Do international institutions matter? Socialization and international bureaucrats
| Author | Sara Connolly,Zuzana Murdoch,Hussein Kassim,Benny Geys |
| DOI | 10.1177/1354066118809156 |
| Published date | 01 September 2019 |
| Date | 01 September 2019 |
research-article2018
EJ R
I
Article
European Journal of
International Relations
Do international institutions
2019, Vol. 25(3) 852 –877
© The Author(s) 2018
matter? Socialization and
Article reuse guidelines:
sagepub.com/journals-permissions
https://doi.org/10.1177/1354066118809156
DOI: 10.1177/1354066118809156
international bureaucrats
journals.sagepub.com/home/ejt
Zuzana Murdoch
University of Bergen, Norway
Hussein Kassim
University of East Anglia, UK
Sara Connolly
University of East Anglia, UK
Benny Geys
Norwegian Business School (BI), Norway
Abstract
A key component of (neo-)functionalist and constructivist approaches to the study
of international organizations concerns staff socialization. Existing analyses of how,
or indeed whether, staff develop more pro-internationalist attitudes over time draw
predominantly on cross-sectional data. Yet, such data cannot address (self-)selection
issues or capture the inherently temporal nature of attitude change. This article
proposes an innovative approach to the study of international socialization using an
explicitly longitudinal design. Analysing two waves of a large-scale survey conducted
within the European Commission in 2008 and 2014, it examines the beliefs and values
of the same individuals over time and exploits exogenous organizational changes to
identify causal effects. Furthermore, the article theorizes and assesses specified scope
conditions affecting socialization processes. Showing that international institutions do,
in fact, influence value acquisition by individual bureaucrats, our results contest the
widely held view that international organizations are not a socializing environment. Our
Corresponding author:
Zuzana Murdoch, Department of Administration and Organization Theory, University of Bergen,
Christiesgate 19, NO-5007 Bergen, Norway.
Email: zuzana.murdoch@uib.no
Murdoch et al.
853
analysis also demonstrates that age at entry and gender significantly affect the intensity
of such value change.
Keywords
Civil servants, European Commission, international bureaucrats, international
organizations, socialization
Introduction
International Relations (IR) scholars studying intergovernmental organizations (IGOs)
have mainly been preoccupied with the conditions underlying the emergence, prolifera-
tion and autonomy of those institutions.1 Still, how, and indeed whether, IGOs shape the
values of the people who work for them is no less important. As functionalist scholars
including Ernst B. Haas and James Sewell first asserted, the extent to which individuals
acquire internationalist attitudes while working for an IGO contributes to that organiza-
tion’s capacity to establish its independence from member state principals (Checkel,
2005; Haas, 1964; Sewell, 1966; Wolf, 1973). The experience, expertise and values of
IGO staff are therefore fundamental, and identifying the conditions under which indi-
viduals acquire internationalist values, as well as the factors that affect their acquisition,
is an important undertaking with significant implications.2
Previous empirical studies of value acquisition by IGO staff have provided decidedly
mixed results (see, e.g., Gheciu, 2005; Lewis, 2005; versus Beyers, 2005; Hooghe,
2005).3 Yet, if the values of personnel affect the capacity of IGOs to fulfil their missions,
the absence of a clear understanding of value acquisition is a serious gap. Drawing on
individual-level data and employing a longitudinal research design appropriate to the
analysis of socialization as a diachronic process, this article revisits the socializing power
of IGOs and aims to advance its understanding. It also takes up the challenge of testing
and refining the scope conditions under which IGOs can be expected to trigger socializa-
tion effects. Scope conditions are singled out since they provide an opportunity to better
understand the often weak and ambiguous socialization effects observed in previous
studies (Johnston, 2005; Wolf, 1973; Zürn and Checkel, 2005). Our empirical analysis
thereby takes the European Commission as a critical case.4 Matching information across
two surveys of individuals’ values and attitudes conducted six years apart — in 2008 and
2014 — our results demonstrate that IGOs are, indeed, ‘social environments’ that shape
the values of international bureaucrats (Johnston, 2001).
