The Leadership Quarterly 30 (2019) 298–319

Contents lists available at ScienceDirect

The Leadership Quarterly

journal homepage: www.elsevier.com/locate/leaqua

From classroom to conscription. Leadership emergence in childhood and T

early adulthood☆
Therese Reitana, , Sten-Åke Stenbergb
⁎

a
Academy of Public Administration, Södertörn University, S-141 89 Huddinge, Sweden
b
Swedish Institute for Social Research, Stockholm University, S-106 91 Stockholm, Sweden

ARTICLE INFO ABSTRACT

Keywords: Research on leadership emergence has mainly focused on adulthood and relied on retrospective accounts of
Leadership assessment childhood factors. Based on a prospective cohort study of 7719 boys born in 1953, of which 5928 were later
Childhood drafted, we explored individual differences in leadership emergence in childhood and early adulthood. The data
Conscripts set consisted of register data from different time points and a survey of the cohort in the 6th grade. As expected,
Cohort study
cognitive ability, tallness and muscular power were decisive for assessment of military officer suitability.
Cragg-hurdle regression
However, we also found a moderate to strong impact on this assessment score from social class, leadership
aspirations in childhood, birth order, self-regulatory skills in school, parental support, and previous participation
in extracurricular activities. Similarly, social class, cognitive ability, developmental experiences, and birth order
were important for nominations as class party organizer in the 6th grade. Delay of gratification was not asso-
ciated with these nominations at all and was neutralized by cognitive ability in relation to officer suitability. The
results strongly support a life-cycle approach to leadership emergence.

Jack spoke. “We've got to decide about being rescued.” There was a
buzz. One of the small boys, Henry, said that he wanted to go home.
“Shut up,” said Ralph absently. He lifted the conch. “Seems to me
we ought to have a chief decide things.”
“A chief! A chief.”
“I ought to be chief,” said Jack with simple arrogance, “because I'm
chapter chorister and head boy. I can sing C sharp.” Another buzz.
“Well then,” said Jack, “I- -” He hesitated.
The dark boy, Roger, stirred at last and spoke up. “Let's have a
vote.”
“Yes!!”
“Vote for chief!”
“Let's vote- -”
The toy of voting was almost as pleasing as the conch. Jack started
to protest but the clamor changed from the general wish for a chief to
an election by acclaim of Ralph himself. None of the boys could have
found good reason for this; what intelligence had been shown was
traceable to Piggy while the most obvious leader was Jack. But there
was a stillness about Ralph as he sat that marked him out: there was his
size, and attractive appearance, and most obscurely, yet most power-
fully, there was the conch.
(William Golding: Lord of the Flies, chapter one)
Introduction
William Golding's legendary novel Lord of the Flies from 1954 dra-
matically portrays human behavior as it unfolds among an unspecified
number of boys approximately 6–12 years old who find themselves in
an extreme, life-threatening situation after their plane has crashed.
With no adults around, one of their first challenges is to identify a
leader and the novel becomes a story of group dynamics, leadership
emergence and practice in a critical situation. Golding mercilessly re-
veals the truly dynamic character of leadership, where individual
characteristics, power resources, and situational factors are put to play
in complex interactions between potential leaders and followers
(Zacharatos, Barling, & Kelloway, 2000). Furthermore, the novel de-
monstrates how leadership is practiced already early in life (French &
Stright, 1991) and Golding thereby raises several fundamental issues
with relevance for leadership studies: What makes a leader? Do some
individuals simply have the right attributes (being tall and attractive or
able to sing C sharp) or is leadership something that comes from
practice (as head of the choir)? Do leaders emerge through social in-
terplay whereby someone is nominated by peers? Are situational factors
essential for leadership emergence (e.g. who is holding the conch – the
symbol of power – or what leadership qualifications are in demand on a

☆
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
⁎
Corresponding author.
E-mail addresses: therese.reitan@sh.se (T. Reitan), stenake@sofi.su.se (S.-Å. Stenberg).

