Content uploaded by Tanjina Rahman
Author content
All content in this area was uploaded by Tanjina Rahman on Apr 07, 2022
Content may be subject to copyright.
Dhaka Univ. J. Sci. 65(1) 73-76, 2017 (January)
*
Author for correspondence. e-mail: israt@isrt.ac.bd
Human Trafficking and Displacement in South Asia: An Econometric Analysis
Tanjina Rahman
1
, Md. Israt Rayhan
1*
and Nayeem Sultana
2
1
Institute of Statistical Research and Training (ISRT)), Dhaka University, Dhaka-1000, Bangladesh
2
Department of Development Studies, Dhaka University, Dhaka-1000, Bangladesh
(Received: 21 September 2016; Accepted: 22 November 2016)
Abstract
Human trafficking has received increased media and national attention. Despite concerted efforts to combat human trafficking, the trade in
persons persists and in fact continues to grow. This paper describes the relationship and distinction between trafficking and ethnic
fragmentation, conflict, internally displaced person by different measures of control. To explain the relationship between these factors, this
study uses a Probit regression model. It appears that ethnic conflict leads the internal displacement of individuals from networks of family
and community, and their access to economic and social safety nets.
Keywords: Trafficking, Displacement, Probit, Conflict, Econometrics.
I. Introduction
Human trafficking has an impact on the individuals. More
than 130 countries are affected by human trafficking.
Globally, 29 percent of human trafficking occurs after
crossing a border. Human traffickers target economic
migrants and those migrating because of lack of education,
political and climate instability, discrimination and conflict
(UN GIFT
1
).
Every stage of the trafficking process can involve physical,
sexual and psychological abuse and violence, deprivation
and torture, the forced use of substances, manipulation,
economic exploitation and abusive working and living
conditions. Some push factors behind asylum migration,
namely (i) repression of minorities or ethnic conflict, (ii)
civil war, (iii) high numbers of internally displaced persons
(IDPs) relative to total population, (iv) poverty as reflected
in low per capita income, (v) low position on the Human
Development Index (HDI), (vi) low life expectancy, (vii)
high population density and (viii) high adult literacy rate;
mirror closely the push factors that induce trafficking,
namely (i) poverty, (ii) lack of educational opportunities and
(iii) armed conflict (Castles and Loughna
2
; Akee et al.
3
).
Perry and McEwing
4
have found that social determinants are
central to the processes that mitigate and facilitate the sale
and exploitation of women and children in Southeast Asia.
Specifically, the facilitation of education and empowerment,
along with the creation and enforcement of effective
policies, could lessen the vulnerability of women and
children to modern-day slavery. Koettl
5
illustrated that
whenever people are forced or lured into exploitation – no
matter if movement of victims is involved – it is considered
human trafficking. Arhin
6
suggested new conceptual tools to
understand the dynamics and relationships between
trafficking in persons and diaspora networks. A diaspora
approach provides a more nuanced and in-depth mode of
analyzing human-trafficking cases, and takes into account
the intersections between traffickers, victims and diaspora
communities within the human-trafficking process.
In South Asia, there are many countries used as origin,
transit and destination countries for trafficking. Victims are
sent to other countries in the region and to other parts of the
world. Even more prevalent is the movement of persons
within the countries for exploitation in various forms. Even
though there are no definite numbers of victims, it is
estimated that 150,000 victims are trafficked from the South
Asia region annually (WDI
7
). Many studies have revealed
that trafficking in women and children is on the rise in Asia.
Bangladesh is a source and transit country for men, women,
and children subjected to trafficking in persons, specifically
forced labor and forced prostitution. A significant share of
Bangladeshi victims are men recruited for work overseas
with fraudulent employment offers who are subsequently
exploited under conditions of forced labor or debt bondage.
Children both boys and girls are trafficked within
Bangladesh for commercial sexual exploitation, bonded
labor, and forced labor. In Nepal, Afghanistan, Pakistan and
the Philippines, displaced children are at risk of child labor,
trafficking and forced recruitment (WDI
7
). The peculiar
situation in South Asia is that communities of same ethnic
group are spread across the international borders. As the
displacements in Afghanistan and Myanmar have direct
implications for the broader South Asian context. Pakistan,
India and Bangladesh, three important countries of the
subcontinent, face displacements simultaneously due to
development, conflicts and natural disasters. Pakistan,
Afghanistan and India had the highest number of reported
IDPs, accounting for more than a third of the regions
displaced population.
Any economic or social policy that is deemed to confer
additional benefits purely to a particular ethnic group can be
a cause for dissent and conflict. An additional cause of
ethnic conflict stems from the inability of international,
national and regional powers to adequately provide security
for minority groups. However, it can be argued that ethnic
fragmentation (or the share of an ethnic group in the total
population) within a country does not necessarily imply
dissent even in the face of perceived unjust economic and
social policies (Akee et al.
3
). Lindstrom and Moore
8
in fact
find support for the hypothesis that ethnic fragmentation is
positively correlated with conflict. About 40 million people
are displaced globally with 15 million refugees (UNHCR
9
).
