-KNAW.
YAMAGUCHI, K. (1991), Event History
Analysis, Beverly Hills/New York, Sage.
09-kok.qxp 16/07/2007 16:24 Page 228
229
PRINCIPLES AND PROSPECTS OF THE LIFE COURSE PARADIGM
Historical demography seems to have taken a
decisive “micro-turn” by focusing more and
more on life courses as a central concept.
According to its advocates, the “life course
approach” represents nothing less than a new
paradigm for (historical) demography. This
article evaluates this claim, by looking at
research achievements and by discussing
potential obstacles to fulfilling the promises of
this new approach.
The adoption of the perspective of the life
course has already resulted in many innovative
contributions, in particular on topics disregarded
by “traditional” historical demography
(e.g. the study of orphanhood, life-cycle
service, comparative studies on family dynamics).
Also, research fields such as ageing,
migration and fertility have benefited from
the collection and analysis of life course data.
Notwithstanding these advances, the further
development of the life course approach
may be obstructed for a number of reasons.
These have to do with the concept itself,
with the statistical methods used for
studying life courses, with the common
research practices in our discipline, and,
finally, with the “attractiveness” of our field.
The concept of “agency”—the capacity for
planning one’s own life—is clearly difficult
to apply in historical research. However, not
knowing individual people’s motives and
goals should not refrain us from hypothesizing
about the most likely goals of people in
specific places and periods. The method of
event history analysis still falls short of elucidating
life courses as processes. More elaborate
models should reveal the impact of past
experiences, of behavioural alternatives and
of interactions among processes and
contexts. However, the learning curve for
historical demography threatens to become
too steep, in particular when highly advanced
techniques need to be mastered. More
(inter)disciplinary team work might solve
this problem. Also, the life course approach
demands rich and complicated datasets.
This can hardly be accomplished by ordinary
faculty-based research teams, let alone
by individual researchers. Research (and
training) practices in this field will have to
shift towards data-sharing within collaboratories
and towards exploiting public use
datasets. Finally, to attract scholars from
other fields, research needs to be focused on
policy-relevant topics such as population
ageing, below replacements fertility, migration
and ethnic integration.
SUMMARY
RÉSUMÉ
La démographie historique semble avoir pris
un tournant « micro » en se concentrant de
plus en plus sur l’analyse des parcours de vie.
Si l’on suit ses avocats, il ne s’agit de rien
moins que d’un nouveau paradigme de la
démographie (historique). Cet article propose
une vue critique de cette affirmation, en
examinant les succès en termes de recherche et
en discutant les obstacles potentiels à la réussite
des promesses de cette nouvelle approche.
L’adoption de l’analyse des parcours de vie a
déjà fourni des contributions très neuves par
rapport aux sujets ignorés de la démographie
historique traditionnelle (les orphelins, le
service domestique dans le cycle de vie, les
études comparées sur les dynamiques familiales).
Des sujets comme le vieillissement,
les migrations et la fécondité ont également
bénéficié de la collecte et de l’analyse des
données du parcours de vie.
En dépit de ces avancées, le développement
de cette approche peut être compromis par
plusieurs facteurs : le concept lui-même, les
méthodes statistiques utilisées et les
pratiques de recherches dominantes ainsi
que l’attraction de notre champ. Tout
d’abord l’affirmation de la capacité d’organiser
sa propre vie est d’application difficile
dans les recherches historiques. Néanmoins,
l’ignorance des motifs et des buts des individus
ne devrait pas dissuader d’identifier
quels étaient les plus probables dans certains
lieux et certaines périodes. Ensuite, la
méthode de l’« event history analysis » ne
09-kok.qxp 16/07/2007 16:24 Page 229
230
JAN KOK
permet pas de clarifier les parcours de vie en
tant que processus. Des modèles plus élaborés
devraient permettre de préciser l’impact
des expériences passées, des conduites alternatives,
des interactions entre les processus
et les contextes. Enfin, le coût d’apprentissage
de la démographie historique risque de
devenir trop élevé, en particulier si des techniques
encore plus poussées doivent être
utilisées. Des équipes interdisciplinaires
pourraient résoudre cette difficulté. Par
ailleurs, l’approche du parcours de la vie
suppose des bases de données compliquées
et très riches. Tout cela ne peut être accompli
par les équipes universitaires traditionnelles
et les chercheurs solitaires. Dans ce
domaine, les pratiques de recherche (et d’apprentissage)
doivent s’orienter vers le partage
de données, la collaboration et l’exploitation
de banques de données publiques. Finalement,
afin d’attirer les chercheurs d’autres
domaines, la recherche doit privilégier des
questions à fort enjeu politique aujourd’hui
comme le niveau de la fécondité (enjeu du
remplacement), les migrations et l’intégration
des populations étrangères.
09-kok.qxp 16/07/2007 16:24 Page 230