https://doi.org/10.1016/j.leaqua.2018.11.006
Received 3 December 2017; Received in revised form 15 November 2018; Accepted 22 November 2018
Available online 07 December 2018
1048-9843/ © 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T. Reitan, S.-Å. Stenberg
desert island)? Moreover, are leadership positions and attributes in
childhood sustainable through to early adulthood? If Ralph and Jack
were to be drafted later in life, (they were actually saved by a naval
officer) would they be considered “officer material”?
These questions originate from a) a deep-seated tradition of distin-
guishing between individual and environmental factors in relation to
leadership emergence and development, b) a less prominent interest in
the influence of childhood patterns in adulthood, the continuity of
specific factors, and the interplay between them across the life course.
The dichotomy between individual and environmental factors has
proven to be very persistent in leadership studies. Although the notion
of a zero-sum battle is increasingly difficult to uphold, the distinction
seems to be continuously useful in sorting causes “outside the body”
from those “inside the body” (Duster, 2006). That factors, inside or
outside the body, are mutually exclusive is, however, not a viable
presumption (Jencks, 1980). Rather, we need to understand how per-
sonal, interpersonal and environmental factors interact under different
temporal and contextual circumstances.
Leadership studies have not denied or ignored the impact of child-
hood and adolescence for leadership emergence in adulthood.
Numerous references to the biographical history of individual leaders
are telling examples (Cunha, Lewis, Rego, & Smith, 2017; Shoup, 2005).
Understanding leadership among children and youths has been central
to psychology, sociology, and pedagogy albeit sometimes under other
headings, e.g. “social interaction”, “socio-metric status”, “deviant peer
influences”, “interpersonal development” and so forth (Belsky,
Steinberg, & Draper, 1991; Dodge, Dishion, & Lansford, 2006; Gest,
Graham-Bermann, & Hartup, 2002; Lerner & Steinberg, 2009). Still,
life-cycle perspectives have not been nearly as prominent in leadership
studies as for example the relative importance of personal and inter-
personal influences on leadership emergence, or the nature and con-
sequences of transformational/charismatic leadership styles. The em-
phasis in leadership studies has also concentrated on conditions and
events in adulthood (Murphy, 2011).
Despite the seemingly uncontroversial notion that childhood and
adolescence are crucial for adult leadership emergence and practice,
there is a scarcity of empirical research validating this perception (Day,
2011; Gottfried et al., 2011). This shortage is not only an issue of
quantity but also of methodological weaknesses. Single case- or small
sample-studies dominate, often based on retrospective reports of
childhood experiences, and selected populations such as students in
leadership programs or military cadets. Prospective studies of leader-
ship development have similarly been oriented towards adults within
organizations, not on developmental roots from childhood (Hannah,
Avolio, Luthans, & Harms, 2008; Strang & Kuhnert, 2009).
The nature of leadership development is inherently multilevel and
longitudinal. Whether we are dealing with formal or informal leader-
ship, within formal or informal organizations, individuals are nested
within teams, which in turn are nested within larger fields and struc-
tures (Day, 2011; Day & Harrison, 2007). Moreover, an individual's
journey into leadership roles (whether formal or informal) and the how
leadership practice develops is a matter of how intrapersonal and in-
terpersonal processes evolve over time and across the life cycle (Day,
Fleenor, Atwater, Sturm, & McKee, 2014; Murphy & Johnson, 2011).
Time is often neglected or poorly analyzed in leadership studies
(Castillo & Trinh, 2018).
Adolescence and early adulthood are periods in which the socio-
economic and ideological environment is thought to be particularly
important for foundations of adulthood (career choices, family life,
lifestyle, political formation). The impact of various factors on leader-
ship development is therefore bound to vary depending on the timing in
a person's life and the duration, magnitude, and periodicity of impact
(Castillo & Trinh, 2018). Rauthmann, Sherman, and Funder (2015)
refer to five categories of “cues”, i.e. elements in the environmental
structures which need to be interpreted and acted upon: 1) persons,
relationships, and social interactions, 2) objects, 3) events/activities, 4)
299
The Leadership Quarterly 30 (2019) 298–319
locations, and 5) time. It is, then, not sufficient to ask who, what, and
which – but also where and when. The “when?” question may, in turn,
refer to different aspects of time such as the magnitude, trends, dura-
tion, and periodicity of certain features/influences (Castillo & Trinh,
2018) – as well as when in history and when in a person's life course.
The aim of this study is, then, to analyze leadership development in
childhood and young adulthood from an individual, inter-personal and
contextual perspective. More specifically the study examines the impact
of individual differences in middle childhood on assessments of leadership
capacity in early adulthood.
Both of the stated objectives concern assessments of leader capacity,
not formal leadership role occupancy or leadership performance. More
specifically, the focus on developmental processes that promote, en-
courage, and assist in an individual's leadership potential in early life
and that may lead to a recognition of leadership potential. We use
leadership development and emergence interchangeably in order to
describe a process whereby individuals develop into leadership, and not
the development of leadership.
The study is based on the Project Metropolitan which included all
children born in 1953 and who were residing in the Stockholm me-
tropolitan area in 1963 (Stenberg & Vågerö, 2006). They were followed
from birth to their early 30s, but this study focuses on the period be-
tween middle childhood and early adulthood and only on the boys
(N = 7719). The latter is because the main outcome measure is as-
sessment of leadership capability in connection with conscription.
Apart from the study's longitudinal and prospective design, its major
advantages are the large sample size and the richness of data and data
sources – not least survey data from when the boys were aged 12–13.
Leadership in childhood and adolescence
The process in which a person becomes a leader, or exercises lea-
dership, is dynamic and iterative (Murphy & Johnson, 2011). How in-
dividuals develop a leader identity, seek arenas in which leader capa-
cities can be developed, and are recognized and supported as leaders by
peers and significant others is therefore fundamental. And, these pro-
cesses start from a very early age (Fukada, Fukada, & Hicks, 1994;
Mawson, 2011). A life course approach to leadership also implies that
events, learning experiences, peer recognition and so forth have varying
importance and impact depending on when in life they occur.
An investigation of the “seeds of leadership” (Murphy & Johnson,
2011) from childhood on will unavoidably touch base with sociological,
epidemiological and demographic life-course studies or developmental
psychology and the distinction between age, period, and cohort effects
(Glenn, 1976; Palmore, 1978; Yang & Land, 2016), although these are
often difficult to disentangle empirically. Empirical studies of leader-
ship in a life-course perspective, including childhood and adolescence,
is no less challenging from a methodological point of view. In addition,
we would argue that there are two distinct approaches to “leadership in
childhood and adolescence” that should be observed when tapping into
this field of research:
First, leadership in childhood and adolescence is often associated
with issues of anti-sociality, destructive leader-follower interaction, and
risk behavior. Fiction, theater and film are replete with illustrations of
the harsh and sometimes fatal consequences of (negative) social inter-
action among children and youths. Research studies in this area will
typically focus on deterring negative leadership and destructive group
processes, as well as studies of long-term “consequences of popularity”
(Allen, Porter, McFarland, Marsh, & McElhaney, 2005). Consequently,
research has often been driven by an explicit aim to promote pro-social
behavior, improve school environments, or enhance cognitive and so-
cial development among youngsters (Hoyt & Kennedy, 2008; Klepp,
Halper, & Perry, 2018; McCullough, Ashbridge, & Pegg, 1994;
Whitehead, 2009).
Second, leadership in pre-adulthood is linked to a search for the
roots of leadership emergence and leadership practice as such. That is,
T. Reitan, S.-Å. Stenberg
how to develop the right type of leaders and not primarily how to foster
“better citizens”. These studies often have adolescence or early adult-
hood as the starting point – typically departing from high-school, stu-
dent, or military recruit populations – where the outcomes usually focus
on leadership role occupancy, leadership practice and development,
occupational, participatory and career patterns in adulthood (Gayles &
Baker, 2015; Judge, Ilies, & Colbert, 2004; Kuhn & Weinberger, 2005;
McFarland & Thomas, 2006). This approach is, nonetheless, no devoid
of interventionistic ambitions. For example, Whitehead calls for an
improvement of leadership development programs in high school –
because “society cannot rely solely on sports or club activities to de-
velop young leaders” (Whitehead, 2009, p. 866).
Our study primarily falls into the latter category of approaches to
leadership in relation to childhood and adolescence. This study revolves
around the “roots” of leadership in adulthood, i.e. the assessment of
leadership capability in connection with conscription. However, we will
also apply a “within-childhood” approach to leadership by empirically
scrutinizing factors related to the choice of (informal) leaders among a
group of children.
Longitudinal studies of leadership development from early life
The shortage of investigations of leadership development from
childhood to adulthood does not only concern the mere quantity of
studies, but also problems related to insufficient data. Many studies
with an attempted longitudinal approach include data referring to two
points in time with only a short period in between (Li, Arvey, & Song,
2011). Studies with a time horizon of several years or longer are rare,
and cross-sectional designs dominate (Barling & Weatherhead, 2016;
Daly, Egan, & O'Reilly, 2015; Murphy, 2011). The lack of data does not
only concern limited points of observation. Given the interactional
character of leadership – there cannot be any leadership without fol-
lowers – a major weakness in leadership studies at large is the shortage
of relational and subjective data. That is, the “who” issues as suggested
by Rauthmann et al. (2015) and data pertaining and/or stemming from
peers, followers, parents, colleagues, subordinates, and so forth. Some
objective measures of conditions during childhood and adolescence can
be reconstructed retrospectively (e.g. socioeconomic status of the fa-
mily, educational attainment, participation in extracurricular activ-
ities), whereas subjective perceptions of relations and events cannot.
One of the main advantages of long-term studies starting from pre-
adulthood is that they comprise data on a population unselected for
specific criteria such as participation in leadership programs, leadership
role occupancy and so forth (Gottfried et al., 2011). Such research is
invaluable to avoid depending on data solely from populations who are
already selected into leadership contexts. The Fullerton Longitudinal
Study is one of few examples of prospective, long-term studies. It
started with 130 children and their families who were followed from
infancy to adulthood (Gottfried et al., 2011; Gottfried, Gottfried, &
Guerin, 2006; Reichard et al., 2011). The U.S. National Longitudinal
Survey of Youth, a large multi-wave and multisource data set, has also
been employed to study the both emergence of leaders and non‑leaders;
“understanding who does not emerge as a leader might point to per-
sonal or contextual factors that limit the pool of future leaders (Barling
& Weatherhead, 2016, p. 1305). The Project Metropolitan was not
specifically designed to study leadership issues, but it includes a broad
selection of data at the individual and inter-personal level with re-
levance to a life-cycle approach to leadership development.
An integrated model of leader individual differences in a life-
course perspective
The underlying theoretical framework of our study is adopted from
Zaccaro, Green, Dubrow, and Kolze (2018), emphasizing the additive
and pattern combinations of individual leader differences from child-
hood on. As shown in Fig. 1, the model includes two main categories of
300
The Leadership Quarterly 30 (2019) 298–319
leader individual differences; foundational traits and leadership capa-
cities. Foundational traits are relatively stable over time and situations,
i.e. “distal” traits. They are defined as individual attributes that orient a
person towards seeking leadership roles and predispose him or her to-
wards behavioral strategies likely to be successful in that role (p 6).
Some of these traits and leadership capacities have demonstrated sig-
nificant heritability estimates in empirical studies. These traits are also
causal precursors to the development and expression of more specific
leadership capacities, including e.g. cognitive and problem-solving
skills, social skills, motivational orientation towards leadership and
leader self-efficacy. Leadership capacities are relatively mutable, or
“proximal”, in that they can change and grow as a function of particular
developmental activities and experiences. It is important to note, then,
that foundational traits not only predispose individuals towards lea-
dership roles per se, but also orient them towards particular develop-
mental activities and increase their readiness for such situations (p 7).
As the authors point out, the suggested theoretical model can be
deconstructed into simpler models for empirical testing without losing
its overall conceptual integrity (Zaccaro et al., 2018, p. 35). It is
therefore important to point out the following: First, we have marked
“Genetic predisposition” with a dotted line in Fig. 1. This is not to
suggest that genetic factors are not of interest or importance – the role
of genes in relation to leader individual differences and outcomes has
received increased attention in recent years (Ilies, Arvey, & Bouchard,
2006; Tuncdogan, Acar, & Stam, 2017) – but we do not have such data
and genetic factors are not included in this empirical study. Second, we
do not study the actual execution of leadership in different situations.
So, leadership capacity in this study is reduced to a question of assess-
ment of leadership capacity by others. Third, time and context are not
specifically addressed in the original Zaccaro et al. framework but are
of course implicit in the distinction between mutable and stable traits,
and the definition of foundational traits as being relatively stable over
time and situations (Zaccaro et al., 2018, p. 6).1 The framework also
incorporates biological maturation and age-normative socialization
(early childhood factors) but it lacks a manifest link to life-course stu-
dies and life span models of leader development (Murphy & Johnson,
2011). To highlight this issue we have added “historical and societal
context” to the theoretical framework. These are not variables in the
proper sense and are not directly connected to the other categories,2 but
including this dimension emphasizes the importance of addressing the
“when?” question (Rauthmann et al., 2015) – or the cohort and period
effects, as it were (Palmore, 1978). Contextual knowledge is important
in relation to interpretation and generalizability. This particular study
of leadership development, for example, concerns an interval of
5–8 years from childhood to early adulthood among a cohort growing
up in an expanding welfare state at a time when in principle all males
were drafted, in a country not facing any immediate threat of war. Had
the country been at risk of war the escalating need for army officers
could potentially have influenced the assessment of leadership capacity.
1
Whether an attribute is stable or mutable naturally depends on when and
how it is measured. Socioeconomic status of a person's family of origin is a
foundational trait if measured retrospectively (Li et al., 2011), but may still
change if measured prospectively. Even a seemingly given attribute such as
birth order may change throughout the life course, for example in connection
with family restructuring involving children from different relationships
(Rosenberg & Hajal, 1985). Moreover, establishing degree of stability is not
only an issue of measurement but also determining how much change is ac-
ceptable before the attribute is no longer “stable” (Castillo & Trinh, 2018).
2
It is, however, not hard to find evidence that historical context(s) may have
a direct impact on other categories – even genetic predisposition (Stanner &
Yudkin, 2001). Societal expectations are also bound to differ depending on the
historical circumstances. Elder (1974) argues that children who grew up during
the Great Depression did better in school, were more likely to enter higher
education, and were generally more satisfied with life – largely because they
had to learn to handle challenges and difficulties at an early age.
T. Reitan, S.-Å. Stenberg
Leadership Learning Capacities and
Skills
Early Life Experiences
Family Academic intrinsic motivation
socioeconomic status
Birth order
Parental support
Foundational traits
Cognitive ability
Self-regulatory skills
Genetic
Motivation to lead
Predisposition
Physical characteristics
Historical and societal context
Fig. 1. A framework for leadership emergence in childhood and early adulthood, adapted from Zaccaro et al. (2018).
In the following, the main categories and variables in the model are
presented in further detail.
Early life experiences
Family socioeconomic status. Family socioeconomic status has not been a
prominent factor in studies of leadership emergence (Martin, Innis, &
Ward, 2017), but already 70 years ago, Stogdill identified > 100 data-
driven articles in which the most consistent correlates of leadership
emergence were SES, various indicators of intelligence, in addition to
personality traits, social skills and emotional control (Stogdill, 1948).
Contrary to Elder's (1974) aforementioned presumption about the
potential benefits of hardships in childhood on leadership emergence
later in life, early exposure to poverty is generally assumed to hinder
later leadership role occupancy by e.g. reducing access to relevant
education and development of skills and personal mastery (Barling &
Weatherhead, 2016). The economic and social environments that
individuals experience during their formative years has also been
found to have an impact on the magnitude of genetic influences on
leadership later in life (Zhang, Ilies, & Arvey, 2009). Based on a study of
twins they found that heritable traits were more strongly associated
with leadership role occupations for individuals from lower social
classes compared to those from higher classes. In short, individuals with
lower access to privileges and opportunities provided by a higher SES
background have to “rely” on heritable traits to a larger degree. The
association is not simplistic though; the impact of childhood
deprivation on leadership emergence has been found to differ
between men and women (Barling & Weatherhead, 2016). In fact, Li
et al. (2011) found that (higher) family socioeconomic status exerted an
adverse effect on leadership advancement for women.
Birth order. Birth order studies have attempted to empirically test the
common notion that the first born often takes on a leadership role, not
only in relation to his/her siblings but also among peers and in later
organizational and work life. There is substantial empirical evidence
that birth order is related to leadership in adulthood (Andeweg & Van
den Berg, 2003; Eckstein et al., 2010). First-born children, including
singletons, are overrepresented among political leaders in all parts of
301
The Leadership Quarterly 30 (2019) 298–319
Developmental Experiences
Membership in sports
club/scouts
Frequency of
extracurricular
activities
Leadership Capacities
Peer ascription of
leadership in childhood
Assessment of
leadership capability at
conscription
the world (Andeweg & Van den Berg, 2003). The mechanisms behind
this pattern have not been fully explored, but the main arguments
usually revolve around child-parental relations or sibling relations.
Some argue that first-born receive more attention, resources and are
subject to higher expectations from their parents. Others maintain the
importance of power struggles between siblings and first-born
children's experience as both leaders and followers (Andeweg & Van
den Berg, 2003). Birth order also figures as a control variable in
leadership studies (McCullough et al., 1994).
Parental support.
Ralph paddled backwards down the slope, immersed his mouth and
blew a jet of water into the air. Then he lifted his chin and spoke.
‘I could swim when I was five. Daddy taught me. He's a commander
in the Navy. When he gets leave he'll come and rescue us. What's
your father?’
Piggy flushed suddenly.
‘My dad's dead,’ he said quickly, ‘and my mum – ’
For children and adolescents, the family environment is critical for
their development and well-being on many levels. According to Zhang
et al. (2009), family environments such as socioeconomic resources and
parental support are more mutable and influential than other aspects of
adolescents' social environments, such as school, peers, neighborhoods.
However, from the adolescent's point of view they are still relatively
non-mutable and factors that the individual are subject to more than
being active participants in. Parents may support their offspring emo-
tionally by promoting self-confidence and achievement orientation, by
acting as role models, by encouraging the engagement in certain ac-
tivities (Brungardt, 1997). Parents may of course also play a role in
introducing their children to networks and arenas that are otherwise
not equally accessible.
Leadership learning capacities and skills
Academic intrinsic motivation. Intrinsic motivation includes enjoyment
of an activity as such without the receipt or promise of external
rewards. Academic intrinsic motivation refers to the motivation for
learning or doing schoolwork out of curiosity or the pleasure of
T. Reitan, S.-Å. Stenberg
learning, not primarily because one has to (Gottfried et al., 2011).
Academic intrinsic motivation is, in turn, important to consider in
relation to leadership. It includes attributes that are likely to contribute
to leadership, such as an orientation towards mastery, curiosity, per-
sistence, and engaging in new and challenging tasks. Based on data
from the Fullerton Longitudinal Study, Gottfried et al. found that aca-
demic intrinsic motivation in childhood and adolescence provided a
foundation for motivation to lead in adulthood. From childhood to
adolescence, individuals with higher academic intrinsic motivation
were more likely to want to lead because of enjoyment and without
regard of rewards or costs and are therefore less likely to avoid lea-
dership roles based on such calculations (Gottfried et al., 2011).
Foundational traits
Cognitive ability. Childhood intelligence or cognitive ability have been
shown to predict a wide array of short- and long-term outcomes. The
more obvious are related to school performance, educational transition,
in addition to occupational and income status (Fergusson, Horwood, &
Ridder, 2005). But, even long-term effects of childhood IQ on survival
to old age have been uncovered (Whalley & Deary, 2001).
Leadership tasks are generally cognitively demanding in terms of
performance requirements and higher intelligence increases the like-
lihood of engaging in such tasks (Zaccaro et al., 2018). In the same way
as cognitive ability in adulthood is likely to influence leadership
emergence and leadership development, results from a large study of
two representative British cohorts suggest that childhood cognitive
ability may also profoundly shape leadership trajectories in later life
(Daly et al., 2015).
Questions have, nonetheless, been raised about the validity of IQ in
relation to both leadership and other developmental or social outcomes.
Based on the Fullerton Longitudinal Study, Guerin et al. (2011) found
that adolescent IQ was not associated with either social skill or lea-
dership potential in adulthood. Instead, they propose the inclusion of
early individual differences in temperament and/or personality in
conceptual frameworks of leader development, or alternative con-
ceptualizations of cognitive ability. Moreover, the impact of cognitive
abilities is highly related to situational factors (Fiedler, 2002).
Be that as it may, there are good reasons for including cognitive
ability in childhood in this model. It is associated with leader role oc-
cupancy in adulthood, not least through educational attainment (Daly
et al., 2015). Cognitive ability may also contribute positively to mental
and physical health, which could facilitate leader role emergence for
example through access to developmental experiences conducive to
leadership emergence (Daly et al., 2015; Deary, Weiss, & Batty, 2010).
Research on military cadets showed that cognitive ability, physical
fitness, self-esteem, and prior experience associated with leader emer-
gence (Day & Sin 2011). Although cognitive ability in itself cannot
predict leadership role occupancy (Li et al., 2011), there are good
reasons to believe cognitive ability contributes to leadership emergence
through different pathways in childhood and between childhood and
early adulthood. The association between cognitive ability and lea-
dership is, nonetheless, complicated by the fact that individuals higher
in intelligence are more likely to be perceived as leaders by followers
(Lord, De Vader, & Alliger, 1986). However, the value of intelligence in
terms of social status among peers could be an intricate act of balance in
childhood when group identification and ordinal rank can be highly
important for expectations and behavior (Elsner & Isphording, 2018).
There is evidence suggesting that follower perceptions of leadership
qualities follow a curvilinear function of intelligence. Too wide dis-
crepancies between leaders and followers may for example jeopardize
the perception of unity and common interests in a group. There are
therefore good reasons to test non-linear functions of IQ (Antonakis,
House, & Simonton, 2017; Simonton, 1985).
Self-regulatory skills. Whether a form of cognitive capacity – as
suggested in the overview of major leadership trait reviews and meta-
302
The Leadership Quarterly 30 (2019) 298–319
analyses by Zaccaro et al. (2018) – or “simply” a trait-like capacity the
ability to exert self-control not only varies between individuals, but
seems to do so already at an early age (Tuncdogan et al., 2017). More
importantly, this capacity has been associated with a number of
outcomes later in life. The large body of research spurred by the
“marshmallow tests” of Walter Mischel has aimed to establish a link
between the ability to delay gratification in childhood with e.g.
educational outcomes, health, criminality, substance abuse and so
forth later in life (Tuncdogan et al., 2017).
Individual differences in e.g. cognitive capacities, personality and
temperament, and ability to regulate emotional responses have been
identified as crucial in the acquisition and improvement of leadership
skills (Lord & Hall, 2005), not least in a lifespan approach to leader
development (Murphy, 2011). Measures of self-regulation have also
been found to vary between age groups, partly related to the fact that
cognitive abilities develop in stages (Gollner, Ballhausen, Kliegel, &
Forstmeier, 2017).
Self-control, defined as “impulsive acting out without regard for
consequences” and rated by teachers when the child, was included as a
covariate in a British longitudinal cohort study of leadership role oc-
cupancy (Daly et al., 2015). Overall, rule-breaking or norm-challenging
behavior in childhood and adolescence are strong warning signs of
continued social deviancy and poor outcomes in many areas (Zhang,
Welte, & Wieczorek, 2002). However, others have argued that not all
problem behaviors are detrimental to educational achievement or life
prospects. Firstly, a large majority of adolescents engage in some form
of rule-breaking behavior at some point although very few continue
into an antisocial life-path (Bandura, 1986; Moffitt, 2015). Secondly, if
one accepts that leadership and entrepreneurship share many simila-
rities (Reid, Anglin, Baur, Short, & Buckley, 2018) and that en-
trepreneurs almost by definition are rule-breakers (Zhang et al., 2009),
modestly rebellious behavior could even be beneficial if it increases
self-knowledge and helps the individual explore the limits and con-
sequences of norm-breaking behavior. Obschonka (2016), for example,
distinguishes between breaking rules and breaking the law, pointing
out how adolescent rule breaking can be a valuable asset for future
entrepreneurship.
Motivation to lead. Motivation is a broad category of individual traits
that are more or less mutable while also susceptible to environmental
influence. Motivation is partly a general ability that children develop as
they mature (Baumeister & Vohs, 2003) or specifically directed towards
becoming the person they want to be (Banaji & Prentice, 1994). Mastery
motivation in young children has been defined as a psychological force
that stimulates an individual to attempt to independently in a focused
and persistent manner, to solve a problem or master a skill or task
“which is at least moderately challenging for him or her” (Morgan,
Harmon, & Maslin-Cole, 1990, p. 319). The latter qualifying term is an
attempt to separate motivation from general cognitive ability.
When it comes to leadership motivation specifically, Lord and Hall
(2005) argue that leader identity is essential for leader skill develop-
ment. To emerge as a leader in the first place, though, is often the result
of a long and dynamic process by which an individual becomes aware
that leaders exist and his/her own aspirations of becoming a leader.
Already in early childhood we learn to recognize that certain people
have a special authority (parents, teachers, and so on), without perhaps
being defined as leaders. As children grow older, some of them also
start experimenting or engaging in leadership activities through extra-
curricular activities (Day, Harrison, & Halpin, 2008) and learning self-
regulation (Murphy, 2002). Developing preferences for certain occu-
pations starts at an early age and advances in stages as children acquire
a better understanding of themselves and the world of work. Children's
occupational preferences are, then, likely to reflect the awareness of
specific occupations and their content, but also a perception of occu-
pational characteristics such as degree of leadership (Stockard &
McGee, 1990).
T. Reitan, S.-Å. Stenberg
Considerable research has supported the relative permanence of
childhood occupational preferences and the concordance between ca-
reer aspirations of youngsters and their actual careers (Wahl &
Blackhurst, 2000). A follow-up study of two British cohorts revealed
that teenage aspirations were good predictors of adult occupational
attainment. Teenagers with high aspirations were more likely to enter a
professional or managerial career compared to their less ambitious
peers. The socio-historical context is, nonetheless, extremely important:
Due to changes in social, economic, and education policies, the two
cohorts – born only 12 years apart – grew up under very different cir-
cumstances. Increasing unemployment, loss of manufacturing jobs, and
technological transformation meant that the younger cohort were under
much greater pressure to continue full-time education and to acquire
formal qualifications and steer away from certain occupations (Schoon
& Parsons, 2002).
Physical characteristics. In his seminal overview of leadership research
Stogdill (1948) noted that athletic ability and “physical prowess”
appeared to be associated with leadership status in boys' gangs and
groups. Piggy in The Lord of the Flies is an intelligent and rational boy
who in that respect would be a potential leadership candidate.
However, he has serious flaws in other areas; apart from being
cowardly, immature, and sensitive, he is overweight, has poor
eyesight and asthma. In contrast, Ralph's somewhat inferior cognitive
ability is compensated for by his tall and athletic approach. He is
described as a potential boxer “as far as width and heaviness of
shoulders [goes]” – his brawns are clearly an important element in
his leadership emergence. The question has continued to intrigue
researchers and there is some support for the notion that tall men are
more likely to attain leadership positions compared to shorter men
(Judge & Cable, 2004; Lindqvist, 2011), although there is a probable
bias in reporting (Elgar, 2016). The possible association has been linked
to psychological research suggesting that tall people are seen as more
persuasive, impressive and capable (Stulp, Buunk, Verhulst, & Pollet,
2013). Interestingly enough there also seems to be evidence of a
reciprocal connection in the sense that people in prestigious positions
are perceived to be taller than they are (Lindqvist, 2011). Lundborg,
Nystedt, and Rooth (2009) found that the well-established “height
premium” in wages in fact may be explained by physical capacity,
cognitive, and non-cognitive skills. However, Blaker et al. (2013) found
a gender difference in this respect; although part of the height premium
may be attributed to increased cognitive capacity tall males still seem to
have an added advantage of being perceived as more dominant and
healthy compared to women.
The role of physical characteristics as foundational leader traits has
seen a resurgence in recent years, linking “new” scientific disciplines
such as neuroscience, behavioral genetics, and human biology to lea-
dership studies (Antonakis, Day, & Schyns, 2012; Arvey, Rotundo,
Johnson, Zhang, & McGue, 2006; Chaturvedi, Zyphur, Arvey, Avolio, &
Larsson, 2012; Dinh & Lord, 2012). Studies have for example shown
significant associations between facial appearance and different aspects
of leadership (Little, 2014).
Developmental experiences
In the campaign between leadership candidates in The Lord of the
Flies Jack invokes his experience as “chapter chorister and head boy”. It
might not have been the most impressive merit for his potential fol-
lowers, but Jack was clearly on to something. As children start school
and grow a bit older, they begin to observe and recognize effective and
ineffective adult leaders (Ruhm, 1992). Participation in organized ac-
tivities such as choirs, musical bands, theater groups, scouts, and sports
clubs, may provide invaluable opportunities for such observations, ex-
posure to role models, learning and practicing leadership skills (Dobosz
& Beaty, 1999; Karagianni & Jude Montgomery, 2018; Murphy, 2011;
Murphy & Johnson, 2011; Ruhm, 1992). The suggested framework
suggests that early development experiences can actualize genetic
303
The Leadership Quarterly 30 (2019) 298–319
predispositions (such experiences can also alter genetic structures), and
facilitate the contribution of foundational traits to particular leadership
capacities. Little empirical work has, however, been done on specific
developmental experiences in relation to different leadership capacities
(Murphy & Johnson, 2011; Zaccaro et al., 2018). Parental encourage-
ment and socioeconomic status are bound to influence the extent to
which children participate in sports and other activities, and not least
which activities (McNeal, 1998) – which is another reason to include
both in the theoretical framework.
Leadership capacity. As mentioned at the outset this study does not
analyze the actual performance of leadership, but rather the assessment
of leader capacity by others. Formal assessment and ascription of
leadership capability is a widely studied topic, not least in relation to
gender and leadership emergence (Rosette, Leonardelli, & Phillips,
2008; Sczesny & Kühnen, 2004; Solansky, 2010).
The main aim of this paper is to analyze the impact of factors in
childhood and early adulthood on the formal assessment of officer
suitability in connection with conscription. However, informal leader-
ship ascription is essential in understanding group behavior and the
self-reinforcing processes leading towards leadership role occupancy
(Paunova, 2015). At no time in life is this more important than in
childhood and adolescence. Children in middle childhood spend a great
deal of time trying to learn the social rules and self-presentational
strategies necessary for inclusion – and are acutely aware of pecking
orders and their precariousness (Lease, Musgrove, & Axelrod, 2002).
Observational studies of group dynamics and socio-metric measures
have been used for decades to investigate social isolation, status and
general acceptance among children (Y. Almquist, 2009, 2011; Horrocks
& Thompson, 1946; Lease et al., 2002; V. Östberg, 2003; V. Östberg &
Modin, 2008). The school survey also includes a socio-metric measure
whereby pupils were asked to anonymously nominate classmates as
potential class party organizers, i.e. as a form of informal ascription of
leadership. Because of an endogeneity problem this measure cannot be
included in the main analytical model, but it is analyzed as a dependent
variable using most of the other variables in the main model.
Method
Participants, design, and procedures
The sample and data for this study stem from the Project
Metropolitan, which was primarily a sociological study of social mo-
bility, conformity and deviance in a life-course perspective. It was
originally planned as a joint effort between Stockholm, Oslo,
Copenhagen, and Helsinki, but ended up including only Stockholm and
Copenhagen (Stenberg, 2018). The Swedish project covered all children
who were born in 1953 and who were residing in the Stockholm Me-
tropolitan area in 1963 (totaling 7398 girls and 7719 boys). The data
were collected from the cohort members themselves, their guardians,
and from registries and covers the time period between 1953 and 1985
(Janson, 1966; Stenberg & Vågerö, 2006). One of the central sources of
information was the school survey performed in 1966, when the cohort
members were 12–13 years old. The survey was distributed by teachers
in the classroom and completed by the children themselves. The school
survey included questions about attitudes towards school and educa-
tion, plans for the future, occupational aspirations, hobbies and leisure
activities. It also included three different cognitive tests, and questions
about relations to their classmates and parental support and expecta-
tions. Even children who were not included in the study sample, but
who were present in the classroom were given a survey form. Their
responses were included in the project if these responses had bearing on
the study cohort members (e.g. nominations for class party organizer)
and they were also included in the information about class size. In-
formation was also collected from registers and censuses of population
and housing – e.g. family income and social class – and conscription
T. Reitan, S.-Å. Stenberg
data were collected for the boys.3
The focus of our study is the assessment of leadership capacity
measured as the assessment of military officer suitability in connection
with conscription. At that time (early 1970s) practically all Swedish
males were drafted for military service. The draft usually took place at
the age of 18 or 19, and consisted of interviews, psychological tests, and
medical examinations over a three-day period. The draft process served
several purposes. First, it was to ensure that all male citizens were
called in and fulfilled their societal duties. Only those who were already
serving in the military forces, seamen, and the severely handicapped
(physically or mentally) were exempt from appearing before the draft
board. Second, this was an arena for assembling and producing in-
formation. The draft boards collected data from other registers, and
produced new data through the examinations and tests performed
during the three-day assessment period. Apart from serving the needs of
the army, these data played an important role in policy documents and
the production of official statistics (National Board of Health and
Welfare, 2009; Public Committee for the Education of Conscripts,
1986). Furthermore, the availability and structure of this comprehen-
sive and comparable information for whole cohorts over a long time
period, came to provide an invaluable asset for research on for example
health, social mobility and labor market participation (Karnehed,
Rasmussen, Hemmingsson, & Tynelius, 2012; Lindqvist & Vestman,
2011; Lundin, Lundberg, Hallsten, Ottosson, & Hemmingsson, 2010;
Sidorchuk, Hemmingsson, Romelsjö, & Allebeck, 2012). The third
purpose was to process and use information for screening and selection
of military recruits. Recruits were deemed either fit or unfit for service,4
and those who were fit for service had to be matched with military
requirements and needs.
The first selection of recruits for assessment of potential officer
suitability was very cursory and primarily based on test results from the
draft (IQ, body height, muscular capacity, medical data) and/or
average school mark in the 9th grade, without a prolonged personal
encounter. Inductees who were selected to be assessed for military of-
ficer suitability5 were interviewed according to the guidelines from the
Central Conscription Authority. At the outset, the draft board had ac-
cess to basic demographic information, test data from the draft as well
as average school mark at the end of the 9th grade. The evaluation was
to be based on intellectual capacity (measured in various tests, school
marks, and other previous accomplishments), attitudes towards the
military, emotional stability, ability to be accepted by others, social
interaction skills, occupational ambitions, as well as physical fitness
(Central Conscription Authority, 1952). According to the guidelines,
intellectual capacity generally accounted for approximately 40% of the
variation in officer suitability. The recruit's motivation and ambition
was also mentioned as decisive factors, and evaluators were advised not
to let intellectual capacity influence the judgment of other factors too
greatly (personality, stability, social skills, etc.). Evaluating the recruit's
emotional stability was part of the purpose of the interview and was an
element in the general assessment of officer suitability It should be
mentioned that the average time allotted for each recruit was 12 min, so
the breadth and depth of topics was limited. The Central Conscription
Authority was responsible for filing draft data. Data for most of the
variables were provided to the Project Metropolitan.
This study is based on the 6928 draftees from the original cohort of
7719 boys.6 We are not able to specify the various grounds for attrition
but the loss of observations was not random given the strict criteria for
3
Codebooks and information about the Project are available at http://www.
stockholmbirthcohort.su.se/about-the-project/original-data-1953-1983.
4 7
In addition, the board could exempt others temporarily or indefinitely, and
allocate yet others to limited service or to serve as conscientious objectors.
5
Befälsskattning in Swedish.
6
The study has been approved by the Regional Ethical Review Board in
Stockholm.
304
The Leadership Quarterly 30 (2019) 298–319
exemption (e.g. mental and physical illness or handicap). Even pre-
mature deaths and emigration are most probably biased. We have
therefore, for our purposes, based our analyses on the individuals who
were drafted and represented in the Conscription Board's registry.
Informal ascription of leadership status among children is analyzed
socio-metrically with the use of a question from the school survey
asking them to nominate classmates for the role as class party orga-
nizer. The question is assumed to reflect both social status in the class
and general popularity, but also a perceived ability to lead a group and
get things done. Peer status ratings based on the school survey have
previously been used in studies of e.g. educational attainment and
health-related outcomes in adulthood (Y. Almquist, 2011; Y. B.
Almquist & Östberg, 2013; Almquist, Modin, & Östberg, 2010).
An overview of the variables in this study, the sources of informa-
tion, the year the variables refer to, and non-response rates is presented
in Table 1. The variables are presented in further detail in the following:
Dependent variables
In connection with the draft leadership capacity, or more precisely
military officer suitability, was assessed on a scale from 1 to 9, with 1
representing the lowest and 9 the highest values. As mentioned in the
last section the evaluators were advised not to let intellectual capacity
influence the judgment of other factors too greatly. In practice, how-
ever, inductees with an IQ of 4 or less were generally not selected for
assessment. Overall, 31% of the males in our study did not pass this first
hurdle, and only 15 individuals (of 5928) with an IQ of 4 or less were in
fact assessed for military officer suitability. The non-assessed in-
dividuals were assigned the value 0 on military officer suitability. In
Fig. 2 the left bar represents non-selected recruits, whereas the selectees
show a near normal distribution on the 1–9 scale.
In the analysis of factors related to informal ascription of leadership
status among children we use the question in the school survey the
children were asked to name up to three classmates who they would
like to see as a class party organizer. The distribution of number of peer
nominations is shown in Fig. 3. The largest category consisted of those
who got no votes, while the maximum number of nominations was 26.
Independent variables
Family socioeconomic status was measured by social class based on
the father's occupation in 1963.7 The information was originally col-
lected from population register data, and coded into five categories by
the original project group: Upper and upper middle class, Lower middle
class, officials, and non-agricultural employees, Lower middle class or
entrepreneurs, Working class (skilled workers), Working class (un-
skilled workers).
The population registry was also the source of information about
birth order and was calculated on the base of data on older and younger
siblings. Sibling status refers to the situation in 1963, when the cohort
members were 10–11 years old, and may of course have changed later if
for example an only child got a younger sibling after that age.
Parental support is an index constructed by the original project staff
and based on 10 questions in the school survey about the family's at-
titudes towards higher education (e.g. “Do your parents think that
having a higher education will give a more secure future?”).8
Academic intrinsic motivation was measured by an index based on 10
questions in the school survey about interest in schoolwork (e.g. whe-
ther they read other books than textbooks about things they studied at
If information about the father was missing, data concerning the mother was
used instead. This was also the case if the mother had an occupation in the
upper-middle class or higher and the father's occupation was working class. For
more information on how these groups were constructed, see Codebook II.
8
See variable SCHOL16 in Codebook I for a full list of questions.
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 1
Overview of variables, data sources, year information refers to and age of cohort members, and non-response rates.
Variable Database Source of information Year information refers to Non-response (%) (N = 6928)
(age of cohort members)