Therefore, the specific objectives of this study are: to
determine the link between ethnic conflict and human
trafficking, to determine the relationship between ethnic
fragmentation and human trafficking, to assess the efficacy
and relevance with the issues of displacement in South Asia.
74 Tanjina Rahman, Md. Israt Rayhan and Nayeem Sultana
II. Data and Methodology
To determine the relationship between internally displaced
persons and refugees, ethnic fragmentation and conflict,
data are collected from various sources described below.
Data on incidence of trafficking are compiled from country-
by-country descriptive accounts and the number of
Internally Displaced Persons (IDPs) and IDP-like situations
are data collected by United Nations High Commissioner for
Refugees (UNHCR
9
). Conflict measures are collected from
the Uppsala Conflict Data Program (UCDP)/International
Peace Research Institute, Oslo (PRIO) Armed Conflict
Dataset
10
. The Gross Domestic Product (GDP) and the
landlocked indicator were obtained from the World
Development Indicator (WDI
7
). From the review of
literature this study found that trafficking is the binary
response variable. The description of other variables is
given below in the table.
Table 1. Description of Variables
Variable Description of variable
Human trafficking
trafficking
Incidence of trafficking(host
-
source country),(0/1)
Fragmentation
ethnic ethnic fractionalization index
religion
religious fractionalization index
language
language fractionalization index
IDPs/Refugees
refugees/IDPs refugee and internally displaced
persons, (0/1)
Conflict
cumulative intensity
Cumulative intensity level of
conflict
intensity intensity level of conflict: 1-
minor, 2- war
count
Number of conflicts within a
country
Here, a country is designated as a Host country for
trafficked victims only if 747 cases were reported in the past
year. Country host-source pairs of trafficking are coded
from these above reports for the year 2015.
The variables fragmentation measures are taken from Akee
et al.
3
, where fragmentation (ethnic, religious or linguistic)
is defined as
where
is the share
of group in country . Ethnic, religious
and language fractionalization cover a larger range of
countries and various aspects of fragmentation. As Akee et
al.
3
discuss, these three indices are correlat
ed and this study employs them separately in the
estimations.
For various measures of conflict in our estimations, this
study uses two measures that capture the intensity of
conflict: (i) the cumulative intensity dummy takes into
account the history of the conflict. It takes the value 0 if the
conflict has resulted in less than 1,000 battle-related deaths
and 1 otherwise and (ii) the level intensity of conflict is
measured by distinguishing between either a minor conflict
or a war (where a minor conflict has less than 25 battle-
related deaths per year for every year in the period, while a
war is defined as 1 where more than 25 battle-related deaths
per year for every year in the period). A count measures the
number of conflicts within a country. Finally, a more
complex measure is utilized that differentiates between the
types of conflict into three categories.
To explain the behavior of a dichotomous dependent
variable this study has to use a suitably chosen CDF. The
estimating model that emerges from the normal CDF is
popularly known as the Probit model (Greene
11
), uses
binomial response variables. In the Probit model, the inverse
standard normal distribution of the probability is modeled as
a linear combination of the predictors. Considering a latent
variable,
this model linearly depends on
and the error term
, here
if the latent variable is
positive and 0 otherwise, now the form is,
!"#
$ "
#
% "&
The latent variable is interpreted as the utility difference
between choosing
'()". The probability that can
be derived from the latent variable and the decision
rule.
*
i
+,
i
*
% "+,
= *,
% "+,
i
= *
% ,
+,
i
= -
./
0
1
2
3
= Φ
/
0
1
2
3
Assuming that the error term has a standard normal
distribution,
4", we have the equation, 5 Φ6
.
Where Φ is the standard normal CDF. The inverse
transformation which gives the linear prediction as a
function of the probability is, 6
Φ
.
5, The
transformation function in the Probit model is the CDF of
the standard normal distribution.
*
i
+,
i
Φ
/
0
1
2
3
= 7Φ898
/
0
1
2
.∞
If the error term has a standard normal distribution, then it is
the Probit model.
III. Analyses and Results
To determine the link between ethnic conflicts and
international trafficking, we estimate the direct effect of
ethnic fragmentation, various types of external and internal
conflicts, presence of IDPs/refugees in a source country on
the incidence of trafficking between countries.
This model is presented below:
:;<=
>
?
@A@@9*
B
;C=D;:
Human Trafficking and Displacement in South Asia: An Econometric Analysis 75
Where trafficking is the binary dependent variable for the
incidence of trafficking from country i to country j (source i
to host j). This variable takes the value 1 if an incidence of
trafficking from country i to country j is reported and 0
otherwise. The variable frag measures fragmentation in the
source country of trafficking. It is measured continuously
from 0 to 1 while frag
2
is the squared value of the
fragmentation variable. Three measures, ethnic, religious
and language fragmentation, are included in turn in the
different regression specifications. The dummy variable
refugeeidp indicates the presence of refugees as well as
internally displaced persons in the source country. The
variable conflict captures the various measures of conflict in
a source country. This study includes these various
measures in separate regression specifications for each
fragmentation measure (ethnic, religious and linguistic).
