Content uploaded by Islam Aburok
Author content
All content in this area was uploaded by Islam Aburok on May 06, 2020
Content may be subject to copyright.
The impact of Syrian Refugees on Income Gender Inequality
Case Study-Syrian Refugees in Jordan
Islam W. H. Aburok
Abstract
This paper investigates the impact of Syrian refugees (SRs) on gender income
inequality as a dimension of their impact on the labor market in The Kingdom of Jordan. the
paper uses the methodology of Difference in Differences based on a panel data set for two
years, 2010 and 2016. The main findings of the study are the following. First, at the governorate
level, the SRs have a positive significant impact on the rates of an individual's income in
Jordan. This impact has increased the rates of income in the treated governorates rather than
the control governorates. The change in rates of income between governorates increases
income inequality between governorates. Second, at the gender level, the existence of SRs in
Jordan increased the rates of income of males rather than females. This difference increases the
income inequality between males and females. Third, at an individual fixed effect level, the
SRs have shown an insignificant impact on the income rates of natives. Fourth, at the level of
the work sector (formal/informal), the SRs affected significantly on the rates of income of the
individual. The rates of income of natives who work in the formal sector increased rather than
rates of who work in the informal sector. Generally, the SRs in Jordan affected the income
inequality at gender and between governorates levels. However, other control factors contribute
majorly to increase the gap of income between natives such as education, age, and size of
households. Also, other factors related to the work policies and characteristics of the Jordanian
labor market contribute to increasing gender income inequality.
Keywords: Immigration; income inequality; gender inequality; Syrian Refugees; Jordan,
Refugees; labor market.
1. Introduction
In the second decade of the twenty-first century, as an extension of Arab spring that
sparkled in 2010. The Syrian crisis has emerged in 2011 through an intensive civil war that
caused the largest movement of refugees in the world recently. It has forced Syrians to seek
safety elsewhere. Seven years later, approximately 13 million Syrians have eloped their homes,
based on the recent statistics of the United Nations High Commissioner for Refugees (UNHCR,
2020).
The destination countries of the SRs are the neighboring countries, The Kingdom of
Jordan, Lebanon, Egypt, Turkey, and European countries such as Germany, Netherlands,
Sweden, and Greece. Jordan is one of the neighboring countries that received a high number
of SRs which reached 656,103 in March 2020 with a percentage of 10% of the total population.
(UNHCR, 2020).
The SRs in Jordan live inside and outside the camps. The percentage of SRs who live
inside camps is around 19% and they are deployed in five camps over three governorates. They
are concentrated mainly in the Zaatari camp and Azraq camp in Mafraq and Irbid governorates,
respectively. While most SRs live outside the camps, they concentrated mainly in the
governorates of Amman, Mafraq, and Irbid with a percentage of 81% (UNHCR, 2020)
compares to 76% in 2016 (UNHCR, 2016). Originally, most SRs in Jordan came from the
governorates of Dar'aa (48%), Homs (19%), Aleppo (10%), and Damascus (17%). They are
particularly from the rural areas of these governorates and 2% of them returned to Syria (Tiltnes
et al., 2019). Around 93% of SRs live under the poverty line. However, the World Food
Program (WFP) provides monthly food or cash vouchers (WFP, 2019).
Regarding the characteristics of SRs in Jordan, they are extremely young, about 48%
are below the age of 15. The participation rate of SRs who lives outside camps reaches 7% for
women and 51% for men. However, they have a high unemployment rate of 57%. In contrast,
66% of Jordanian men and 18% of women participate in the labor market as shown in Fig. 5
and Fig. 6, respectively in Appendix (B). The unemployment rates among Jordanians increased
from 14.5% before the influx of SRs to 22% after that. The SRs work mainly in informal jobs.
While 10% work formally since it is difficult to get legal work permits (Stave & Hillesund,
2015). However, the government does not hinder their movement across the country (Carrion,
2015). The SRs work mainly in sectors of construction, food services, repair industry,
manufacturing, wholesale and retail trade, accommodation, and in the occupation of craft and
related trade workers, service and sales workers, and elementary occupations, Fig. 7 in
Appendix (B) shows the percentage of employment by sector for SRs and Jordanian (Stave &
Hillesund, 2015).
Many researchers study the impact of SRs on hosting countries from different
perspectives. Some studies used the same data source and hosting country of this study and
they conclude different results. For instance, Malaeb & Wahba (2018) conclude that the
existence of SRs in Jordan led to a decrease in the labor force of immigrants and Jordanians
over time, but it is not significant. Besides, the SRs led to a decrease in work hours but without
an effect on their hourly wages. Therefore, the SRs affected significantly negative on total
wages. In addition, Fallah et al. (2019) use the same methodology and data source of this study.