Outcomes
Military officer suitability National Conscription Board register Draft officer 1971–73 (18–20) 0
Nominations as class party organizer School survey Class mates 1966 (12−13) 16
Independent variables
Family socioeconomic status Population register Coded by original project 1963 (9–10) 15
Birth order Census 1964 Household members 1964 (10−11) 0
Parental support School survey Child 1966 (12–13) 10
Academic intrinsic motivation School survey Child 1966 (12–13) 10
IQ School survey Child 1966 (12–13) 13
Leader occupation aspiration School survey Child 1966 (12–13) 13
Delay of gratification School survey Child 1966 (12–13) 13
Expulsion from classroom School survey Child 1966 (12–13) 10
Membership in sports club/scouts School survey Child 1966 (12–13) 13
Frequency of extracurricular activity School survey Child 1966 (12–13) 11
Muscular capacity National Conscription Board register Draft officer 1971–73 (18–20) 5
Body height National Conscription Board register Draft officer 1971–73 (18–20) 5

Recruits with value 0 were not assessed

30
20
Percent
10
0

0 1 2 3 4 5 6 7 8 9
Assessment scores

Fig. 2. Assessment scores for military officer suitability (percent).

school, and whether they found homework dull).9
Lastly, we included the children's IQ in the 6th grade. The measure
is based on a spatial, verbal, and numerical sub-test as part of the school
survey. Each sub-test rendered a maximum score of 40 and, so, the
highest possible score overall was 120.10 The IQ scores in this study
group varied between 12 and 116, with a mean of 70. In order to test
the non-linear effect of IQ we also included the squared IQ score
(Simonton, 1985).
Motivation to be a leader was based on the pupils' answers to
questions about what they want to be when they grow up. The children
were presented with a list of 56 different occupations and for each
occupation the child was to indicate whether he/she “would like to
become” (1), “do not know” (2), or “would not like to become” (3) this
occupation. As a measure of motivation to lead we included two oc-
cupations which are closely associated with leadership functions;
manager11 and (military) officer. Leader occupation aspiration was coded
into three categories depending on how many of these occupations the
boy had chosen (both, one, or none).
Self-regulatory skills were measured in two ways: The first item is a
measure of capacity to suppress immediate gratification for the benefit
of long-term reward. In the school survey the children were presented
with a hypothetical question asking whether they would choose if they
had the prospect of receiving SEK 100 now or SEK 1000 in five years,
comparable to the test of self-control and delayed gratification in the
classical “Stanford Marshmallow experiment” (Mischel & Ebbesen,
1970). The second measure is based on a question in the school survey,
which may reflect an ability to regulate emotions and curb dis-
obedience, namely whether the pupil was ever told to leave the classroom
during a lesson because of something he did.

11
“Direktör” in the original Swedish version. This has been translated to
9
See variable SCHOL18 in the codebook for full list of questions. “manager” in Codebook I, but “director” is actually more appropriate because it
10
See Codebook I. has a clearer connotation to an executive or leadership role.

305
T. Reitan, S.-Å. Stenberg
30
20
Percent
10
0
0 5
Fig. 3. Nominations for class party organizer (percent).
The foundational traits in the theoretical framework also consist of
physical characteristics. Two measures from the draft procedures were
included as indicators of physical stature; body height (in centimeters)
and muscular power (in Newton, grouped on a scale 1–9).
Developmental experiences are important for the practice of lea-
dership skills and exposure to arenas and situations that may be helpful
for this purpose. In the school survey, the children were asked whether
they were members of different types of associations and clubs, and the
frequency of activity in such associations. They were allowed to provide
up to three different types of memberships and for our purpose, we
focused on membership in a sports clubs and/or the scouts. Both alter-
natives emphasize physical and mental training, teamwork, goal at-
tainment and competitiveness. The pupils were also asked about how
often they were busy with club or association activities, based on a
presumption that the “dosage” of developmental experiences also
matters. Frequency of extracurricular activities was coded into four ca-
tegories; not a member of an organized activity, never, sometimes,
weekly.
Means, standard deviations, and correlations are presented in
Table 2a and the distribution between categories for categorical vari-
ables are presented in Table 2b.
Procedure
When it comes to the assessment of military officer suitability, we
are dealing with a bounded distribution. Assessment of military officer
suitability is bounded by zero, i.e. the value assigned to draftees who
were not assessed at all. When the dependent variable is bounded a
hurdle regression is an appropriate statistical technique.
Hurdle models were first proposed by Cragg (1971). A hurdle model
resembles a tobit model, but the main difference is that a hurdle model
provides separate equations for the bounded and unbounded outcomes
whereas a tobit model uses the same equation for both. Hurdle models
assume that the unbounded outcomes are the result of passing an initial
“hurdle”. The model consists of two equations. In the first equation, we
estimate the chance of proceeding to the unbounded selection, i.e. of
passing the “hurdle” (values 0 or 1).12
Because IQ was a central criterion at this stage, we included IQ in
12
Estimated with the churdle (Cragg-Hurdle) command in Stata 15.1.
The Leadership Quarterly 30 (2019) 298–319
10 15 20 25
Number of votes
the first equation. However, we did not include the IQ score from the
draft, but rather IQ as measured in the school survey. The reasons are
threefold: First, there was – not unexpectedly – a substantial overlap
between IQ in the 6th grade and at the draft (r = 0.64) and as such
these measures are congruent but not fully overlapping. Second – and
more importantly – neither the draft board nor the inductee had any
knowledge of the IQ score from 6th grade and as such it was not
“contaminated” by the draft situation where inductees might have
wanted to manipulate results if they did not wish to be assessed for
leadership capacity or to serve in the military at all. Third, the IQ score
from the 6th grade was more widely dispersed, with scores varying
between 12 and 116, compared to a grouped score of 0–9 for IQ at the
draft. In the first equation, we also included body height and muscular
capacity, which were directly available from the draft board. Muscular
power was easier to manipulate than body height but are overall highly
correlated and were easily accessible and interpretable measures that
may have influenced the initial categorization of candidates for as-
sessment albeit not as clearly as in the case of IQ.
The second equation provides an estimate for the scores on the 1–9
scale among those who were selected for assessment. Because the
children in the original Project Metropolitan were grouped in 612
school classes in 1966 (the year the school survey was performed),
standard errors are probably not unbiased. Therefore, we also calcu-
lated clustered standard errors. Contrary to the first equation, all
variables were included in the second equation. Hence, the model in-
cludes information that was or might have been known by the draft
board and/or inductee at the outset (birth order, family socioeconomic
status), information which none of the parties had knowledge about or
remembered (e.g. parental support, leader occupation aspiration). The
model also includes information that may have emerged during the
assessment procedure (such as participation in sports).
In order to study the effect of factors relating to early life experi-
ences, leadership learning capacities and skills, foundational traits, and
developmental experiences in childhood on leadership emergence “in
the classroom”, we performed a negative binominal regression. This is
an appropriate method for over-dispersed count data, without an ex-
cessive number of zeros, where the conditional variance exceeds the
conditional mean (see Fig. 3). The dependent variable is number of
nominations the cohort member received from his classmates (anon-
ymously) for the potential role as class party organizer. The in-
dependent variables are all those related to childhood, i.e. all variables
306
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 2a
Descriptive statistics: Means, standard deviations, and (Pearson r) correlations (N = 6928).
Variable M SD 1 2 3 4 5 6 7 8 9 10 11 12 13

1. Military officer 3.72 3.05

suitability
2. Nominations for class 2.47 3.09 0.18⁎⁎⁎
party organizer

Family socio-economic status

3. Unskilled worker 0.17 0.37 −0.18⁎⁎⁎ −0.07⁎⁎⁎
4. Working class, skilled 0.23 0.42 −0.12⁎⁎⁎ −0.07⁎⁎⁎ −0.24⁎⁎⁎
workers
5. Lower middle class, 0.08 0.26 0.00 0.02 −0.13⁎⁎⁎ −0.15⁎⁎⁎
entre-preneurs
6. Lower middle class 0.36 0.48 0.07⁎⁎⁎ 0.04⁎⁎ −0.34⁎⁎⁎ −0.41⁎⁎⁎ −0.21⁎⁎⁎
7. Upper, upper-middle 0.17 0.37 0.22⁎⁎⁎ 0.09⁎⁎⁎ −0.20⁎⁎⁎ −0.24⁎⁎⁎ −0.13⁎⁎⁎ −0.34⁎⁎⁎
class

