Untold Stories: Biases and Selection Effects in Research with Victims of Trafficking for Sexual Exploitation
| Published date | 01 August 2010 |
| Date | 01 August 2010 |
| DOI | http://doi.org/10.1111/j.1468-2435.2010.00628.x |
| Author | Rebecca Surtees,Anette Brunovskis |
Untold Stories: Biases and
Selection Effects in Research with
Victims of Trafficking for Sexual
Exploitation
Anette Brunovskis* and Rebecca Surtees**
ABSTRACT
Recent discussions of trafficking research have included calls for more
innovative studies and new methodologies in order to move beyond the
current trafficking narrative, which is often based on unrepresentative sam-
ples and overly simplified images. While new methods can potentially play
a role in expanding the knowledge base on trafficking, this article argues
that the solution is not entirely about applying new methods, but as much
about using current methods to greater effect and with careful attention to
their limitations and ethical constraints. Drawing on the authors’ experi-
ence in researching trafficking issues in a number of projects over the past
decade, the article outlines and exemplifies some of the methodological
and ethical issues to be considered and accommodated when conducting
research with trafficked persons – including unrepresentative samples;
access to respondents; selection biases by ‘‘gatekeepers’’ and self selection
by potential respondents. Such considerations should inform not only how
research is undertaken but also how this information is read and under-
stood. Moreover, many of these considerations equally apply when consid-
ering the application of new methods within this field. The article
maintains that a better understanding of how these issues come into play
and inform trafficking research will translate into tools for conducting
improved research in this field and, by implication, new perspectives on
human trafficking.
* Fafo, Institute of Applied International Studies, Oslo, Norway.
** NEXUS Institute, Washington, DC.
2010 The Authors
Published by Blackwell Publishing Ltd., Journal Compilation 2010 IOM
9600 Garsington Road, Oxford OX4 2DQ, UK, International Migration Vol. 48 (4) 2010
and 350 Main Street, Malden, MA 02148, USA. ISSN 0020-7985
doi:10.1111/j.1468-2435.2010.00628.x
MIGRATION
Edited by Elzbieta Gozdziak, Georgetown University
INTRODUCTION
Trafficking in human beings has long been a topic high on the political
agenda in many countries and for many international organisations.
Research on trafficking is an important source of information about
the phenomenon and forms the foundation of policy and interventions.
Simultaneously, however, there are many scholars and activists who
voice concern over the image of trafficking that emerges from this col-
lective body of research, thus raising questions about its usefulness.
While the severe harm inflicted upon some trafficking victims is
well documented, some concerns are tied to whether the image of the
extremely brutalised victim is representative or sufficiently common to
warrant current programmatic and policy response (Weitzer, 2007:
456).
1
This image may not fit some women who have been trafficked
but are subjected to more mundane pressure and control (Kelly, 2002:
34). As Zimmerman and Watts (2003: 3) note, ‘‘It should not be
assumed […] that all women who have been trafficked are traumatized,
consider themselves victims, detest their captors, or wish to escape or
go home’’. These concerns illustrate the importance of accessing
trafficking victims more broadly, documenting and analysing not only
those narratives that are by now established, but also those stories that
are presently untold.
While research into the topic has expanded over the past decade, it has
also been accompanied by a call for more innovative studies and new
methodologies in order to move beyond the victim narrative and to pro-
duce conclusions that can be applied beyond each specific sample. A
recent survey of human trafficking research concludes that most empiri-
cal research on the topic is based on qualitative methodology and small
samples, which makes generalizations and extrapolation a problematic
exercise. The report calls for the use of methodologies such as capture-
recapture or respondent driven sampling (RDS) in order to broaden the
scope of trafficking research (Gozdziak and Bump, 2008: 10). Similarly,
Cwikel and Hoban (2005) recommend’’…using creative research and
investigative methods and accessing unlikely key informants, including
brothel workers, managers, and traffickers or their representatives’’.
And, in Southeast Asia in 2007, the United Nations Interagency Project
(UNIAP) initiated a competition to find innovative and rigorous meth-
ods for estimating numbers of trafficking victims in a given geographic
area and ⁄or sector (UNIAP, 2008). Indeed many recent discussions and
publications on trafficking research stress methodological limitations
2Brunovskis and Surtees
2010 The Authors
Journal Compilation 2010 IOM
and, in many quarters, there has been a call for ‘‘new methods’’ and
‘‘new approaches’’ in trafficking research and data collection (see, for
instance, Aghazaram et al., 2008; Cwikel and Hoban, 2005; Gozdziak
and Bump, 2008; Harvard University, 2009; ILO, 2006; Laczko, 2007;
Savona and Stefanizzi, 2007).
We agree that much of the current research on trafficking is not particu-
larly representative; that the field could benefit from a larger variety of
methodological approaches. And we contend that trafficking research is
constrained by several methodological limitations, related to both practi-
cal considerations and ethical concerns. However, we also believe that
the solution is not entirely (or perhaps even primarily) about new meth-
ods. Rather, we would argue that it is as much about how we use cur-
rent methods to greater effect and with careful attention to their
limitations and ethical constraints. Moreover, where new methodologies
are applied in this field, careful attention is needed to how to undertake
ethically sound research with trafficking victims, including in terms of
respondent recruitment and selection effects.
Over the course of a number of trafficking research projects we have
often found that what seemed like good ideas when designing projects at
our desks were hopelessly unfeasible when confronted with the practical
and ethical concerns in the field, warranting adjustment and rethinking
of our methodological approach. The purpose of this article is to outline
some issues to be considered and accommodated when conducting
research with trafficked persons. We also seek to exemplify these issues,
drawing on research we and others have conducted. These concerns
should inform not only how research is undertaken but also the reading
of current research on human trafficking, highlighting both strengths
and weaknesses of the largely qualitative body of research. It is
important to think critically about who is included (and who is omitted)
in current analyses of human trafficking and which perspectives and
aspects are considered and even privileged. These concerns and in-field
limitations – both practical and ethical – should also be taken into
account when designing new research projects, particularly those
employing new methodologies that have, to a limited extent, been tested
in the field.
Our main focus in this article is on qualitative research, rather than
quantitative. Nevertheless, many of the ethical and practical concerns will
be equally valid regardless of methodological orientation, notably so
when research involves primary data collection with victims. A number
Untold stories: research on sex trafficking 3
2010 The Authors
Journal Compilation 2010 IOM
To continue reading
Request your trial