Table 2. Probit regression : Marginal Effects of Ethnic, Religion and Language Fragmentation
Variables
(Ethnic)
Estimated Coefficient
[SD error]
Variables
(Religion)
Estimated Coefficient
[SD error]
Variables
(Language)
Estimated
Coefficient
[SD error]
Ethnic
5.20*
Religion
4.18**
Language
8.93**
[2.12]
[1.58]
[3.31]
Ethnic
squared
-6.33** Religion
squared
-5.89*
Language squared
-9.64*
[2.43]
[2.11]
[4.36]
Refugeeidp
1.08
Refugeeidp
.28
Refugeeidp
2.54*
[.61]
[.53]
[1.08]
Cum-intensity
-.97
Cum-intensity
-2.32
Cum-intensity
-3.92
[.69]
[1.84]
[2.39]
Intensity
.75
Intensity
5.81*
Intensity
4.53
[.53]
[2.60]
[2.91]
Count
-.19
Count
.24
Count
.42
[.14]
[.14]
[.25]
*significant at 10%, ** significant at 5% and *** significant at 1% level of significance.
The impact of ethnic, religion, and language fragmentation
on the incidence of trafficking have been estimated the
various conflict measures. From the above table 2, it is
illustrated that higher ethnic fragmentation increases the
likelihood of trafficking from a country while the coefficient
on the squared term on ethnic fragmentation is negative and
significant under all the conflict measures. This implies that
ethnic fragmentation increases the likelihood of trafficking
but at a decreasing rate. A possible explanation of this result
might be that, higher ethnic fragmentation allows
middleman or traffickers to easily target members of
different ethnic groups and take advantages of the limited
information. A higher likelihood of trafficking is associated
positively but insignificantly with a host country under the
cumulative intensity, level of intensity and count measures
of conflict. The presence of IDPs/refugees in the host
country has positive impact the likelihood of trafficking.
I. Conclusion
Human trafficking is an issue of major international
discussion and concern. This study is an attempt to
determine the relationship between ethnic, religious,
fragmentation, and different types of conflict. In order to
study the qualitative and binary response variable, probit
regression model is used. By estimating the various conflict
measures this study found that ethnicity, religion, language,
refugee status and level intensity are significant, and that
matches with relevant literature (Akee et al.
3
, Arhin
6
).
Human trafficking involves transnational movement of
people, one important related area of debate is internally
displaced persons (IDPs). Another concern is ethnic
conflicts. Trafficking for commercial sexual exploitation is
the most virulent form of trafficking in the region. The
movement of young girls from South Asian countries is
common, taking place either between countries or within
countries. There is further movement to the Middle East as
well as other destinations. Internal displacement due to
conflict in some countries, poverty and lack of employment
opportunities increase the vulnerability to being trafficked.
Every stage of the trafficking process can involve physical,
sexual and psychological abuse and violence, deprivation
and torture, the forced use of substances, manipulation,
economic exploitation and abusive working and living
conditions. Unlike most other violent crime, trafficking
usually involves prolonged and repeated trauma. The
situation of IDPs has been more acute compared to refugees
or migrants in the absence of protection from international
organizations or states pursuing concrete policies in this
regard.
76 Tanjina Rahman, Md. Israt Rayhan and Nayeem Sultana
References
1. UN GIFT (Global Initiative to Fight Human Trafficking),
2012. An Introduction to Human Trafficking:
Vulnerability, Impact and Action. www.ungift.org.
2. Castles, S. and S. Loughna, 2003. Trends in Asylum
Migration to Industrialized Countries: 1990-2001;
Discussion Paper No. 2003/31. World Institute of
Development Research (WIDER), Helsinki.
3. Akee, R., A. K. Basu, A. Bedi and N. Chau, 2010.
Combating trafficking in Women and Children: A Review
of International and National legislation. Cooperation
Failures and perverse Economic Incentives. 2, 1-24.
4. Perry, K. M. and L. McEwing, 2013. How do social
determinants affect human trafficking in Southeast Asia,
and what can we do about it? A systematic review. Health
and Human Rights.15 (2), 138-159.
5. Koettl, J., 2009. Human Trafficking, Modern Day Slavery,
and Economic Exploitation. SP Discussion Paper, No.
0911, World Bank.
6. Arhin, A., 2016. A Diaspora Approach to Understanding
Human Trafficking for Labor Exploitation. Journal of
Human Trafficking. 2 (1), 78-98.
7. World Development Indicators (WDI), 2015. Trafficking
in Persons Report. Washington DC: The World Bank.
8. Lindstrom, R. and W. H. Moore, 1995. Deprived, Rational
or Both? Minorities Rebel Revisited, Journal of Political
and Military Sociology. 23, 167-190.
9. UNHCR, 2015. The State of the Worlds Refugees: Fifty
Years of Humanitarian Action. Oxford: Oxford University
Press.
10
. UCDP/PRIO, 2015. UCDP/PRIO Armed Conflict Dataset.
ww.ucdp.uu.se/database, last accessed 30
th
June 2016.
11. Greene, W. H., 2013. Econometric Analysis, Fifth edition,
Pearson, New York University.