Birth order
8.Youngest child 0.30 0.46 −0.02 0.05⁎⁎⁎ −0.03⁎ −0.04⁎⁎ 0.01 0.01 0.05⁎⁎⁎
9. Only child 0.21 0.41 0.02 −0.01 0.01 0.03⁎ −0.01 0.04⁎⁎ −0.08⁎⁎⁎ −0.34⁎⁎⁎
10. First born 0.32 0.47 0.05⁎⁎⁎ −0.01 −0.00 0.00 −0.02 0.02 −0.02 −0.45⁎⁎⁎ −0.35⁎⁎⁎
11. Middle child 0.17 0.38 −0.05⁎⁎⁎ −0.04⁎⁎ 0.03⁎ 0.01 0.03⁎ −0.07⁎⁎⁎ 0.04⁎⁎ −0.30⁎⁎⁎ −0.23⁎⁎⁎ −0.31⁎⁎⁎
12. Parental support 6.10 2.26 0.36⁎⁎⁎ 0.17⁎⁎⁎ −0.19⁎⁎⁎ −0.13⁎⁎⁎ −0.00 0.10⁎⁎⁎ 0.21⁎⁎⁎ −0.00 0.04⁎⁎ 0.03⁎ −0.07⁎⁎⁎
13. Academic intrinsic 4.87 2.46 0.13⁎⁎⁎ 0.06⁎⁎⁎ −0.06⁎⁎⁎ −0.02 −0.03⁎ 0.03⁎ 0.06⁎⁎⁎ −0.06⁎⁎⁎ 0.05⁎⁎⁎ 0.05⁎⁎⁎ −0.04⁎⁎ 0.25⁎⁎⁎
motivation
14. IQ 69.47 18.09 0.54⁎⁎⁎ 0.20⁎⁎⁎ −0.19⁎⁎⁎ −0.13⁎⁎⁎ 0.01 0.09⁎⁎⁎ 0.21⁎⁎⁎ −0.02 0.03⁎ 0.04⁎⁎ −0.06⁎⁎⁎ 0.43⁎⁎⁎ 0.07⁎⁎⁎

Delay of gratification
15. Certainly SEK 100 0.06 0.24 −0.07⁎⁎⁎ −0.01 0.05⁎⁎⁎ 0.02 −0.04⁎⁎ 0.01 −0.05⁎⁎⁎ −0.00 −0.01 −0.01 0.03⁎ −0.08⁎⁎⁎ −0.05⁎⁎⁎
now
16. Probably SEK 100 0.05 0.22 −0.03⁎ −0.04⁎ 0.02 0.02 −0.02 −0.01 −0.00 0.00 0.00 0.00 −0.01 −0.03⁎⁎ −0.01
now
17. Cannot choose 0.06 0.25 −0.06 ⁎⁎⁎
−0.03 ⁎
−0.00 0.01 0.01 0.01 −0.03 ⁎
0.02 −0.01 −0.01 −0.01 −0.05⁎⁎⁎
0.00
18. Probably SEK 1000 0.33 0.47 0.02 0.00 0.00 0.00 −0.02 −0.00 0.01 0.01 0.00 −0.00 −0.02 0.01 −0.01
in five years
19. Certainly SEK 1000 0.49 0.50 0.06⁎⁎⁎ 0.03⁎ −0.03⁎ −0.03⁎ 0.04⁎⁎ 0.00 0.03⁎ −0.02 0.00 0.01 0.01 0.07⁎⁎⁎ 0.03⁎
in five years

Expulsion from classroom

20. Never expelled 0.26 0.44 0.03⁎ −0.02 −0.01 0.02 −0.01 −0.01 0.01 −0.05⁎⁎⁎ 0.03⁎ 0.04⁎⁎ −0.02 0.04⁎⁎ 0.18⁎⁎⁎
21. Once 0.54 0.50 0.03⁎ 0.02 −0.00 −0.02 −0.00 0.01 0.01 0.03⁎ −0.02 −0.00 −0.02 0.03⁎ 0.05⁎⁎⁎
22. Twice or more 0.21 0.41 −0.06⁎⁎⁎ 0.00 0.01 0.00 0.02 −0.00 −0.02 0.01 −0.01 −0.04⁎⁎⁎ 0.04⁎⁎⁎ −0.08⁎⁎⁎ −0.25⁎⁎⁎

Leader occupation aspiration

23. None 0.41 0.49 −0.14⁎⁎⁎ −0.02 0.05⁎⁎⁎ 0.04⁎⁎ 0.02 −0.04⁎⁎ −0.06⁎⁎⁎ −0.01 −0.02 −0.00 0.04⁎⁎ −0.16⁎⁎⁎ −0.03⁎
24. One 0.43 0.50 0.11⁎⁎⁎ 0.05⁎⁎⁎ −0.07⁎⁎⁎ −0.02 −0.02 0.04⁎⁎ 0.05⁎⁎⁎ −0.01 0.02 0.00 −0.02 0.13⁎⁎⁎ 0.02
25. Both 0.16 0.37 0.03⁎ −0.04⁎⁎ 0.02 −0.02 −0.01 −0.00 0.01 0.02 −0.01 0.00 −0.03 0.04⁎⁎ 0.01
26. Body height 179.66 6.51 0.17⁎⁎⁎ 0.01 −0.05⁎⁎⁎ −0.03⁎ −0.01 0.03⁎⁎ 0.05⁎⁎⁎ 0.02 0.02 0.00 −0.05 ⁎⁎⁎
0.10⁎⁎⁎
0.03⁎
27. Muscular capacity 5.18 1.69 0.11 ⁎⁎⁎
0.01 0.02 −0.01 0.02 −0.02 −0.00 −0.01 −0.01 −0.01 0.03 ⁎⁎
−0.01 0.00

Membership in scouts and/or sports club

28. Neither 0.60 0.49 −0.13⁎⁎⁎ −0.11⁎⁎⁎ 0.01 0.02 −0.00 −0.02 −0.00 −0.01 0.02 −0.01 0.01 −0.07⁎⁎⁎ −0.03⁎⁎
29. Scouts 0.06 0.23 0.12⁎⁎⁎ 0.06⁎⁎⁎ −0.06⁎⁎⁎ −0.04⁎⁎ −0.01 0.04⁎⁎ 0.06⁎⁎⁎ −0.03⁎⁎ 0.01 0.04⁎⁎⁎ −0.03⁎ 0.09⁎⁎⁎ 0.04⁎⁎
30. Sport 0.34 0.47 0.07⁎⁎⁎ 0.07⁎⁎⁎ 0.03⁎ 0.00 0.00 −0.00 −0.03⁎ 0.03⁎ −0.02 −0.01 0.01 0.02 0.01
31. Both 0.01 0.09 0.06⁎⁎⁎ 0.04⁎⁎ −0.03⁎ −0.02 0.00 0.03⁎ 0.01 −0.02 −0.01 0.02 0.01 0.05⁎⁎⁎ 0.04⁎⁎⁎

Frequency extracurricular activity

32. Never 0.11 0.32 −0.12⁎⁎⁎ −0.07⁎⁎⁎ 0.05⁎⁎⁎ 0.01 0.03 −0.04⁎⁎ −0.03⁎ 0.03⁎ 0.02 −0.05⁎⁎⁎ 0.02 −0.08⁎⁎⁎ −0.04⁎⁎
33. Sometimes 0.14 0.35 −0.01 0.01 −0.04⁎⁎ −0.01 0.01 −0.00 0.04⁎⁎ 0.01 0.01 −0.01 −0.00 0.01 0.02
34. Weekly 0.46 0.50 0.10⁎⁎⁎ 0.12⁎⁎⁎ 0.00 −0.02 −0.01 0.02 −0.01 −0.01 −0.02 0.03⁎ −0.00 0.06⁎⁎⁎ 0.04⁎⁎⁎
35. Not member in a club 0.28 0.45 −0.03 −0.08⁎⁎⁎ −0.01 0.01 −0.02 0.01 −0.01 −0.02 0.01 0.02 −0.01 −0.01 −0.03⁎
or association

Variable 14 15 16 17 18 19 20 21 22 23 24 25 26

14. IQ

Delay of gratification
15. Certainly SEK 100 now −0.10⁎⁎⁎
16. Probably SEK 100 now −0.04⁎⁎ −0.06⁎⁎⁎
17. Cannot choose −0.06⁎⁎⁎ −0.07⁎⁎⁎ −0.06⁎⁎⁎
18. Probably SEK 1000 in five years 0.06⁎⁎⁎ −0.18⁎⁎⁎ −0.17⁎⁎⁎ −0.19⁎⁎⁎
19. Certainly SEK 1000 in five years 0.04⁎⁎ −0.25⁎⁎⁎ −0.23⁎⁎⁎ −0.26⁎⁎⁎ −0.69⁎⁎⁎

Expulsion from classroom

20. Never expelled 0.07⁎⁎⁎ −0.02 0.03⁎⁎ 0.02 −0.01 −0.01
21. Once −0.03⁎ −0.01 −0.01 −0.01 0.00 0.01 −0.63⁎⁎⁎
22. Twice or more −0.05⁎⁎⁎ 0.03⁎ −0.02 −0.01 0.00 −0.00 −0.30⁎⁎⁎ −0.55⁎⁎⁎

(continued on next page)

307
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 2a (continued)

Variable 14 15 16 17 18 19 20 21 22 23 24 25 26

Leader occupation aspiration

23. None −0.13⁎⁎⁎ 0.04⁎⁎ 0.03⁎ 0.05⁎⁎⁎ −0.03 −0.03⁎⁎ 0.02 −0.01 −0.00
24. One 0.12⁎⁎⁎ −0.04⁎⁎ −0.02 −0.02 0.02 0.02 0.01 0.00 −0.01 −0.72⁎⁎⁎
25. Both 0.02 −0.00 −0.01 −0.04⁎⁎ 0.00 0.02 −0.03⁎ 0.01 0.02 −0.36⁎⁎⁎ −0.38⁎⁎⁎
26. Body height 0.14⁎⁎⁎ −0.02 −0.00 −0.02 −0.01 0.03 0.01 0.00 −0.01 −0.04⁎⁎ 0.02 0.03⁎

Variable 27 28 29 30 31 32 33 34 35

14. IQ −0.00 −0.07 ⁎⁎⁎

0.10 ⁎⁎⁎
0.01 0.06 ⁎⁎⁎
−0.13 ⁎⁎⁎
0.00 0.04 ⁎⁎
0.05⁎⁎⁎

Delay of gratification
15. Certainly SEK 100 now −0.03⁎ 0.03⁎ −0.02 −0.02 −0.01 0.03⁎ −0.01 −0.03⁎ 0.01
16. Probably SEK 100 now −0.00 0.01 −0.01 −0.01 −0.01 0.00 0.01 −0.02 0.02
17. Cannot choose −0.02 0.01 0.00 −0.01 −0.02 0.02 0.01 −0.02 0.00
18. Probably SEK 1000 in five years −0.01 0.01 −0.03⁎ −0.00 0.01 −0.02 0.02 −0.02 0.03⁎
19. Certainly SEK 1000 in five years 0.03⁎ −0.04⁎⁎ 0.04⁎⁎ 0.02 0.01 −0.01 −0.02 0.05⁎⁎⁎ −0.04⁎⁎

Expulsion from classroom

20. Never expelled −0.06⁎⁎⁎ 0.08⁎⁎⁎ 0.01 −0.08⁎⁎⁎ −0.02 0.02 −0.00 −0.07⁎⁎⁎ 0.07⁎⁎⁎
21. Once 0.01 −0.03⁎ 0.00 0.03⁎ 0.01 −0.02 −0.00 0.03⁎⁎ −0.02
22. Twice or more 0.05⁎⁎⁎ −0.05⁎⁎⁎ −0.01 0.06⁎⁎⁎ 0.01 0.01 0.01 0.04⁎⁎ −0.05⁎⁎⁎

Leader occupation aspiration

23. None −0.02 0.06⁎⁎⁎ −0.04⁎⁎ −0.04⁎⁎ −0.01 0.04⁎⁎ 0.00 −0.05⁎⁎⁎ 0.02
24. One 0.02 −0.05⁎⁎⁎ 0.02 0.04⁎⁎ 0.02 −0.06⁎⁎⁎ −0.00 0.04⁎⁎ 0.00
25. Both 0.01 −0.01 0.02 0.00 −0.02 0.02 0.00 0.02 −0.04⁎⁎
26. Body height 0.23⁎⁎⁎ 0.00 0.02 −0.01 0.00 −0.04⁎⁎ 0.02 −0.02 0.02

Variable 27 28 29 30 31 32 33 34 35

27. Muscular capacity

Membership in scouts and/or sports club

28. Neither −0.07⁎⁎⁎
29. Scouts −0.01 −0.30⁎⁎⁎
30. Sport 0.08⁎⁎⁎ −0.87⁎⁎⁎ −0.18⁎⁎⁎
31. Both −0.01 −0.11⁎⁎⁎ −0.02 −0.07⁎⁎⁎

Frequency extracurricular activity

32. Never −0.02 0.19⁎⁎⁎ −0.08⁎⁎⁎ −0.15⁎⁎⁎ −0.03⁎
33. Sometimes −0.01 0.02 −0.05⁎⁎⁎ 0.01 −0.03⁎ −0.15⁎⁎⁎
34. Weekly 0.05⁎⁎⁎ −0.59⁎⁎⁎ 0.23⁎⁎⁎ 0.48⁎⁎⁎ 0.09⁎⁎⁎ −0.33⁎⁎⁎ −0.38⁎⁎⁎
35. Not member in a club or association −0.03⁎ 0.51⁎⁎⁎ −0.15⁎⁎⁎ −0.44⁎⁎⁎ −0.06⁎⁎⁎ −0.22⁎⁎⁎ −0.26⁎⁎⁎ −0.58⁎⁎⁎

apart from those pertaining to the draft.
In both the analysis of military officer suitability and nominations
for class party organizer, the models were tested with and without fixed
effects of school classes by use of a Chow test. In the first case there
were no significant differences between the models
(prob > Chi2 = 0.149) and therefore we did not use fixed effects in
this case. In the latter case, however, the Chow test did show a sig-
nificant difference between the models with and without fixed effects.
Therefore, the analyses of nominations for class party organizer include
fixed effects for school classes (Halaby, 2004).
Missing data
Although the data set is very comprehensive and has a high cov-
erage rate, there are of course missing data. First, the Conscription
Board register included information on 6928 individuals from the ori-
ginal sample of 7719. The loss of observations between these points is
caused by death, emigration, or exemption from draft due to religious
faith, severe handicap, already serving in the military, working as a
seaman or other valid grounds for exemption. It is, however, not pos-
sible to break down the loss of observations by different causes. Because
the vast majority of male cohorts were drafted at the time, the no-show
group is likely to be quite select and it is therefore not reasonable to
impute values for missing observations. The point of departure in this
study is, then, males in the Stockholm Birth Cohort who were registered
in the National Conscription Board register (N = 6928).
In addition to missing observations, there was item non-response on
most variables (see Table 1).13 There are several ways of handling in-
complete data with different gains and losses. For item non-response we
performed multiple imputations.14
Post-estimations
All post-estimations were based on listwise deletion of missing data.
For the Hurdle model we calculated the average marginal effects (dy/
dx), i.e. how each additional change in x is associated with a change in
y.15 In order to post-estimate effects in the negative binominal regres-
sion analysis we calculated the incidence ratio relationship (IRR).
13
Among the draftees 390 individuals (5.3%) had not participated in the
school survey at all, presumably mainly due to absence from school when the
survey was carried out.
14
Multiple imputation is not supported by Stata for Cragg-Hurdle or negative
binominal regressions, but the cmdok command allows unsupported estima-
tions with multiple imputations estimates. Because the data set is large and
consists of both categorical and continuous variables imputations were per-
formed with fully conditional specification (FCS). The estimated model has a
minimum degree of freedom > 100. We used the Stata command “how_ma-
ny_imputations” to determine the number of imputations (von Hippel, 2018).
15
Using the “margins, dydx (*)” command in Stata.

