Early Return of Sri Lankan Migrants in the Middle East
| Published date | 01 October 1988 |
| Date | 01 October 1988 |
| DOI | http://doi.org/10.1111/j.1468-2435.1988.tb00660.x |
| Author | F. Eelens |
Early
Return
of
Sri
Lankan
Migrants
in the
Middle
East
F.
EELENS*
INTRODUCTION
The process of contract migration holds many risks, and can lead to severe stress,
especially in its early phases. Due to loneliness, unfavourable working conditions,
or
economic and familial problems, migrants may give up their hope of building up a new
and better life in another country, and return prematurely. This may have important
economic and social consequences for the migrant and his family. Contrary to the
abundant literature about the situation of migrants in the receiving countries, scientific
literature about ‘failed migration’ and early return migration is sadly lacking. In this
article we will attempt to partially bridge this gap by looking at early return migration of
Sri Lankan contract workers to the Middle East. We will pay particular attention to the
reasons why a considerable number of migrants return to their country of origin
prematurely. We will also deal with differentials in the duration of stay in the host
country,
as
well
as
some of the consequences of early return migration for the migrant and
his family.
This article forms part of a large-scale study into the economic, social and demographic
aspects of Sri Lankan labour migration to the Middle East which is being carried out
jointly
by
the Netherlands Interdisciplinary Demographic Institute, the University of
Leiden (the Netherlands) and the University
of
Colombo. The data used
for
this article
are based primarily on a random survey held among 899 return migrants, 858 family
members
of
current migrants and 409 non-migrants. The survey was held in the districts
of Colombo and neighbouring Gampaha in the period December 1985 to July 1986. In
addition to the survey, anthropological research into the consequences of labour
migration at the individual, familial and community level was carried out in
four
communities: in the towns of Colombo and Matale, and in two villages in southern Sri
Lanka.
SRI LANKAN LABOUR MIGRATION
TO
THE MIDDLE EAST
In 1983, about 3 million Asian contract migrants were working in the oil exporting
countries
of
the Middle East (Asian Pop. Studies, 1985). Compared to other countries in
*
Netherlands Interdisciplinary Demographic Institute, The Hague (Netherlands).
40
1
the Asian Region, Sri Lanka entered the Middle Eastern labour market relatively late. It
was only after the United Democratic National Party won the elections in 1977 that
labour migration to the Middle East was stimulated by the Sri Lankan government. At
that time, a number of other Asian countries already had large contingents of contract
workers in the Gulf. According to Demery (1986) some 205,000 Indians and some
354,000 Pakistanis were working in the Middle East in 1975. In 198
1
these numbers had
already increased to about
800,000
for both of these countries. Sri Lankan labour
migration also witnessed a dramatic increase in the early eighties. According to an
estimation made by the Sri Lankan Ministry of Plan Implementation, some 200,000 Sri
Lankans were working in Western Asia in mid-1983 (Asian Pop. Studies, 1985).
Compared to other labour sending countries in Asia, Sri Lanka is unique in at least two
ways. Firstly,
Sri
Lanka is the only country where more females than males leave for
overseas employment. About
70%
of
all Sri Lankan migrants to the Middle East are
women. Of those women, 60% are married, and 67% of the married female migrants
leave one
or
more children behind. Secondly, since most Sri Lankan contract workers are
engaged in unskilled and low-pay jobs, their average yearly remittances are by far the
lowest of all labour exporting countries in Asia (Eelens
&
Schampers, 1988).
DURATION OF STAY IN THE MIDDLE EAST
The first important step in any analysis of return migration
is
to determine how long
migrants stay in the receiving country and which migrants return prematurely to their
home country. When trying to do this in
a
statistically sound manner, the researcher is
faced with several methodological problems. First of all, when calculating the average
duration of stay, one should not only deal with return migrants but also with current
migrants. After all, when applying such a selection, early return migrants are
overrepresented in the sample. This would lead to
a
considerable underestimation of the
average duration of stay in the Middle East. Even if the information were to be restricted
to the last completed migration,
as
is the case in most migration studies, there would still
be a significant bias. In such samples, migrants with
a
longer duration of stay are strongly
underrepresented for they have not had the opportunity to find another job abroad once
they have returned to their home country. This is a well-known problem among
demographers and has been described,
inter alia,
for
post-partum
variables (Page et al.,
1982). In order to avoid such an important bias, we must apply a life-table method in
which migration
-
not the migrant
-
is taken as the unit of analysis, and in which all
instances of migration
-
not only the last completed one
-
are incorporated in the
analysis.
The results
of
the life-table analysis are given in Figure
1
(see page 41 1). The survival
function (l(x) in Figure 1) gives the number of migrants who, for every
1000
initial
departures, are still in the Middle East at the beginning of period
x.
The probability of
remaining in the country of emigration is calculated both
for
housemaids and for
non-housemaids. As can be seen, the proportion of housemaids who return prematurely
is far higher than the proportion of non-housemaids. About 12% of the housemaids
return home within the first year of their stay in the Middle East; 20.6% come home
before having served 18 months, while another 6% return home between 18 and 22
months after their
arrival
in
the
Middle
East
(I).
In
the
case
of
non-housemaids, this
is
only 8.3%, 14.6% and Soh, respectively. The average duration of stay was found to be
24.4 months for housemaids and 37.85 months for non-housemaids. It is interesting to
note that the average duration abroad is only 22.55 months for housemaids and 25.86
months for non-housemaids, if we restrict the analysis to return migrants (2). In the case
402
To continue reading
Request your trial