It is estimated that 940,921 Syrians work in Turkey as of 2017,17 most of them men and some working
from an early age, see Figures 3 and 4. Of these, 862,039 (91.6 per cent) do so informally, occupying
low-skilled positions in sectors where productivity is relatively low. These facts raise questions with
regards the quality of the jobs held by Syrians and the kind of industries that provide them. This
section analyzes, to the extent possible given the data limitations, where and how Syrian refugees work,
with an emphasis on wages and hours worked. It should be noted that no information is available for
working children in the HLFS.18 This section, thus, covers the 813,781 Syrian workers who are 15 years
old or older.
Source: HLFS 2017 and author’s own calculations. Notes: The table shows (i) the number of Syrian refugees working
in each economic activity, (ii) the nationality-specific share of Syrian workers in each economic activity and (iii) the
sector-specific share of Syrian workers. TCLF refers to textile, clothing, leather and footwear, ISIC rev.4 codes 13, 14
Employment structure. The economic activities and occupations held by Syrian refugees are determined by their informal working arrangements. As such, sectors like trade, construction and manufacturing, which have historically showcased high informality rates, became the main receivers of
Syrian labour. In total, these three sectors account for 79.1 per cent of the Syrian workers (see Table
1), however, one of these sectors stands out in terms of employment: Textile, clothing, leather and
footwear industries (TCLF) provide jobs to almost one in every three Syrians. The reasons behind
this sector’s attachment to Syrian labour are commented in Section 3.1, together with an assessment
of how much money is saved by its employers by hiring -most- Syrians informally. In addition to
these activities, a relatively high share of Syrians are working in a black box called ‘other services’.
In practice, a combined analysis of ISCO and ISIC codes at the 2 digit level allows us to shed some
light finding that these services refer -mostly- to repair of personal goods and clothing activities (the
latter are garment workers in occupational terms but not in terms of economic activity) vehicle and
domestic cleaners as well as possibly some waste pickers/scrap collectors.
There are cases where the concentration of Syrians working in particular activities have produced
an impact at the national level despite the fact that Syrians only represent 2.9 per cent of all workers in
Turkey. Some candidates to witness this impact include the TCLF industries, where Syrians represent
16.1 per cent of the Turkish workers and the construction sector, with 5.2 per cent. The impact
includes profitability boosts due to the low salaries and the informal nature of the workers but also
the crowding out of national workers -especially those with few years of schooling- who may not find
an informal job with the same ease as before.
Source: HLFS and author’s own calculations. Notes: The figure shows the share of informal workers in Turkey in three
economic activities for the years 2005, 2011 and 2017. The 2017 figures are estimated with and without the presence of
Syrian refugees and use the weight adjustments proposed in Appendix C. ‘TCLF’ refers to the textile, garment, leather
and footwear industries.
In addition to enhancing the survivability of certain companies, Syrian refugees are also modifying
nation-wide informality rates. This is especially so in the garment sector -see Figure 7- which currently.
showcases an informality rate of 38.5 per cent with Syrian refugees and 30.4 without them.19 Higher
informality rates can also be seen in the construction sector as well as in the food and beverage
service activities. In these cases, though, the difference between the informality rates attained with
and without Syrian refugees are more modest. The effect of Syrian refugees at the country level is
smaller but still significant; the share of informal workers (including those self-employed) goes from
34.4 per cent without Syrians to 36.0 per cent with them. The difference when only employees are
taken into account is slightly larger, going from 17.4 per cent to 20.1 per cent but in either case the
strong trend towards formal work arrangements that Turkey has witnessed in recent times has been
Source: HLFS 2017 and author’s own calculations. Notes: The table shows (i) the number of Syrian refugees by occupation, (ii) the population-specific share of Syrian refugees and informal Turkish workers in each occupation. Subsistence
agricultural workers are excluded from the Turkish group for better comparability. Occupations are measured using
A high share of Syrians can be found in low and semi-skilled occupations. This is partly due to the
low number of years in school but also the informal nature of the jobs held and the type of sector where
they are employed. Indeed, the occupations held can be traced back to the economic activities they
work for; for instance, among labourers in manufacturing and construction (ISCO-08, group 93), more
than a third work in the construction sector while half of them do so in the manufacturing industry.