308
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 2b effects (dy/dx) which represent the conditional average change in y

Descriptive statistics: distribution among categorical variables from a one-unit change in x. The confidence intervals also provide a
(N = 6928). useful hint of the level of uncertainty for each variable.
Variable Percent Among the variables measuring early life experiences, Table 4
suggests that family socioeconomic status in childhood strongly influ-
Early life experiences ences leadership assessment, and the association is positive and linear.
Family socioeconomic status
It is worth pointing out that we have controlled for other factors such as
Unskilled workers 3
Working class, skilled workers 16 IQ, which almost consistently varies between social groups in favor of
Lower middle class, entrepreneurs 22 higher social classes (Fischbein, 1980).
Lower middle class 7 The effect is not great – 0.12 units on the 1–9 scale – but Table 4
Upper and upper-middle class 35
shows that being firstborn does have an effect on suitability assessment
Missing 16
Birth order
compared to being the youngest among siblings. Parental support also
Youngest child 30 showed a positive but weak effect on leadership assessment. The same
Only child 21 applies to academic intrinsic motivation.
First born 32 Most notable in Table 4 is the effect of IQ also on the actual as-
Middle child 17
sessment of officer suitability. For every increase of one standard de-
Foundational traits
Delay of gratification viation on IQ, we see a rise of 1.33 points on the 1–9 assessment scale.
Certainly SEK 100 now 5 Delay of gratification (“marshmallow” measure) showed no effect
Probably SEK 100 now 5 on leadership assessment, but the other measure of self-regulatory skills
Cannot choose 6
showed a slight impact: Being expelled from the classroom was slightly
Probably SEK 1000 in five years 30
Certainly SEK 1000 in five years 44
“beneficial” for leadership assessment if the draftee had been expelled
Missing 10 on one occasion.
Expulsion from classroom Draftees who aspired to become leadership occupations when they
Never expelled 23 were in the 6th grade (an army officer or director) did receive a higher
Once 48
assessment score compared to peers who did not have such ambitions.
Twice or more 19
Missing 10 The average increase in assessment ratings was around 0.17 regardless
Leader occupation aspiration of whether they had wished to become one, or both, of these occupa-
None 36 tions.
One 38 Among other foundational traits, Table 4 shows that a one-cen-
Both (director and manager) 14
Missing 13
timeter increase in body height had an average impact on suitability
Developmental experiences assessment of 0.03 points. This may not seem much at first glance, but
Membership in scouts and/or sport club equals an average increase of 0.3 on the leadership assessment score
Neither 52 between a body height of 1.65 and 1.75 m. Muscular power seems to
Scouts 5
have approximately the same impact as body height, but they cannot be
Sport 30
Both 1 compared directly because they are measured on different scales.
Missing 13 When it comes to developmental experiences being a member of a
Frequency extracurricular activity sports club or the scouts is clearly influential for the assessment of
Never 10 leadership capability. As shown in Table 4 the association is particu-
Sometimes 13
Weekly 41
larly strong for scouting, with an average increase of 0.53 on the officer
Not member in a club or association 25 suitability score, compared to a 0.31 increase for sports club member-
Missing 11 ship. The difference between these two is statistically significant at the
0.001 level. Moreover, the “dosage” of developmental experiences was
important in the sense that frequency of extracurricular activities meant
Stata does not calculate elasticities for fractional variables (ey/ex) an increase in officer suitability score. There was, however, no sig-
so we have presented marginal effects in Figs. 4 and 5 with dy/dx. nificant difference between the intensity of activities (sometimes or
weekly) – the important thing is to participate.
Results In order to visualize marginal effects we present these graphically
for IQ and family socioeconomic status in Figs. 4 and 5. IQ is a foun-
Two Cragg-Hurdle regression models are presented in Table 3. One dational trait of the individual and family socioeconomic status is an
presents coefficients and standard errors using listwise deletion and the environmental factor describing early life experiences. Table 4 portrays
other using multiple imputation. Within each model the first two col- the average marginal effects, while Figs. 4 and 5 illustrate the effects for
umns present coefficients and standard errors for being chosen for as- each value of the respective variables thereby also visualizing the
sessment at all, i.e. passing the “hurdle”, and the next two columns curvilinear pattern for IQ.
presenting ditto for the actual assessment on the 1–9 scale for those
who did pass the “hurdle”. Informal ascription of leadership: Nominations as class party organizer
First, we note that the results do not differ much between the two
models based on listwise deletion and multiple imputation respectively. As a secondary analysis we studied the informal ascription of lea-
Second, IQ had a significant, positive and curvilinear effect on passing dership status in childhood through a negative binominal regression on
the “hurdle”. Moreover, body height and muscular power also showed a nominations as class party organizer (in the 6th grade). The results are
significant impact in the first step – as could be expected. Third, all the presented in Table 5. In the same way as for the Cragg-Hurdle regres-
included factors also showed a significant impact on the actual assess- sion, Table 5 presents results for both listwise deletion and multiple
ment of military officer suitability, but the curvilinear effect of IQ was imputation. Again, we see very slight differences between these two
not significant in this stage. models.
The coefficients in Table 3 are not directly interpretable. We cannot Interpretable post-estimations of regression effects (incidence rate
conclude that a certain change in x produces a given change in y. To ratios, IRR) are presented in Table 6. Keeping all other covariates at a
obtain the effect of a covariate on the model, we calculated marginal constant level (means), IRR measures ratio of certain events for

309
T. Reitan, S.-Å. Stenberg
8
Predictive margins with 95 % CIs
6
4
2
0
-4
Fig. 4. Conditional mean estimates of military officer suitability.
5
Predictive margins with 95 % CIs
4
3
1 2
Fig. 5. Conditional mean estimates of military officer suitability.
categorical variables. That is, the number of events (in this case no-
minations as class party organizer) per unit of time (school lesson),
divided by the groups we are studying – analogous to percent for
continuous variables. The interpretation of IRR is similar to that of dy/
dx, but a more familiar comparison would be odds ratios (OR). If the
confidence interval includes 1, the ratio is not significant.
For family socioeconomic status there was a significant impact for all
classes except the two lowest (unskilled and skilled workers). If the boy
came from the highest social class, the incidence rate increased 0.46 times
compared to the lowest social class, controlling for the other variables in
the model. As far as birth order is concerned, the youngest siblings were
the only group to show a significant impact on nominations. Boys who
perceived to have strong parental support were also more likely to receive
more nominations. Academic intrinsic motivation did not, however, show
any effect on nominations for class party organizer.
Among the foundational traits, IQ showed a strong effect on
The Leadership Quarterly 30 (2019) 298–319
-2 0 2
IQ (standardized)
3 4 5
Social class 1963 (1 highest, 2 lowest)
nominations. In contrast to the model for assessment of military officer
suitability, there was no curvilinear pattern between IQ and nomina-
tions for class party organizer.
As in the case of military officer assessment, delay of gratification
did not have any impact on nominations whatsoever. Expulsion from
the classroom, however, was influential – and there seemed to be a
difference between being expelled once compared to several times
(p = 0.08). Wanting to become a leader, i.e. aspiring to become a di-
rector or manager, did not increase the chances of receiving many
nominations as class party organizer. In fact, wanting to become both a
director and a manager significantly decreased the likelihood of being
nominated for class party organizer.
Developmental experiences are more tangible and known to other
pupils, which may be one reason that participation in sports and in
extracurricular activities was clearly positively associated with nomi-
nations. For those who were most active (weekly) the impact on
310
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 3
Cragg hurdle regression estimates for score on assessment of military officer suitability (0–9, truncated at 0). Clustered standard errors by 612 school classes.
Listwise deletionb (N = 5069) Multiple imputationc (N = 6928)

Estimates for being assessed Estimates for assessment Estimates for being assessed Estimates for assessment

Independent variables Coefficient S.E. Coefficient S.E. Coefficient S.E. Coefficient S.E.

Constant −2.12⁎⁎⁎ 0.58 −0.17 0.94 −2.15⁎⁎⁎ 0.56 −0.66 0.92

Early life experiences

Family socioeconomic status
Unskilled workersa
Working class, skilled workers 0.33⁎ 0.14 0.33⁎⁎ 0.13
Lower middle class, entrepreneurs 0.53⁎⁎ 0.18 0.41⁎⁎ 0.15
Lower middle class 0.66⁎⁎⁎ 0.11 0.65⁎⁎⁎ 0.10
Upper & upper-middle class 1.01⁎⁎⁎ 0.12 1.03⁎⁎⁎ 0.12
Birth order
Youngest childa
Only child 0.01 0.09 0.05 0.09
First born 0.16⁎ 0.08 0.15⁎ 0.08
Middle child −0.02 0.10 −0.01 0.09
Parental support 0.11⁎⁎⁎ 0.02 0.11⁎⁎⁎ 0.02

Leadership learning capacities and skills

Academic intrinsic motivation 0.07⁎⁎⁎ 0.01 0.07⁎⁎⁎ 0.01

Foundational traits
IQ in 6th grade (standardized) 0.87⁎⁎⁎ 0.03 0.45⁎⁎⁎ 0.05 0.79⁎⁎⁎ 0.02 0.52⁎⁎⁎ 0.06
IQ squared −0.11⁎⁎⁎ 0.02 0.03 0.03 −0.11⁎⁎⁎ 0.02 0.01 0.03
Delay of gratification
Certainly SEK 100 nowa
Probably SEK 100 now 0.09 0.22 −0.01 0.21
Cannot choose −0.04 0.20 −0.08 0.19
Probably SEK 1000 in five years 0.13 0.16 0.08 0.15
Certainly SEK 1000 in five years 0.12 0.16 0.07 0.15
Expulsion from classroom
Never expelleda
Once 0.15 0.08 0.12 0.07
Twice or more −0.10 0.10 −0.12 0.10
Both
Leader occupation aspiration
Nonea
One 0.22⁎⁎ 0.07 0.21⁎⁎ 0.07
Both 0.23⁎ 0.10 0.21⁎ 0.09
Body height 0.02⁎⁎⁎ 0.00 0.01⁎ 0.01 0.02⁎⁎⁎ 0.00 0.01⁎⁎ 0.01
Muscular capacity 0.03⁎⁎ 0.01 0.19⁎⁎⁎ 0.02 0.04⁎⁎ 0.01 0.18⁎⁎⁎ 0.02

Developmental experiences
Membership in scouts and/or sports club
Neithera
Scouts 0.72⁎⁎⁎ 0.14 0.74⁎⁎⁎ 0.13
Sport 0.42⁎⁎⁎ 0.08 0.36⁎⁎⁎ 0.08
Both 0.69⁎⁎ 0.26 0.75⁎⁎ 0.27
Frequency extracurricular activity
Nevera
Sometimes 0.29⁎ 0.13 0.24⁎ 0.13
Weekly 0.34⁎⁎ 0.13 0.34⁎⁎ 0.13
Not member in a club or association 0.23 0.14 0.18 0.12

a
Reference category.
b
Pseudo R2 = 0.11.
c
20 imputations.
⁎
p < 0.05.
⁎⁎
p < 0.01.
⁎⁎⁎
p < 0.001.

nominations was almost twice as strong compared to those who were while the air rang. Ralph looked at him, eager to offer something.
less active (sometimes). It should be noted, though, that scouting was ‘The choir belongs to you, of course.’
not favorable vis-à-vis nominations as class party organizer. ‘They could be the army – ’.
‘Or hunters – ’.
Discussion ‘They could be – ’.
The suffusion drained away from Jack's face. Ralph waved again for
Ralph counted. silence.
‘I'm chief then.’ ‘Jack's in charge of the choir. They can be – what do you want them
The circle of boys broke into applause. Even the choir applauded; to be?’
and the freckles on Jack's face disappeared under a blush of morti- ‘Hunters.’
fication. He started up, then changed his mind and sat down again

311
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 4 the 6th grade. All of these factors turned out to be important for lea-
Conditional mean estimates of assessment of military officer suitability. Cragg dership emergence although in varying ways.
hurdle regression (N = 5069). Social class was an influential factor for leadership ascription both
dy/dxa 95% CI in childhood and for assessment of leadership capability at the draft. As
noted in so much social research, the imprint of social class is evident at
Independent variables an early age and seems to be maintained and perhaps emphasized in
young adulthood. Position among siblings also turned out to be influ-
Early life experiences
Family socioeconomic status ential in this study, but in diverging directions. As could be expected,
Unskilled workersb the eldest sibling was more likely to be considered officer material. The
Working class, skilled workers 0.24 0.04 0.43 pattern was reverse for arranging a class party, with the younger sibling
Lower middle class, entrepreneurs 0.39 0.12 0.64
being the favored choice. This is perhaps surprising but maybe rea-
Lower middle class 0.48 0.32 0.64
Upper and upper-middle class 0.74 0.58 0.90
sonable given the different tasks at hand. Studies have shown that later-
Birth order born children are more popular among peers compared to first-borns. A
Youngest childb suggested explanation is that younger siblings are more likely to de-
Only child 0.01 −0.11 0.13 velop interpersonal or social skills in order to compensate for their
First born 0.12 0.00 0.23
power deficit vis-à-vis older siblings (Miller & Maruyama, 1976). Social
Middle child −0.01 −0.15 0.12
Parental support 0.08 0.06 0.10 skills are definitely relevant for party organizers, while the military are
Leadership learning capacities and skills probably searching for that extra sense of responsibility and authority
Academic intrinsic motivation 0.05 0.03 0.07 often associated with first-born.
Foundational traits
Parental support also showed a positive effect on leadership emer-
IQ in sixth grade (standardized) 1.33 1.27 1.41
Delay of gratification
gence in both instances, although not as notably in early adulthood. It is
Certainly SEK 100 nowb plausible that the “influence” of parent support has worn off by the time
Probably SEK 100 now 0.07 −0.23 0.37 the cohort has reached early adulthood or rather transported through
Cannot choose −0.03 −0.31 0.26 educational and career choices, attitudes, emotional and physical
Probably SEK 1000 in five years 0.09 −0.14 0.32
status, etc. Parental support may provide the offspring with a general
Certainly SEK 1000 in five years 0.09 −0.14 0.32
Expulsion from classroom self-confidence, but it may also have materialized in terms of en-
Never expelledb couragement of leadership role occupancy or even specific attitudes
Once 0.11 −0.00 0.22 towards the military as such (Legree et al., 2000). In contrast, academic
Twice or more −0.08 −0.22 0.07
intrinsic motivation did not have a particularly strong impact on lea-
Leader occupation aspiration
Noneb
dership emergence in any instance.
One 0.16 0.06 0.26 Among the foundational traits, IQ stands out as particularly im-
Both 0.17 0.03 0.31 portant for leadership emergence, i.e. ascription or assessment of lea-
Body height 0.03 0.02 0.04 dership qualities. As such the results are compatible with those from a
Muscular capacity 0.17 0.13 0.21
large-scale, long-term prospective study of two British cohorts in which
Developmental experiences
Membership in scouts and/or sports club childhood cognitive ability was found to be highly influential for actual
Neitherb leadership role occupation at various stages in adulthood (Daly et al.,
Scouts 0.53 0.34 0.73 2015). In our study, the patterns of association did differ somewhat
Sports 0.31 0.19 0.43 between childhood and early adulthood though. In the former case the
Both 0.50 0.12 0.89
Frequency extracurricular activity
impact of IQ showed a linear patterns, while the effect of IQ wore off at
Neverb the highest levels in relation to military officer suitability. These results
Sometimes 0.21 0.02 0.40 suggest that there is a limit to the value of IQ, so to speak, in the sense
Weekly 0.25 0.08 0.42 that too much of a good thing can be problematic (Antonakis et al.,
Not member in a club or association 0.17 −0.02 0.36
2017).
a
dy/dx for factor levels is the discrete change from the base level. The well-known “marshmallow” test has been used in different
b
Reference category. forms to measure a component of self-control thought to be associated
with various outcomes. Our analyses showed that this factor had no
effect whatsoever in any of the models. Recent studies have shown that
the ability to delay gratification is an aspect of self-regulation that is
Commanding a troop at war is one thing – organizing a class party closely associated with cognitive ability and sensitive to inclusion of
in the 6th grade something quite different. Having said that, deciding control variables (Burks, Carpenter, Goette, & Rustichini, 2009;
on which qualities are relevant for the task and identifying individuals Dohmen, Falk, Huffman, & Sunde, 2010; Watts, Duncan, & Quan,
who have these qualities can certainly be challenging in both cases. For 2018). A closer look at this relationship showed that delay of gratifi-
Jack and Ralph, two youngsters struggling to survive on the desert is- cation had no connection to choice of class party organizer at all (re-
land, leadership is more similar to commanding troops than directing a gardless of inclusion of control variables), while its effect was neu-
choir – or organizing a party. tralized by IQ in relation to officer suitability. Other self-regulatory
Leadership ascription in childhood and leadership assessment in skills were of some interest though. Being expelled from the classroom
early adulthood takes place in different contexts at different points in (even on several occasions) was positively linked to nominations a class
the life cycle, and is steered by diverging considerations. Still, this study party organizer, but the effect had diminished by the time the boys
has shown that the process of identifying leadership qualities share were drafted. A certain degree of rebelliousness can be advantageous
many similarities despite varying contexts and phases in life. Also, the vis-à-vis leadership emergence, but maybe less so in a military context
study provides ample support for the application of a longitudinal and where discipline and order are fundamental. As discussed in the lim-
life-cycle approach to leadership emergence as represented by the itations section, peer nominations may simply reflect perceived popu-
theoretical framework in Fig. 1. larity rather than leadership qualities and popularity in childhood and
The early life experiences analyzed in this study focused on the adolescence may certainly be a double-edged sword. Allen et al. (2005)
socioeconomic situation in the family of origin, the pupil's rank among found that popularity in adolescence was associated with positive
siblings, and parental support as perceived by the pupil when he was in markers of social development and functioning, but also increased

312
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 5
Negative binominal regression estimates for nominations as class party organizer.a
Listwise deletion. Clustered standard errors by 607 school Multiple imputation.c Clustered standard errors by 612 school
classes.b N = 4654 classes. N = 6928

Independent variables Coefficient S.E. Coefficient S.E.