Socialization and international bureaucrats: The state of
the literature and beyond
A 2005 special issue of International Organization on ‘International Institutions and
Socialization in Europe’ offered the first systematic attempt to examine value acquisition
by IGO staff. Although contributors to the volume fell on both sides of the debate, inter-
pretation of the findings by subsequent IR scholarship has been less equivocal. The
854
European Journal of International Relations 25(3)
special issue is routinely referenced as the locus classicus of a demonstration that there
is no evidence of international socialization.5 Whichever way the special issue is read,
the contributions to it share two generic shortcomings of analyses of international social-
ization in the IR literature that derive ultimately from limitations of the data available to
scholars at the time.
Socialization as a diachronic process
The first shortcoming is methodological. Socialization is a diachronic process that con-
cerns the acquisition or change in values over time. Nevertheless, virtually all existing
studies of socialization in IGOs depend on cross-sectional data sets (see, e.g., Beyers,
2005; Hooghe, 2005; Lewis, 2005; Suvarierol et al., 2013). Possible socialization effects
are typically inferred from ‘the number of years someone was involved in a particular
venue’ (Beyers, 2010: 914). Yet, an analysis of the values held at one point in time by
individuals who have worked in the organization for periods of differing length cannot
resolve potential (self-)selection issues or capture the temporal nature of socialization
processes (Beyers, 2010; Martin and Simmons, 1998). It necessarily involves either
inferring changes in values stated at time t or the correct recollection of values reported
at the time of recruitment. Hence, the validity of such an approach is questionable, and it
certainly does not allow causal inferences about individual-level socialization. A single
cross-section has severe and inherent limitations.
In contrast, this article employs a longitudinal two-wave survey research design that
is inspired by the use of longitudinal data in private-sector settings and (sub)national
public administrations in the organizational socialization literature (Gleibs et al., 2008;
Hatmaker and Park, 2014; Kjeldsen and Jacobsen, 2013; Saks and Ashforth, 1997). To
the best of our knowledge, we are the first to apply a similar approach to the study of
IGOs. This is possible due to the availability of data from two projects on the European
Commission — ‘The European Commission in Question’ and ‘European Commission:
Facing the Future’ (for more details, see later) — which collected detailed information
on staff backgrounds, beliefs and values. The surveys were administered six years apart
— in 2008 and 2014, respectively — so as to permit investigation of changes over time.
Since protecting the anonymity of staff was a condition for gaining access, it was neces-
sary to develop a technique that links the two samples and makes inter-temporal com-
parisons possible without revealing individual identities. We present and employ this
technique in the following to evaluate attitude change in the same individuals over time.
Socialization scope conditions
A second widely observed shortcoming of IR scholarship on socialization is theoretical.
The institutional and individual scope conditions that can facilitate or impede processes
of value acquisition are often unspecified, or where they are specified, are not submitted
to rigorous testing (Ashforth et al., 2007; Zürn and Checkel, 2005). The nature of our
sample (discussed in detail later) makes it possible to refine the scope conditions under
which IGOs can be expected to trigger socialization effects. We focus on three
elements.
Murdoch et al.
855
First, we test whether socialization processes are impeded when institutions are in
flux because such processes are by their very nature ‘highly contingent and potentially
shaped by exogenous and sudden events’ impacting upon individuals’ exposure to an
organization (Beyers, 2010: 917; see also Hooghe, 2005). This takes a step towards
addressing the limited empirical research ‘on the contextual factors that facilitate and
constrain socialization practices and outcomes’ (Ashforth et al., 2007: 31).6 Second,
intensified gender-equality policies in many IGOs imply that women are set to make up
a larger share of staff in coming years (Ban, 2013), which raises the question of whether
gender is one of ‘the characteristics of individual agents … [that] retard or propel the
socialization process’ (Johnston, 2001: 506). Such differences between men and women
can arise for a number of reasons, and we evaluate whether gender affects individuals’
openness to changing their attitudes towards those of the organization. Finally, our third
scope condition relates to age at entry in the IGO, which is directly linked to the oft-
stated impact of previous socialization experiences (Checkel, 2005; Cohen, 2017;
...
To continue reading
Request your trial