Likewise, the majority (89.1 per cent) of stationary plant and machine operators are employed in the
TCLF industries. Aside from the traceability between economic activities and occupations, what stands
out of the occupational distribution is that 32.0 per cent of Syrian work in low-skilled occupations while
60.9 per cent do so in semi-skilled ones. This also implies that, only 7.2 per cent of them work in highskilled occupations, compared to 21.0 per cent of Turkish natives.
In principle, the occupations held by Syrian refugees (most of whom are informally employed)
and informal Turkish workers20 would be expected to be of similar nature, putting Syrians forward as
direct competitors of informal Turkish workers for the same jobs. However, a closer look at the last two
columns of Table 3, which shows nationality-specific employment shares by occupation, reveals that this
is not exactly the case. Syrian refugees have a much higher chance to be employed as manufacturing
labourers, machine operators and garment workers (groups 75, 81 and 93) than informally employed
natives, i.e. Syrians might be, to some extent, complementing -instead of substituting- Turkish labour
occupying jobs natives might not willing to perform.
Even though in purely economic terms Turkey might momentarily benefit from having Syrians with
skills that most complement those of natives it is unclear whether the country would be better off in
the long-term. On the one hand, keeping a large share of Syrians in low skill occupations hinders their
chances for a quick integration while maintaining a drain on many social services. The slow integration
might also cluster Syrians in poor neighborhoods, potentially generating social conflict. On the other
hand, cheap labour might foster what would otherwise constitute unprofitable economic activities,
taking valuable resources from long-term sustainable, more productive sectors.
Another negative side of Syrians being clustered in low-skilled, informal jobs has to do with the
existence of overqualified workers. It is estimated that 84,479 Syrians -roughly 10 per cent of all
Syrian workers- work with tertiary degrees. However, as it can be seen in Figure 8, only 28.9 per
cent managed to find a job in the professional category. By digging into the specific occupations
held by Syrian workers in the professional category it is possible to find that half of them work as
teachers without a work permit, narrowing further down the range of available professions for this
group. A plausible reason for the lack of access to high-skilled jobs could be related to the existence
of a language barrier, however, this should not be interpreted like teaching Turkish to random Syrians
would improve their labour market experiences in the short-run; Syrians are relatively clustered in a
few provinces and chances are their customers/co-workers are Syrian too. High-skilled Syrian refugees,
though, might benefit greatly -themselves and their hosting country- from more integration, including
In addition to the language related issues, problems validating foreign degrees and acknowledgment
of prior learning are likely to lower the value of Syrians capacities, particularly in highly regulated
professions where signals matter as much -if not more- than actual human capital. Together with
the two commented reasons, the fact that high skilled jobs are usually provided under formal work
arrangements conform a triple threat to well educated Syrians in their quest to use their abilities.
Source: 2017 HLFS and author’s own calculations. Notes: The figure shows the education and nationality-specific
occupational distribution for Syrian refugee and Turkish native workers with tertiary education. Occupations measured
with ISCO-08 codes.
Although high-skilled Syrians might be among the ones who lost the most in terms of foregone
earnings due to the lack of work permits, informal work arrangements are pervasive across Syrian
refugees with 9 in 10 working without being registered in the social security institution. Generally,
employees hired informally tend to work in small businesses. This is for two reasons; firstly, because
this practice is unlawful and it is easier to conceal the activity if the premises are not visible. Secondly,
because these businesses tend to be less profitable than bigger companies and, thus, might need to
rely on a cheaper workforce. At the international level examples of this behaviour are easy to find;21
in Mexico 81.6 per cent of informal employees work in businesses with less than 10 employees while
in India and in Egypt the percentages are, respectively, 86.7 and 88.4 per cent. A similar share of
informal employees working in micro-enterprises is also found in Turkey for natives, 78.5 per cent, with
almost half of them working in hard to spot (non-regular) places such as irregular/mobile premises or
at home -see Figure 9(a).