Constant 0.62⁎⁎⁎ 0.15 0.52⁎⁎⁎ 0.13

Early life experiences
Family socioeconomic status
Unskilled workersd
Working class, skilled workers −0.00 0.07 −0.00 0.07
Lower middle class, entrepreneurs 0.21⁎ 0.09 0.17⁎ 0.09
Lower middle class 0.20⁎⁎ 0.06 0.17⁎⁎⁎ 0.06
Upper and upper-middle class 0.38⁎⁎⁎ 0.08 0.33⁎⁎⁎ 0.08
Birth order
Youngest childd
Only child −0.14⁎ 0.06 −0.10⁎ 0.05
First born −0.19⁎⁎⁎ 0.05 −0.12⁎⁎ 0.05
Middle child −0.19⁎⁎ 0.07 −0.17⁎⁎ 0.05
Parental support 0.07⁎⁎⁎ 0.01 0.06⁎⁎⁎ 0.01
Leadership learning capacities and skills
Academic intrinsic motivation 0.00 0.00 0.01 0.01
Foundational traits
IQ in 6th grade (standardized) 0.31⁎⁎⁎ 0.023 0.26⁎⁎⁎ 0.03
IQ squared 0.02 0.01 0.02 0.02
Delay of gratification
Certainly SEK 100 nowd
Probably SEK 100 now −0.22 0.13 −0.17 0.11
Cannot choose −0.24 0.12 −0.14 0.09
Probably SEK 1000 in five years −0.09 0.10 −0.04 0.07
Certainly SEK 1000 in five years −0.13 0.10 −0.05 0.07
Expulsion from classroom
Never expelledd
Once 0.12⁎ 0.05 0.12⁎⁎ 0.04
Twice or more 0.22⁎⁎ 0.07 0.19⁎⁎ 0.06
Leader occupation aspiration
Noned
One −0.00 0.05 −0.00 0.04
Both −0.11 0.06 −0.12⁎ 0.05
Developmental experiences
Membership in scouts and/or sport club
Neitherd
Scouts −0.01 0.11 0.06 0.09
Sport 0.15⁎⁎ 0.05 0.13⁎⁎ 0.05
Both 0.31 0.19 0.23 0.17
Frequency extracurricular activity
Neverd
Sometimes 0.18⁎ 0.08 0.13⁎ 0.07
Weekly 0.33⁎⁎⁎ 0.08 0.31⁎⁎⁎ 0.06
Not member in a club or association 0.10 0.07 0.06 0.06

a
Fixed effects of school classes included.
b
Pseudo R2 = 0.05.
c
20 imputations.
d
Reference category.
⁎
p < 0.05.
⁎⁎
p < 0.01.
⁎⁎⁎
p < 0.001.

likelihood of behavioral changes consistent with peer norms but not
necessarily compatible with norms of the adult society.
Aspirations for leadership occupations in childhood showed a re-
verse pattern. Wanting to become a director or a manager in the 6th
grade did not lead to more nominations as class party organizer –
wanting to become both even had a negative impact – but these as-
pirations made a greater impression on military officer evaluators.
Physical characteristics are as foundational traits obviously closely
linked to genetic predisposition (see Fig. 1), but also influenced by early
childhood experiences (e.g. family economic status) and also by the
historical period and individual life-events (Elder, 1974; Stanner &
Yudkin, 2001). We were only able to include measures of physical
characteristics in the analysis of leadership assessment in connection
with the draft. As has been noted in other studies body height is
strongly associated with perceptions of leadership (Stulp et al., 2013).
The body height bonus was noteworthy, with a 0.3 increase on the
assessment scale for every 10-cm increase in height. The impact of
muscular power was also significant. Although these results confirm
previous research on the importance of body height. We should not
draw too hasty conclusions about the magnitude of this association. The
measures of body height and muscular power were a manifest part of
this assessment process and almost certainly emphasized more in this
particular context compared to other situations involving formal or
informal assessment of leadership qualifications.
Apart from the foundational traits and early childhood experiences,
the importance of developmental experiences was evident in both in-
stances in the cohort's lives. Scouting was not related to peer nomina-
tions in childhood, but was definitely a long-term investment in clearly

313
T. Reitan, S.-Å. Stenberg The Leadership Quarterly 30 (2019) 298–319

Table 6 socialization, shared experiences of certain cohorts, and specific life-

Incidence rate ratios (IRR) of nominations for class party organizer. events with significant impact on individual lives (Baltes, Reese, &
Negative binominal regression (N = 4654)a. Lipsitt, 1980) has not been the aim of this study. However, these issues
IRR 95% CI are relevant when discussing generalizability and the specific empirical
context.
Independent variables We have argued that time (and place) must be taken into account in
a life-cycle approach to leadership development. In practice, historical
Early life experiences
Family socioeconomic status 1963 period cannot be a variable in the traditional sense. Many aspects of
Unskilled workersb human life can be manipulated, e.g. for scientific purposes, but human
Working class, skilled workers 1.00 0.87 1.15 behavior always takes place in a certain historical and geographic
Lower middle class, entrepreneurs 1.23 1.03 1.47
context which is usually fixed. It does matter whether potential leaders
Lower middle class 1.22 1.07 1.38
Upper and upper-middle class 1.46 1.25 1.70
are born in a period of high levels of living and low mortality rates or in
Birth order a society characterized by deprivation and social unrest – perhaps even
Youngest childb with a high risk of not surviving childhood and adolescence at all.
Only child 0.87 0.78 0.98 Leadership is developed and practiced in settings that are not
First born 0.83 0.74 0.92
wholly controllable and it is therefore problematic to apply the golden
Middle child 0.83 0.75 0.94
Parental support 1.07 1.05 1.09 standard in intervention research, randomized controlled trials (RCT),
Leadership learning capacities and skills in leadership research. Moreover, the obvious advantages of long-
Academic intrinsic motivation 1.01 0.99 1.03 itudinal study designs in comparison with cross-sectional approaches
Foundational traits
are also “polluted” by environmental factors and constant change. The
IQ in 6th grade (standardized) 1.37 1.30 1.44
IQ squared 1.02 0.98 1.06
longer the longitudinal approach, the greater the loss of control over
Delay of gratification historical and social change. Generalizability may always be questioned
Certainly SEK 100 nowb in longitudinal studies such as this one.
Probably SEK 100 now 0.80 0.62 1.03 A longitudinal approach is not only a question of repeated mea-
Cannot choose 0.79 0.62 1.00
surements over time, but also about conducting prospective research –
Probably SEK 1000 in five years 0.91 0.75 1.06
Certainly SEK 1000 in five years 0.93 0.80 1.07 studying groups not chosen based on the outcome variable (i.e. those
Expulsion from classroom who are unemployed, have cancer, have become grandparents, − or are
Never expelledb leaders). It is very easy to forget about all the promising computer
Once 1.13 1.03 1.24
programmers who did not become Bill Gates, overlooking one of the
Twice or more 1.25 1.09 1.43
Leader occupation aspiration
most important determinants, namely their date – and we would also
Noneb add place – of birth (Dixon, Horton, & Weir, 2011). This approach
One 1.00 0.92 1.09 obviously entails “an early start”, and schools are extremely suitable
Both 0.90 0.80 1.01 contexts in that respect. They are relatively standardized institutions,
Developmental Experiences
universal in their aim to target everyone within a certain age group.
Membership in scouts and/or sport club
Neitherb This was also the case for the children born in 1953 and residing in the
Scouts 0.99 0.80 1.22 Stockholm metropolitan area in 1963. Moreover, universalism was also
Sport 1.16 1.05 1.29 a key factor at the other end of our study – the draft. At this time, in
Both 1.37 0.95 1.97 principle all males were subject to a drafting procedure after their 18th
Frequency extracurricular activity
Neverb
birthday. Our study would, then, have been less valuable in today's
Sometimes 1.20 1.02 1.41 situation, where obligatory military service was abolished in 2010 and
Weekly 1.39 1.20 1.62 reintroduced on a small scale for both men and women in 2017 (J.
Not member in a club or association 1.10 0.96 1.27 Östberg, 2018; Strand & Berndtsson, 2015). Sweden was also at the
a time a very homogeneous society socially, economically, and ethni-
Fixed effects of school classes included.
b cally. Most mothers were homemakers or working part-time. Income
Reference category.
differences were generally low, and it was an optimistic era in terms of
low unemployment and substantial economic growth. The welfare state
was in full development with large advances in education, social wel-
influencing leadership assessment in early adulthood. Being active in a fare and health care. These children could generally dream about – and
sports club, or generally engaged in extracurricular activities, was de- expect – a better future compared to their parents.
finitely positive in relation to leadership emergence in both childhood The geographical context is easier to reproduce in the sense that the
and early adulthood. If being active in a sports club can be translated Stockholm metropolitan area is still an area with a different social
into “athleticism”, these findings clearly correspond to those presented composition compared to other parts of the country. For example, re-
by Vannatta, Gartstein, Zeller, and Noll (2009). In their study of peer latively fewer fathers were farmers or industrial workers then and this
perceptions of attributes and various behavioral qualities, athleticism is bound to have influenced the children's occupational choices. This
was found to be associated with leadership, particularly for boys. would probably also be the case today. Another common factor is the
age-normative issue related to biological maturation: The cohort con-
Generalizability and relevance for leadership research sists of boys during a crucial formative period in their lives. In grade 6
they are on the verge of puberty and adolescence – an age when all
Although time is an obvious aspect of all human development, it is youngsters are seeking their identity, male norms, and working on peer
often overlooked and poorly studied – in leadership studies and other relations. The next point of measurement is at the other end of the
disciplines alike (Bartolini, 1993; Castillo & Trinh, 2018). Foundational adolescence tunnel when the now young men are confronted with ex-
traits are characteristics that are more stable and with relatively higher tensive demands on them as adult citizens in a setting reeking with
heritability (or intergenerational transmission as may be a more male norms.
common term in some social sciences). Such a proposition cannot be It is possible, though, that the importance of peer relations was not
made without an underlying temporal assumption. Unveiling the re- as crucial for this sample as it might have been in a group of boys
lative importance of biological maturation and age-graded without adult male role models. As children most of the cohort

314
T. Reitan, S.-Å. Stenberg
members lived in two-parent families where fathers – and many mo-
thers – were in employment. Birth order, family socioeconomic status
and cognitive ability are likely to be relevant in most historical and
cultural contexts – but probably to varying degrees and in varying
ways.
Besides the unsolvable issue of how to generalize results from
(historical) longitudinal research, our study also begs for some com-
ments on the choice of setting vis-à-vis leadership studies. In their de-
fense and encouragement of unconventional contexts, Bamberger and
Pratt (2010) particularly mention primary and secondary schools as an
ideal, multilevel setting for exploring boundaries of extant theories and
for generating new theory (p. 667). Another advantage of this study is
that data collection took place in controlled environments; in the
classroom under the supervision of teachers, through use of register
data which could not be manipulated, and through a highly controlled
and standardized draft procedure.
Limitations
Despite the uniqueness of data and many advantages of the pro-
spective design, there are several limitations to this study which need to
be taken into consideration. We have already mentioned that the
Project Metropolitan was not specifically designed for the purpose of
studying leadership. The questions and variables in the dataset are
therefore not directly applicable for this purpose, which means that a)
the variables are not very valid measures of the theoretical categories
we are trying to explore, and b) we lack data on other closely related
concepts and important confounders.
Neither of the dependent variables were specifically designed to
measure leadership qualities in a broad sense. The draft board's primary
goal is of course to find suitable candidates to serve as military officers.
That is, to practice leadership in a very specific context under very rigid
organizational circumstances. Having said that, the universal con-
scription and training of officers has in practice also supplied the labor
market and society at large with carefully selected and well-trained
leaders (Sundevall, 2017). As such, the assessment of military officer
suitability is a reasonable measure of leadership qualities. Access to
socio-metric data from when the cohort was aged 12–13 is of course
unique and invaluable, but nominations for class party organizer is
probably more remotely related to leadership per se and may simply
reflect (perceived) popularity. Popularity can, of course, also be ex-
tremely volatile and be associated with factors that are impossible to
encompass in a survey such as this (e.g. a big brother on the local
football team, newborn kittens at home, owning the latest Beatles re-
cord, or having a father who is an airline pilot). Moreover, physical
appearance, athletic ability and academic competence were found to be
moderately to strongly related to peer ratings of social acceptance
among pupils in elementary school (Vannatta et al., 2009). The relative
importance varied, however, been grades and gender, which reminds us
that the peer nominations in this study came from both boys and girls in
the classroom and it is possible that there were different factors ef-
fecting the likelihood of a boy being nominated compared to a girl
(Chiao, Bowman, & Gill, 2008). Almost half of the boys had been ex-
pelled from the classroom on at least one occasion. This may have been
both expected and accepted for a boy, but a sign of severe deviance and
a source of mistrust for a girl. One mitigating factor in this respect is
that the nominations were made anonymously, in writing, and the
pupils did not know how many nominations they received or from
whom.
There are of course a number of potential weaknesses as far as lack
of information is concerned. Although the dataset is very comprehen-
sive and consists of a variety of information from different sources, we
were not able to include subjective data from the parents, compre-
hensive data on physical appearance, and knowledge about specific
315
The Leadership Quarterly 30 (2019) 298–319
events which may have influenced an individual to a great degree (e.g.
serious illness, loss of a family member, or so). There are abundant
examples of people's lives changing overnight following events that
could not easily be foreseen (Hogan & Kaiser, 2005) and exposure to the
same influences may have very different impact depending on personal
and individual resilience or whether this experience is shared by most
of one's “contemporaries”.
Closely linked to this issue is the fact that not all boys from the
cohort presented at the draft. Around 10% were not present in the
Conscription Board register. It is conceivable that some of the cohort
members were even “non-selected” at an earlier stage, before being
called to the draft. The youths with the least academic motivation,
cognitive ability, developmental experiences, parental support and so
forth might already have gone off to sea, emigrated, been incarcerated,
become seriously ill, or even passed away. Including these individuals
might have produced even more distinct results. It should also be kept
in mind that the cohort was selected in 1963, which means that some of
the potential cohort members (born in 1953) had already passed away
or moved from the area during their first childhood years. This cir-
cumstance is of course important to recognize in studies focusing on
premature mortality or early childhood outcomes.
Another limitation is the lack of data on physical characteristics in
the 6th grade and the limited ditto at the draft. For example, we did not
have access to any data on physical or facial appearance, which could
have been relevant in several different ways in this study. Physical
appearance has for example, been shown to effect academic outcomes
in the sense that students with attractive peers received higher grades
compared to those in a course with average peers. This may simply
reflect a form of discrimination, but may also be explained by increased
motivation and effort when classmates are more attractive (Hernández-
Julián & Peters, 2018). Appearance and leadership ascription is also
relevant in childhood (Vannatta et al., 2009). Facial appearance has
also been linked to inferences about competence, for example among
political leaders (Todorov, Mandisodza, Goren, & Hall, 2005), and fa-
cial appearance was found to influence promotions to top positions in
the army (Mueller & Mazur, 1996).
He was old enough, twelve years and a few months, to have lost the
prominent tummy of childhood; and not yet old enough for ado-
lescence to have made him awkward. You could see now that he
might make a boxer, as far as width and heaviness of shoulders
went, but there was a mildness about his mouth and eyes that
proclaimed no devil.
(Description of Ralph in chapter 1 of Lord of the Flies)
Obviously, a whole range of verbal and non-verbal traits and ges-
tures are possible predictors of leadership ascription in a group
(Ridgeway, 2002; Sanchez-Cortes, Aran, Jayagopi, Schmid Mast, &
Gatica-Perez, 2013).
Another important limitation relates to the use of IQ in the Cragg-
Hurdle analysis. The draft officer had knowledge of the draftees' IQ
scores when deciding on a) who to refer for assessment, and b) the
actual assessment. Including the IQ score from the draft would there-
fore have been problematic, particularly for the first “hurdle”, given
that the data set provides evidence of an informal cut-off score of 5 or
more for being referred for assessment in the first place. Therefore, we
included the IQ score from the 6th grade instead. Nevertheless, there is
a strong correlation between these two measures of IQ and it is likely
that the “cursory” sorting mechanism continued on to the second part
of the process, i.e. the actual assessment of military officer suitability.
Lastly, it should be pointed out that we were not able to correct for
the effects of measurement error, for example regarding cognitive
ability. This was due to software limitations in combining errors-in-
variables in non-linear or complex models such as negative binominal
regressions or Cragg-Hurdle models.
T. Reitan, S.-Å. Stenberg
Conclusions
Individual and environmental conditions in childhood and adoles-
cence are not absent in the discourse on leadership emergence, be it in
early life or later in adult life. For empirical studies in this area we
mostly have to make do with retrospective accounts from those already
in a leader position, cross-sectional data, small (and selected) samples,
secondary data, or data collected at a later stage and from other sources
than our primary study objects.
With access from the Project Metropolitan we were able to remedy
some of these shortcomings in studies of leadership emergence. The
main assets of the data set from the project was its comprehensiveness
in terms of incorporating a whole cohort in a large urban area and in
terms of the breadth of data sources and topics covered. Of particular
importance is the school survey performed when the respondents were
in the 6th grade. The cohort was followed through early adulthood into
their thirties. After the dataset was anonymized in 1986, no more
surveys could be performed but it has been possible to follow the cohort
through registers as they now move into the age of retirement
(Stenberg, 2018).
Our study only includes the male members of the cohort and only
those who were drafted (N = 5928). With data on birth order and fa-
mily socioeconomic status when they were 10, through an extensive
survey when they were in the 6th grade, and on to circumstances re-
lated to the draft, we cover the cohort members' first 20 or so years.
This is a fairly long period as such but is also a particularly formative
period in life. By following the cohort prospectively, we were also able
to avoid departing from study populations already selected for the
outcome we are actually interested in. Having said that, this was not an
investigation of formal leadership or the practice of leadership, but the
assessment or ascription of leadership at two points in the cohort's lives.
Overall, we found that early life experiences in terms of family so-
cioeconomic status, parental support, and birth order were all im-
portant for leadership emergence at both points in life, although
somewhat differently. Interestingly family socioeconomic status did not
affect leadership ascription in childhood, but was all the more sig-
nificant in early adulthood – even when controlling for a number of
other factors, including IQ.
Cognitive ability has been identified as a crucial factor in leadership
emergence, and this study is no exception. Still, cognitive ability is but
one important factor related to leadership emergence and it might even
have its limits. The positive effect of IQ on leadership assessment at the
draft did show a curvilinear pattern. Leadership ascription is contextual
and the children seem to appreciate the younger sibling for organizing
class parties while the draft board is looking for the responsible elder
brother.
As so much social and behavioral research has reported, cognitive
ability and social class showed very strong influences on outcomes both
in childhood and later in early adulthood. And this was also the case in
a society with relatively small class differences. The lesson for leader-
ship studies may be that including childhood factors is relevant not only
for fostering (good) leaders, but also for understanding followers (Elgar,
2016). That is, to observe and highlight not only how some children are
chosen, or choose themselves, as leaders – but also to encourage re-
flection about what qualifications are made relevant and which in fact
are relevant.
For those interested in leadership emergence, the results offer sup-
port for including childhood factors in models attempting to understand
how leaders are formed. For those engaged in fostering leaders they will
find support for encouraging youngsters to engage in sports, scouts, and
other extracurricular activities. Based on the Project Metropolitan data
Kassman (2017) found that taking part in youth organizations and
specifically the scouts did had a positive impact on school marks and
future income. Scouts showed signs of more positive trajectories even
when controlling for differences in their social background and per-
sonal cognitive abilities.
316
The Leadership Quarterly 30 (2019) 298–319
Choosing a leader on a desert island is a question of life and death.
The exit option is not available and the children must act upon limited
knowledge. The choice is not free either (no anonymous votes) but part
of an intricate group formation. It is important to question and discuss
claims to leadership in different situations. Is singing a high C a valid
competence on a desert island? Is physical strength more valuable than
intellectual reasoning? For Piggy, cognitive ability was not a sufficient
asset for leadership – he had too many deficits in other areas – but he
was an important source of knowledge and advanced solutions to dif-
ferent problems. He kept count of the children and tried to uphold order
of speech and social rules. Until he dies in an accident. When the res-
cuing officer asks how many they are on the island, Ralph is not able to
answer.
‘I should have thought,’ said the officer as he visualized the search
before him. ‘I should have though that a pack of British boys – you're
all British aren't you? – would have been able to put up a better
show than that – I mean – ’
‘It was like that at first,’ said Ralph, ‘before things – ’
He stopped.
(Lord of the Flies, chapter 12)
A true life cycle perspective should also take on the practice of
leadership outside work hours and even after working age. We do not
know how many of the Project Metropolitan boys actually ended up as
military officers or in other leadership roles. A true life-cycle perspec-
tive on leadership emergence should of course include the entirety of
adulthood. For Ralph and Jack the events on the desert island might
have put them off any sort of leadership position in the future.
However, in our view, it is highly likely that they do emerge as leaders
in some form or another – possibly as naval officers. Apart from the
assets that made them emerge as leaders in the first place, they can now
also point to previous leadership experience.
References
Allen, J. P., Porter, M. R., McFarland, F. C., Marsh, P., & McElhaney, K. B. (2005). The two
faces of adolescents' success with peers: Adolescent popularity, social adaptation, and
deviant behavior. Child Development, 76(3), 747–760. https://doi.org/10.1111/j.
1467-8624.2005.00875.x.
Almquist, Y. (2009). Peer status in school and adult disease risk: A 30-year follow-up
study of disease-specific morbidity in a Stockholm cohort. Journal of Epidemiology and
Community Health, 63(12), 1028–1034. https://doi.org/10.1136/jech.2009.088377.
Almquist, Y. (2011). Social isolation in the classroom and adult health: A longitudinal
study of a 1953 cohort. Advances in Life Course Research, 16(1), 1–12. https://doi.org/
10.1016/j.alcr.2010.11.001.
Almquist, Y., Modin, B., & Östberg, V. (2010). Childhood social status in society and
school: Implications for the transition to higher levels of education. British Journal of
Sociology of Education, 31(1), 31–45. https://doi.org/10.1080/01425690903385352.
Almquist, Y. B., & Östberg, V. (2013). Social relationships and subsequent health-related
behaviours: Linkages between adolescent peer status and levels of adult smoking in a
Stockholm cohort. Addiction, 108(3), 629–637. https://doi.org/10.1111/j.1360-
0443.2012.04097.x.
Andeweg, R. B., & Van den Berg, S. B. (2003). Linking birth order to political leadership:
The impact of parents or sibling interaction? Political Psychology, 24(3), 605–623.
https://doi.org/10.1111/0162-895X.00343.
Antonakis, J., Day, D. V., & Schyns, B. (2012). Leadership and individual differences: At
the cusp of a renaissance. The Leadership Quarterly, 23(4), 643–650. https://doi.org/
10.1016/j.leaqua.2012.05.002.
Antonakis, J., House, R. J., & Simonton, D. K. (2017). Can super smart leaders suffer from
too much of a good thing? The curvilinear effect of intelligence on perceived lea-
dership behavior. Journal of Applied Psychology, 102(7), 1003. https://doi.org/10.
1037/apl0000221.
Arvey, R. D., Rotundo, M., Johnson, W., Zhang, Z., & McGue, M. (2006). The determi-
nants of leadership role occupancy: Genetic and personality factors. The Leadership
Quarterly, 17(1), 1–20. https://doi.org/10.1016/j.leaqua.2005.10.009.
Baltes, P. B., Reese, H. W., & Lipsitt, L. P. (1980). Life-span developmental psychology.
Annual Review of Psychology, 31(1), 65–110. https://doi.org/10.1146/annurev.ps.31.
020180.000433.
Bamberger, P. A., & Pratt, M. G. (2010). Moving forward by looking back: Reclaiming
unconventional research contexts and samples in organizational scholarship.
Academy of Management Journal, 53(4), 665–671. https://doi.org/10.5465/AMJ.
2010.52814357.
Banaji, M. R., & Prentice, D. A. (1994). The self in social contexts. Annual Review of
Psychology, 45(1), 297–332. https://doi.org/10.1146/annurev.ps.45.020194.
001501.
T. Reitan, S.-Å. Stenberg
Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory. Journal
of Social and Clinical Psychology, 4(3), 359–373. https://doi.org/10.1521/jscp.1986.
4.3.359.
Barling, J., & Weatherhead, J. G. (2016). Persistent exposure to poverty during childhood
limits later leader emergence. Journal of Applied Psychology, 101(9), 1305–1318.
https://doi.org/10.1037/apl0000129.
Bartolini, S. (1993). On time and comparative research. Journal of Theoretical Politics,
5(2), 131–167. https://doi.org/10.1177/0951692893005002001.
Baumeister, R. F., & Vohs, K. D. (2003). Self-regulation and the executive function of the
self. Handbook of Self and Identity, 1, 197–217.
Belsky, J., Steinberg, L., & Draper, P. (1991). Childhood experience, interpersonal de-
velopment, and reproductive strategy: An evolutionary theory of socialization. Child
Development, 62(4), 647–670.
Blaker, N. M., Rompa, I., Dessing, I. H., Vriend, A. F., Herschberg, C., & van Vugt, M.
(2013). The height leadership advantage in men and women: Testing evolutionary
psychology predictions about the perceptions of tall leaders. Group Processes &
Intergroup Relations, 16(1), 17–27. https://doi.org/10.1177/1368430212437211.
Brungardt, C. (1997). The making of leaders: A review of the research in leadership de-
velopment and education. Journal of Leadership Studies, 3(3), 81–95. https://doi.org/
10.1177/107179199700300309.
Burks, S. V., Carpenter, J. P., Goette, L., & Rustichini, A. (2009). Cognitive skills affect
economic preferences, strategic behavior, and job attachment. Proceedings of the
National Academy of Sciences, 106(19), 7745.
Castillo, E. A., & Trinh, M. P. (2018). In search of missing time: A review of the study of
time in leadership research. The Leadership Quarterly, 29, 165–178. https://doi.org/
10.1016/j.leaqua.2017.12.001.
Central Conscription Authority (1952). Instruktion för exploratörer vid
inskrivningsförrätningar [instructions for evaluators in the assessment of military con-
scripts]. Sweden: Lund.
Chaturvedi, S., Zyphur, M. J., Arvey, R. D., Avolio, B. J., & Larsson, G. (2012). The
heritability of emergent leadership: Age and gender as moderating factors. The
Leadership Quarterly, 23(2), 219–232. https://doi.org/10.1016/j.leaqua.2011.08.004.
Chiao, J. Y., Bowman, N. E., & Gill, H. (2008). The political gender gap: Gender bias in
facial inferences that predict voting behavior. PLoS One, 3(10), e3666. https://doi.
org/10.1371/journal.pone.0003666.
Cragg, J. D. (1971). Some statistical models for limited dependent variables with appli-
cation to the demand for durable goods. Econometrica, 39(5), 829–844.
Cunha, M. P.e., Lewis, M., Rego, A., & Smith, W. K. (2017). Biographical methods in
leadership research. In B. Schyns, R. J. Hall, & P. Neves (Eds.). Handbook of methods in
leadership research (pp. 372–402). Cheltenham, UK: Edward Elgar Publishing.
Daly, M., Egan, M., & O'Reilly, F. (2015). Childhood general cognitive ability predicts
leadership role occupancy across life: Evidence from 17,000 cohort study partici-
pants. The Leadership Quarterly, 26(3), 323–341. https://doi.org/10.1016/j.leaqua.
2015.03.006.
Day, D. V. (2011). Integrative perspectives on longitudinal investigations of leader de-
velopment: From childhood through adulthood. The Leadership Quarterly, 22(3),
561–571. https://doi.org/10.1016/j.leaqua.2011.04.012.
Day, D. V., & Sin, H.-P. (2011). Longitudinal tests of an integrative model of leader de-
velopment: Charting and understanding developmental trajectories. The Leadership
Quarterly, 22(3), 545–560. https://doi.org/10.1016/j.leaqua.2011.04.011.
Day, D. V., Fleenor, J. W., Atwater, L. E., Sturm, R. E., & McKee, R. A. (2014). Advances in
leader and leadership development: A review of 25 years of research and theory. The
Leadership Quarterly, 25(1), 63–82. https://doi.org/10.1016/j.leaqua.2013.11.004.
Day, D. V., & Harrison, M. M. (2007). A multilevel, identity-based approach to leadership
development. Human Resource Management Review, 17(4), 360–373.
Day, D. V., Harrison, M. M., & Halpin, S. M. (2008). An integrative approach to leader
development: Connecting adult development, identity, and expertise. New York:
Psychology Press.
Deary, I. J., Weiss, A., & Batty, G. D. (2010). Intelligence and personality as predictors of
illness and death: How researchers in differential psychology and chronic disease
epidemiology are collaborating to understand and address health inequalities.
Psychological Science in the Public Interest, 11(2), 53–79. https://doi.org/10.1177/
1529100610387081.