Surprisingly, the empirical regularity regarding company sizes for informal employees is broken for
Syrian refugees. According to Figure 9(b), 43.8 per cent work informally in businesses of more than 10
people. Moreover, 32.0 per cent of Syrians work in companies with more than 20 employees; something
rare to witness elsewhere in the world of work. On top of the high share of informal employees working
outside of micro-businesses, those working in companies with less than 10 employees also showcase a
higher tendency to do so in regular premises which are arguably easier to spot by the labour inspection.
To sum up, Syrian refugees’ employment structure is uncommon in terms of the combination of
Source: 2017 HLFS and author’s own calculations. Notes: The figure shows (i) the nationality-specific distribution
of company sizes for Syrian refugee and Turkish informal employees (in blue) and (ii) the nationality and company
size-specific share of workplace types for companies with less than 10 employees.
characteristics -informal employment in relatively large companies and high prevalence of the manufacturing sector- and definitely only representative of Syrians in Turkey. For instance, in Jordan only
20.6 per cent (48.2 per cent in Turkey) of Syrians are employed in the manufacturing sector while
36.8 per cent work in the construction sector and 29.8 in trade and hospitality. The above-mentioned
characteristics are not necessarily negative, though, and might play in favour of Syrian refugees’ future
labour market experiences. The relatively high visibility of the businesses where Syrians work could,
in principle, be used by the labour inspection of Turkey to optimize its efforts in reducing informality.
Moreover, the productivity of the manufacturing industries is higher than that of the trade and hospitality sector, increasing the viability of the formalization of Syrian refugees in their current posts.
Obviously, the above-stated remarks would do little good if Syrians are not eligible to request a work
permit. We invite the reader to follow the discussion on section 3.2 with regards the extent to which
Syrian refugees under temporary protection can request a work permit.
Working conditions. High informality rates and jobs clustered in some of the sectors where longer
hours22 are worked in Turkey are likely to shoot up the hours worked by Syrian refugees beyond legal
limits. Indeed, Syrian refugees work, on an average week, 52.3 hours, almost five and four hours more
than, respectively, formal and informal native employees. In addition, more than half (53.7 per cent)
of the Syrian employees work more than 50 hours a week and 34.7 per cent work 60 or more hours.
The complete distribution of average weekly hours is provided in Figure 10(a) for Syrian employees and in Figure 10(b) for, separately, Turkish formal and informal employees. At first sight, the
distribution of hours worked by Syrian resembles neither of the natives’ distributions, however, the
devil is in the detail and a closer look reveals things are not really that different. For instance, weekly
hours worked by formal natives go up by 2.1 when removing the public sector -represented in Figure
10(b) by the peak at 40 hours. That is, without public employment (that belong exclusively to formal
native employees) both, the Syrians and the Turkish distribution share a peak at 48 hours, the most
common arrangement in Turkey. Still, the distribution of hours worked by Syrians is ‘shifted’ to the
right, with a second peak at 60 hours which is characteristic of the informal economy. The distribution
of hours worked by informal natives does not seem to measure up to the Syrian one; the former has
a much ‘heavier’ left tail (individuals who work part-time) than that of Syrian refugees. However,
if we only take into account full-time informal employees, Syrians (55.0 hours) and informal native
employees (55.8 hours) look very much like each other. In sum, it could be said that the Syrian refugee
distribution of hours worked is a mixture of full-time informal Turkish employees (right-hand side of
the distribution) and part-time formal Turkish employees (for the left part of the distribution).
Source: HLFS 2017 and author’s own calculations. Notes: The figure shows the distribution (probability density
function) of average weekly hours worked by (a) Syrian employees, (b) Turkish informal employees and (c) Turkish
formal employees using Kernel density estimation. The height of graph corresponds to the frequency of employees with
a given number of hours worked.
A salient issue of the weekly hours distribution has to do with the low prevalence of part-time
arrangements among Syrians and natives; despite an increase in this mode of work from 10.1 per cent
back in 2005 to 14.4 in 2017 among 15-65 female employees, it still falls short in comparison with some
European countries like France (30.5 per cent, 2016) or the United Kingdom (44.8 per cent, 2016).
This, in turn, might still constitute a barrier of entry to the labour market for women who also carry
out duties at home.