Dinh, J. E., & Lord, R. G. (2012). Implications of dispositional and process views of traits
for individual difference research in leadership. The Leadership Quarterly, 23(4),
651–669. https://doi.org/10.1016/j.leaqua.2012.03.003.
Dixon, J., Horton, S., & Weir, P. (2011). Relative age effects: Implications for leadership
development. The International Journal of Sport and Society, 2(2)https://scholar.
uwindsor.ca/humankineticspub/6.
Dobosz, R. P., & Beaty, L. A. (1999). The relationship between athletic participation and
high school student's leadership ability. Adolescence, 34(133), 215–220.
Dodge, K. A., Dishion, T. J., & Lansford, J. E. (2006). Deviant peer influences in inter-
vention and public policy for youth. Social Policy Report, 20(1).
Dohmen, T., Falk, A., Huffman, D., & Sunde, U. (2010). Are risk aversion and impatience
related to cognitive ability? American Economic Review, 100(3), 1238–1260. https://
doi.org/10.1257/aer.100.3.1238.
Duster, T. (2006). Comparative perspectives and competing explanations: Taking on the
newly configured reductionist challenge to sociology. American Sociological Review,
71(1), 1–15 (10.1177%2F000312240607100101).
Eckstein, D., Aycock, K. J., Sperber, M. A., McDonald, J., Van Wiesner, V., III, Watts, R. E.,
& Ginsburg, P. (2010). A review of 200 birth-order studies: Lifestyle characteristics.
Journal of Individual Psychology, 66(4), 408–434.
Elder, G. H. (1974). Children of the great depression. Chicago: University of Chicago Press.
Elgar, M. A. (2016). Leader selection and leadership outcomes: Height and age in a
sporting model. The Leadership Quarterly, 27(4), 588–601. https://doi.org/10.1016/j.
leaqua.2015.12.005.
317
The Leadership Quarterly 30 (2019) 298–319
Elsner, B., & Isphording, I. E. (2018). Rank, sex, drugs, and crime. Journal of Human
Resources, 53(2), 356–381.
Fergusson, D. M., Horwood, L. J., & Ridder, E. M. (2005). Show me the child at seven II:
Childhood intelligence and later outcomes in adolescence and young adulthood.
Journal of Child Psychology and Psychiatry, 46(8), 850–858. https://doi.org/10.1111/
j.1469-7610.2005.01472.x.
Fiedler, F. E. (2002). The curious role of cognitive resources in leadership. In R. E. Riggio,
S. E. Muprhy, & F. J. Pirozzolo (Eds.). Multiple intelligences and leadership (pp. 91–
104). Mahwah, NJ: Lawrence Erlbaum Associates.
Fischbein, S. (1980). IQ and social class. Intelligence, 4(1), 51–63.
French, D. C., & Stright, A. L. (1991). Emergent leadership in Children's small groups.
Small Group Research, 22(2), 187–199. https://doi.org/10.1177/
1046496491222003.
Fukada, S., Fukada, H., & Hicks, J. (1994). Structure of leadership among preschool
children. The Journal of Genetic Psychology, 155(4), 389–395. https://doi.org/10.
1080/00221325.1994.9914789.
Gayles, J. G., & Baker, A. R. (2015). Opportunities and challenges for first-year student-
athletes transitioning from high school to college. New Directions for Student
Leadership, 2015(147), 43–51. https://doi.org/10.1002/yd.20142.
Gest, S. D., Graham-Bermann, S. A., & Hartup, W. W. (2002). Peer experience: Common
and unique features of number of friendships, social network centrality, and socio-
metric status. Social Development, 10(1), 23–40. https://doi.org/10.1111/1467-9507.
00146.
Glenn, N. D. (1976). Cohort analysts' futile quest: Statistical attempts to separate age,
period and cohort effects. American Sociological Review, 41(5), 900–904. https://doi.
org/10.2307/2094738.
Gollner, L. M., Ballhausen, N., Kliegel, M., & Forstmeier, S. (2017). Delay of gratification,
delay discounting and their associations with age, episodic future thinking, and fu-
ture time perspective. Frontiers in Psychology, 8, 2304. https://doi.org/10.3389/
fpsyg.2017.02304.
Gottfried, A. E., Gottfried, A. W., Reichard, R. J., Guerin, D. W., Oliver, P. H., & Riggio, R.
E. (2011). Motivational roots of leadership: A longitudinal study from childhood
through adulthood. The Leadership Quarterly, 22(3), 510–519. https://doi.org/10.
1016/j.leaqua.2011.04.008.
Gottfried, A. W., Gottfried, A. E., & Guerin, D. W. (2006). The Fullerton longitudinal
study: A long-term investigation of intellectual and motivational giftedness. Journal
for the Education of the Gifted, 29(4), 430–450.
Guerin, D. W., Oliver, P. H., Gottfried, A. W., Gottfried, A. E., Reichard, R. J., & Riggio, R.
E. (2011). Childhood and adolescent antecedents of social skills and leadership po-
tential in adulthood: Temperamental approach/withdrawal and extraversion. The
Leadership Quarterly, 22(3), 482–494. https://doi.org/10.1016/j.leaqua.2011.04.006.
Halaby, C. N. (2004). Panel models in sociological research: Theory into practice. Annual
Review of Sociology, 30, 507–544.
Hannah, S. T., Avolio, B. J., Luthans, F., & Harms, P. D. (2008). Leadership efficacy:
Review and future directions. The Leadership Quarterly, 19(6), 669–692. https://doi.
org/10.1016/j.leaqua.2008.09.007.
Hernández-Julián, R., & Peters, C. (2018). Physical appearance and peer effects in aca-
demic performance. Applied Economics Letters, 25(13), 887–890. https://doi.org/10.
1080/13504851.2017.1380282.
Hogan, R., & Kaiser, R. B. (2005). What we know about leadership. Review of General
Psychology, 9(2), 169–180. https://doi.org/10.1037/1089-2680.9.2.169.
Horrocks, J. E., & Thompson, G. G. (1946). A study of the friendship fluctuations of rural
boys and girls. Journal of Genetic Psychology, 69(2), 189–198. https://doi.org/10.
1080/08856559.1946.10533388.
Hoyt, M. A., & Kennedy, C. L. (2008). Leadership and adolescent girls: A qualitative study
of leadership development. American Journal of Community Psychology, 42(3),
203–219. https://doi.org/10.1007/s10464-008-9206-8.
Ilies, R., Arvey, R. D., & Bouchard, T. J. (2006). Darwinism, behavioral genetics, and
organizational behavior: A review and agenda for future research. Journal of
Organizational Behavior, 27(2), 121–141. https://doi.org/10.1002/job.351.
Janson, C.-G. (1966). Project Metropolitan. Acta Sociologica, 9(1–2), 110–115.
Jencks, C. (1980). Heredity, environment, and public policy reconsidered. American
Sociological Review, 45(5), 723–736. https://doi.org/10.2307/2094892.
Judge, T. A., & Cable, D. M. (2004). The effect of physical height on workplace success
and income: Preliminary test of a theoretical model. Journal of Applied Psychology,
89(3), 428.
Judge, T. A., Ilies, R., & Colbert, A. E. (2004). Intelligence and leadership: A quantitative
review and test of theoretical propositions. Journal of Applied Psychology, 89(3),
542–552.
Karagianni, D., & Jude Montgomery, A. (2018). Developing leadership skills among
adolescents and young adults: A review of leadership programmes. International
Journal of Adolescence and Youth, 23(1), 86–98. https://doi.org/10.1080/02673843.
2017.1292928.
Karnehed, N., Rasmussen, F., Hemmingsson, T., & Tynelius, P. (2012). Obesity and at-
tained education: Cohort study of more than 700,000 Swedish men. Obesity, 14(8),
1421–1428. https://doi.org/10.1038/oby.2006.161.
Kassman, A. (2017). One of all the others – A life course study of scouts, social capital,
and stratification in the Swedish ‘Folkhem’. Journal of Civil Society, 13(1), 71–89.
https://doi.org/10.1080/17448689.2017.1287843.
Klepp, K., Halper, A., & Perry, C. L. (2018). The efficacy of peer leaders in drug abuse
prevention. Journal of School Health, 56(9), 407–411. https://doi.org/10.1111/j.
1746-1561.1986.tb05783.x.
Kuhn, P., & Weinberger, C. (2005). Leadership skills and wages. Journal of Labor
Economics, 23(3), 395–436. https://doi.org/10.1086/430282.
Lease, A. M., Musgrove, K. T., & Axelrod, J. L. (2002). Dimensions of social status in
preadolescent peer groups: Likability, perceived popularity, and social dominance.
T. Reitan, S.-Å. Stenberg
Social Development, 11(4), 508–533.
Legree, P. J., Gade, P. A., Martin, D. E., Fischl, M. A., Wilson, M. J., Nieva, V. F., ...
Laurence, J. (2000). Military enlistment and family dynamics: Youth and parental
perspectives. Military Psychology, 12(1), 31–49. https://doi.org/10.1207/
S15327876MP1201_2.
Lerner, R. M., & Steinberg, L. (2009). Handbook of adolescent psychology, volume 1:
Individual bases of adolescent development. vol. 1. Hoboken, N.J.: John Wiley & Sons.
Li, W.-D., Arvey, R. D., & Song, Z. (2011). The influence of general mental ability, self-
esteem and family socioeconomic status on leadership role occupancy and leader
advancement: The moderating role of gender. The Leadership Quarterly, 22(3),
520–534. https://doi.org/10.1016/j.leaqua.2011.04.009.
Lindqvist, E. (2011). Height and leadership. The Review of Economics and Statistics, 94(4),
1191–1196. https://doi.org/10.1162/REST_a_00239.
Lindqvist, E., & Vestman, R. (2011). The labor market returns to cognitive and non-
cognitive ability: Evidence from the Swedish enlistment. American Economic Journal:
Applied Economics, 3(1), 101–128.
Little, A. C. (2014). Facial appearance and leader choice in different contexts: Evidence
for task contingent selection based on implicit and learned face-behaviour/face-
ability associations. The Leadership Quarterly, 25(5), 865–874. https://doi.org/10.
1016/j.leaqua.2014.04.002.
Lord, R. G., De Vader, C. L., & Alliger, G. M. (1986). A meta-analysis of the relation
between personality traits and leadership perceptions: An application of validity
generalization procedures. Journal of Applied Psychology, 71(3), 402–410.
Lord, R. G., & Hall, R. J. (2005). Identity, deep structure and the development of lea-
dership skill. The Leadership Quarterly, 16(4), 591–615. https://doi.org/10.1016/j.
leaqua.2005.06.003.
Lundborg, P., Nystedt, P., & Rooth, D.-O. (2009). The height premium in earnings: The role of
physical capacity and cognitive and non-cognitive skills. IZA discussion papers 4266.
Berlin: Institute for the Study of Labor.
Lundin, A., Lundberg, I., Hallsten, L., Ottosson, J., & Hemmingsson, T. (2010).
Unemployment and mortality. A longitudinal prospective study on selection and
causation in 49321 Swedish middle-aged men. Journal of Epidemiology & Community
Health, 64(01), 22–28.
Martin, S. R., Innis, B. D., & Ward, R. G. (2017). Social class, leaders and leadership: A
critical review and suggestions for development. Current Opinion in Psychology, 18,
49–54. https://doi.org/10.1016/j.copsyc.2017.08.001.
Mawson, B. (2011). Children's leadership strategies in early childhood. Journal of
Research in Childhood Education, 25(4), 327–338. https://doi.org/10.1080/
02568543.2011.605207.
McCullough, P. M., Ashbridge, D., & Pegg, R. (1994). The effect of self-esteem, family
structure, locus of control, and career goals on adolescent leadership behavior.
Adolescence, 29(115), 605–611.
McFarland, D. A., & Thomas, R. J. (2006). Bowling young: How youth voluntary asso-
ciations influence adult political participation. American Sociological Review, 71(3),
401–425. https://doi.org/10.1177/000312240607100303.
McNeal, R. B. (1998). High school extracurricular activities: Closed structures and stra-
tifying patterns of participation. The Journal of Educational Research, 91(3), 183–191.
https://doi.org/10.1080/00220679809597539.
Miller, N., & Maruyama, G. (1976). Ordinal position and peer popularity. Journal of
Personality and Social Psychology, 33(2), 123–131. https://doi.org/10.1037/0022-
3514.33.2.123.
Mischel, W., & Ebbesen, E. B. (1970). Attention in delay of gratification. Journal of
Personality and Social Psychology, 16(2), 329–337. https://doi.org/10.1037/
h0029815.
Moffitt, T. E. (2015). Life-course-persistent versus adolescence-limited antisocial beha-
vior. In D. Cicchetti, & D. J. Cohen (Eds.). Developmental psychopathology (pp. 570–
598). Hoboken, N.J.: John Wiley & Sons. https://doi.org/10.1002/9780470939406.
ch15.
Morgan, G. A., Harmon, R. J., & Maslin-Cole, C. A. (1990). Mastery motivation: Definition
and measurement. Early Education and Development, 1(5), 318–339.
Mueller, U., & Mazur, A. (1996). Facial dominance of West Point cadets as a predictor of
later military rank. Social Forces, 74(3), 823–850. https://doi.org/10.2307/2580383.
Murphy, S. E. (2002). Leader self-regulation: The role of self-efficacy and multiple in-
telligences. In R. E. Riggio, S. E. Murphy, & F. J. Pirozzolo (Eds.). LEA's organization
and management series. Multiple intelligences and leadership (pp. 163–186). Mahwah,
NJ: Lawrence Erlbaum Associates.
Murphy, S. E. (2011). Providing a foundation for leadership development. In S. E.
Murphy, & R. J. Reichard (Eds.). Early development and leadership: Building the next
generation of leaders (pp. 3–37). New York: Psychology Press/Routledge.
Murphy, S. E., & Johnson, S. K. (2011). The benefits of a long-lens approach to leader
development: Understanding the seeds of leadership. The Leadership Quarterly, 22(3),
459–470. https://doi.org/10.1016/j.leaqua.2011.04.004.
National Board of Health and Welfare (2009). Folkhälsorapport 2009 [Public health report
2009]. Stockholm: Socialstyrelsen.
Obschonka, M. (2016). Adolescent pathways to entrepreneurship. Child Development
Perspectives, 10(3), 196–201. https://doi.org/10.1111/cdep.12185.
Östberg, J. (2018). We want you as our new recruit. Prerequisites for recruitment to and
retention in the Swedish Armed Forces. PhD thesisSweden: University of Karlstad.
Östberg, V. (2003). Children in classrooms: Peer status, status distribution and mental
well-being. Social Science & Medicine, 56(1), 17–29. https://doi.org/10.1016/S0277-
9536(02)00006-0.
Östberg, V., & Modin, B. (2008). Status relations in school and their relevance for health
in a life course perspective: Findings from the Aberdeen children of the 1950's cohort
study. Social Science & Medicine, 66(4), 835–848. https://doi.org/10.1016/j.
socscimed.2007.10.018.
Palmore, E. (1978). When can age, period, and cohort be separated? Social Forces, 57(1),
318
The Leadership Quarterly 30 (2019) 298–319
282–295. https://doi.org/10.1093/sf/57.1.282.
Paunova, M. (2015). The emergence of individual and collective leadership in task
groups: A matter of achievement and ascription. The Leadership Quarterly, 26(6),
935–957. https://doi.org/10.1016/j.leaqua.2015.10.002.
Public Committee for the Education of Conscripts (1986). SOU 1986:43 Befrielse från
värnpliktstjänstgöring [Exemption from military service, Swedish Public Investigation].
Stockholm: Ministry of Defence.
Rauthmann, J. F., Sherman, R. A., & Funder, D. C. (2015). Principles of situation research:
Towards a better understanding of psychological situations. European Journal of
Personality, 29(3), 363–381. https://doi.org/10.1002/per.1994.
Reichard, R. J., Riggio, R. E., Guerin, D. W., Oliver, P. H., Gottfried, A. W., & Gottfried, A.
E. (2011). A longitudinal analysis of relationships between adolescent personality
and intelligence with adult leader emergence and transformational leadership. The
Leadership Quarterly, 22(3), 471–481. https://doi.org/10.1016/j.leaqua.2011.04.005.
Reid, S. W., Anglin, A. H., Baur, J. E., Short, J. C., & Buckley, M. R. (2018). Blazing new
trails or opportunity lost? Evaluating research at the intersection of leadership and
entrepreneurship. The Leadership Quarterly, 29(1), 150–164. https://doi.org/10.
1016/j.leaqua.2017.11.005.
Ridgeway, C. L. (2002). Gender, status, and leadership. Journal of Social Issues, 57(4),
637–655. https://doi.org/10.1111/0022-4537.00233.
Rosenberg, E. B., & Hajal, F. (1985). Stepsibling relationships in remarried families. Social
Casework, 66(5), 287–292. https://doi.org/10.1177/104438948506600504.
Rosette, A. S., Leonardelli, G. J., & Phillips, K. W. (2008). The white standard: Racial bias
in leader categorization. Journal of Applied Psychology, 93(4), 758.
Ruhm, C. (1992). Youth leadership training in curricular, cocurricular, and community pro-
grams. Doctoral dissertationThe University of Wisconsin-Madison.
Sanchez-Cortes, D., Aran, O., Jayagopi, D. B., Schmid Mast, M., & Gatica-Perez, D. (2013).
Emergent leaders through looking and speaking: From audio-visual data to multi-
modal recognition. Journal on Multimodal User Interfaces, 7(1), 39–53. https://doi.
org/10.1007/s12193-012-0101-0.
Schoon, I., & Parsons, S. (2002). Teenage aspirations for future careers and occupational
outcomes. Journal of Vocational Behavior, 60(2), 262–288. https://doi.org/10.1006/
jvbe.2001.1867.
Sczesny, S., & Kühnen, U. (2004). Meta-cognition about biological sex and gender-ste-
reotypic physical appearance: Consequences for the assessment of leadership com-
petence. Personality and Social Psychology Bulletin, 30(1), 13–21. https://doi.org/10.
1177/0146167203258831.
Shoup, J. R. (2005). A collective biography of twelve world-class leaders: A study on devel-
oping exemplary leaders. Lanham, Maryland: University Press of America.
Sidorchuk, A., Hemmingsson, T., Romelsjö, A., & Allebeck, P. (2012). Alcohol use in
adolescence and risk of disability pension: A 39 year follow-up of a population-based
conscription survey. PLoS One, 7(8), e42083. https://doi.org/10.16993/bav.
Simonton, D. K. (1985). Intelligence and personal influence in groups: Four nonlinear
models. Psychological Review, 92(4), 532–547.
Solansky, S. T. (2010). The evaluation of two key leadership development program
components: Leadership skills assessment and leadership mentoring. The Leadership
Quarterly, 21(4), 675–681. https://doi.org/10.1016/j.leaqua.2010.06.009.
Stanner, S. A., & Yudkin, J. S. (2001). Fetal programming and the Leningrad siege study.
Twin Research, 4(5), 287–292. https://doi.org/10.1375/twin.4.5.287.
Stenberg, S.-Å. (2018). Born in 1953. The story about a post-war Swedish cohort, and a
longitudinal research project. Stockholm: Stockholm University Press.
Stenberg, S.-Å., & Vågerö, D. (2006). Cohort profile: The Stockholm birth cohort of 1953.
International Journal of Epidemiology, 35(3), 546–548. https://doi.org/10.1093/ije/
dyi310.
Stockard, J., & McGee, J. (1990). Children's occupational preferences: The influence of
sex and perceptions of occupational characteristics. Journal of Vocational Behavior,
36(3), 287–303. https://doi.org/10.1016/0001-8791(90)90033-X.
Stogdill, R. M. (1948). Personal factors associated with leadership: A survey of the lit-
erature. The Journal of Psychology, 25(1), 35–71. https://doi.org/10.1080/00223980.
1948.9917362.
Strand, S., & Berndtsson, J. (2015). Recruiting the “enterprising soldier”: Military re-
cruitment discourses in Sweden and the United Kingdom. Critical Military Studies,
1(3), 233–248. https://doi.org/10.1080/23337486.2015.1090676.
Strang, S. E., & Kuhnert, K. W. (2009). Personality and leadership developmental levels as
predictors of leader performance. The Leadership Quarterly, 20(3), 421–433. https://
doi.org/10.1016/j.leaqua.2009.03.009.
Stulp, G., Buunk, A. P., Verhulst, S., & Pollet, T. V. (2013). Tall claims? Sense and non-
sense about the importance of height of US presidents. The Leadership Quarterly,
24(1), 159–171. https://doi.org/10.1016/j.leaqua.2012.09.002.
Sundevall, F. (2017). Military education for non-military purposes. History of Education
Review, 46(1), 58–71. https://doi.org/10.1108/HER-05-2016-0024.
Todorov, A., Mandisodza, A. N., Goren, A., & Hall, C. C. (2005). Inferences of competence
from faces predict election outcomes. Science, 308(5728), 1623–1626.
Tuncdogan, A., Acar, O. A., & Stam, D. (2017). Individual differences as antecedents of
leader behavior: Towards an understanding of multi-level outcomes. The Leadership
Quarterly, 28(1), 40–64. https://doi.org/10.1016/j.leaqua.2016.10.011.
Vannatta, K., Gartstein, M. A., Zeller, M., & Noll, R. B. (2009). Peer acceptance and social
behavior during childhood and adolescence: How important are appearance, athle-
ticism, and academic competence? International Journal of Behavioral Development,
33(4), 303–311. https://doi.org/10.1177/0165025408101275.
von Hippel, P. T. (2018). How many imputations do you need? A two-stage calculation
using a quadratic rule. Sociological Methods & Research, 1–20. https://doi.org/10.
1177/0049124117747303.
Wahl, K. H., & Blackhurst, A. (2000). Factors affecting the occupational and educational
aspirations of children and adolescents. Professional School Counseling, 3(5), 367–374.
Watts, T. W., Duncan, G. J., & Quan, H. (2018). Revisiting the marshmallow test: A
T. Reitan, S.-Å. Stenberg
conceptual replication investigating links between early delay of gratification and
later outcomes. Psychological Science, 29(7), 1159–1177. https://doi.org/10.1177/
0956797618761661.
Whalley, L. J., & Deary, I. J. (2001). Longitudinal cohort study of childhood IQ and
survival up to age 76. BMJ, 322(7290), 819.
Whitehead, G. (2009). Adolescent leadership development: Building a case for an au-
thenticity framework. Educational Management Administration & Leadership, 37(6),
847–872. https://doi.org/10.1177/1741143209345441.
Yang, Y., & Land, K. C. (2016). Age-period-cohort analysis: New models, methods, and em-
pirical applications. New York: Chapman and Hall/CRC.
Zaccaro, S. J., Green, J. P., Dubrow, S., & Kolze, M. (2018). Leader individual differences,
situational parameters, and leadership outcomes: A comprehensive review and
319
The Leadership Quarterly 30 (2019) 298–319
integration. The Leadership Quarterly, 29(1), 2–43. https://doi.org/10.1016/j.leaqua.
2017.10.003.
Zacharatos, A., Barling, J., & Kelloway, E. K. (2000). Development and effects of trans-
formational leadership in adolescents. The Leadership Quarterly, 11(2), 211–226.
https://doi.org/10.1016/S1048-9843(00)00041-2.
Zhang, L., Welte, J. W., & Wieczorek, W. F. (2002). Underlying common factors of ado-
lescent problem behaviors. Criminal Justice and Behavior, 29(2), 161–182. https://doi.
org/10.1177/0093854802029002003.
Zhang, Z., Ilies, R., & Arvey, R. D. (2009). Beyond genetic explanations for leadership: The
moderating role of the social environment. Organizational Behavior and Human
Decision Processes, 110(2), 118–128. https://doi.org/10.1016/j.obhdp.2009.06.004.