Another reason that might be driving Syrian women away from the labour market is the lower
wages earned in comparison with men. The unadjusted monthly gender wage gap for Syrian refugees
stands at 23.4 per cent while the hourly counterpart is 9.4 per cent. These are, in both cases, higher
than the gender wage gaps of Turkish citizens.23 Women is not the only disfavored group among
Syrian refugees, young people (15-29 years old) also earn 18.2 per cent less than adults per month and
21.6 per cent per hour worked.
Source: HLFS 2017 and author’s own calculations. Notes: The table shows average hourly and monthly net wages of
Syrian refugees measured in nominal 2017 TRY by sex, age-group and Istanbul/rest of the country. Young refers to
those aged 15-29 and adult to those aged 30-65.
Irrespective of the group we put the focus on, Syrian refugees’ jobs are characterized by low earnings;
the average take-home salary is 1,302 which, in 2017, was below the statutory minimum wage for
full-time employees, 1,404. Minimum wage compliance provides a darker picture when assessed from
an hourly point of view. Syrians tend to work more than the 45 hours required to earn a full-time
salary and, as a result, 3 out of 4 Syrian employees earn less than the minimum wage per hour. The
percentage of Syrians earning below the minimum wage is slightly lower in Istanbul, where refugees
enjoy higher wages than in other regions, however, this difference is likely due to the higher living cost
displayed by the largest city of Turkey and not to the generosity of employers. Perhaps surprisingly
in light of the gender wage differential is the fact that a higher share of men earn below the hourly
minimum wage, 76.8 per cent compared to 68.7. This result is driven by the shape of each genders’
wage distribution; low female earners earn very little but there is a higher percentage of them earning
above the minimum wage than among men.
On average, Turkish natives earn 63.1 per cent more than Syrians and an obvious question is
whether refugees might be discriminated against by employers. One way to answer this question, at
least partially, is by means of a so-called wage regression. This tool allows us to control for job-related
and personal characteristics that might influence hourly wages; for example, Syrians’ low wages might
be partly explained by the fact that they work informally. Some of the variables used in this exercise
include the sex, age24 and educational attainment of the person. In addition, job-related variables
like the economic activity, the occupation and the formality of work arrangements are also taken into
Source: HLFS 2017 and author’s own calculations. Notes: The table shows shows the group-specific percentage of Syrian
employees earning below (<95%) at or around (95%-105%) and above (>105%) the net hourly minimum wage. The
minimum wage is expressed in nominal TRY of 2017.
Results from the wage regression are mixed; on the one hand, we find that informal Syrian employees
are paid the same as Turkish informal employees -other things being held constant. On the other hand,
the few Syrians working formally earn, on average, 10 per cent less than Turkish citizens with similar
characteristics. An even bigger loss, though, is faced by highly educated Syrians who work without
a work permit. By using a modified wage regression26 we can calculate the ‘return to schooling’ -the
increase in wages due to an increase in the number of years spent in school- for Syrians27 as well as
for formal and informal Turkish employees. The results, which can be seen in Figure 11, show for
each of the three groups under analysis the percentage increase in earnings achieved by a higher level
of educational attainment. It should be noted that the comparison is in relative terms and that the
return to schooling of Syrian refugees’ with no formal education is set to 0.
We find that Syrians with tertiary degrees are penalized twice in the labour market. First for not
having a work permit and then for their degrees not being recognized. Indeed, Syrians’ return to a
college degree is 5 times smaller than the one received by a formal Turkish native. As expected, the
informality of work arrangements halves the return to college, from 76.9 per cent to the 37.1 per cent
actually enjoyed by Turkish informal workers. However, it looks like employers of informal employees
also value differently the knowledge brought by Turkish and Syrians, with the latter obtaining a mere
14.6 per cent increase in their hourly wages for their degrees. This would, perhaps, give support to the
importance of job-market signals as pioneered by Spence (1973) in contrast to the importance attached
to sheer human capital.
Source: 2017 HLFS and author’s own calculations. Notes: The figure shows the return to schooling (additional salary
earned for having spent more years in education) of three groups, informal Syrian employees, informal Turkish employees
and formal Turkish employees. Results are extracted from a (log) wage regression where controls for sex, age, occupation,
economic activity, firm size, type of workplace and ownership (public/private) are also included. Syrian refugees with
no schooling constitute the reference point and their return is set to zero.
This post is also available in: English