They conclude that the SRs have no negative impact on the Jordanian labor market. However,
some changes that occurred especially in the work sector include positive significant effect on
hourly wages. In Turkey, Bahcekapili & Cetin (2015) results showed a positive impact for the
SRs on the unemployment rates in Gaziantep and Adıyaman and Kilis (GAK) and significantly
occurred among the unskilled labor groups. However, the SRs affected negatively on economic
growth in Turkey irrespective of positive change in some regions such as GAK. Similarly,
Kuyumcu & Kösematoğlu (2017) conclude that the inflow of SRs in Turkey increased the
supply of informal workers in the informal sector. Therefore, wages declined. Generally. The
SRs have made a negative impact on the Turkish economy. In Germany, the SRs showed a
positive impact on employment rates and national wages (Chatzichristou, 2018).
On the other hand, some studies investigate the impact of SRs on income inequality.
Xu et al. (2016) investigate empirically the impact of immigration on income inequality at the
level of 50 states in the USA for the period from 1996 to 2008. They conclude that the
immigrants have a long term significant positive impact on income inequality at the state level.
Green & Green (2016) investigate the impact of immigration in Canada on the distribution of
earnings during the first five decades of the twentieth century. They conclude that immigrants
were a minor factor in increasing the inequality between natives and other factors play majorly
in increasing inequality. However, immigrants can contribute significantly to widening the gap
of distribution in the hosting countries if they are characterized by the same skills of natives in
the labor market.
This paper emphasizes the question of How do the SRs affect the gender income
inequality of Natives in Jordan? It aims to investigate if the SRs have any impact on
changing the levels of rates of income of employed natives in twelve cities across Jordan.
This research contributes to other research in this context by shed light on one of the
countries that received a high number of Syrian immigrants. In addition, it is the first research
to investigate the impact of the SRs on the gender income inequality of natives in Jordan since
a few researchers illustrated the impact of SRs on the labor market in Jordan. On the other
hand, it attempts to find results that could demonstrate the impact of SRs on the Jordanian labor
market since some studies used the same data source and methodology concluded different
results. For instance, Fallah et al. (2019) conclude that the SRs have no negative impact on the
Jordanian Labour market. In contrast, Malaeb & Wahba (2018) concluded a negative impact
of the SRs on the Labour market. Besides, the concerned researchers about the impact of the
SRs in Jordan concluded different discussions between the negative and the positive impact.
Accordingly, this study uses a major variable in the labor market that could illustrate
the impact of SRs on natives. Also, it supports similar results and discusses and analysis the
different results by evidence.For this purpose, the study uses micro-level panel data across
individuals from the Jordanian Labour Market Panel Survey (JLMPS) over two years, one
before the inflow of SRs (2010), and the second after the inflow (2016). The dependent variable
is the income of natives measured by using a proxy variable for the total monthly wages. While
the independent variable is the percentage of SRs in the selected governorates based on data
from UNHCR (2016). Other control variables are used to determine the effect on the income
of natives for instance; years of schooling, marital status, number of individuals in the
household, age, urban/rural area of living, marital status, and formal or informal job. The paper
uses a difference in difference methodology to measure the average difference in wages
between the treatment group and the control group over time. The paper concludes the result
at three different levels using fixed-effect regression. The first level measures the impact at the
governorate level, the second at the gender level. While the third at individual fixed effect level.
The main findings of this research are the following, at the governorate level, the SRs
have a positive significant impact on the rates of an individual's income in Jordan. This impact
has increased the rates of income in the treated governorates rather than the control
governorates. In addition, the change in rates of income between governorates increases
income inequality between governorates. At the gender level, the existence of SRs in Jordan
increased the rates of income of males rather than females. Third, at an individual fixed effect
level, the SRs have shown an insignificant impact on the income rates of natives. Fourth, at the
level of the work sector (formal/informal), the SRs affected significantly on the rates of income
of individuals based on the type of work. The rates of income of natives who work in the formal
sector increased rather than rates of who work in the informal sector which affected by the
concentration of the SRs workers in the informal sector. Overall, the SRs affected significantly
on the income inequality between males and females and between governorates. Also, the
results showed a major impact of other control variables on the income inequality between
natives. These variables include education, age, size of households, and work policies.
The remainder of the paper includes seven sections. The second section illustrates the
theoretical concept behind the research. Section 3 illustrates a group of related studies and how
they rely on this study. Section 4 includes the data description. Section 5 includes the
methodology. The sixth and seventh sections include results and conclusion, respectively.
2. Conceptual Framework
Regarding the theoretical framework, immigration could come under the scope of the
international trade theory since the movement of immigrants including refugees leads to an
increase in the demand for labor to relative wages (Kuhn & Wooten, 1991). Similarly, Card et
al. (2009) states that the standard economic theory suggests that there is a similarity between
migration and free trade. The migration will create a surplus in returns due to the increase in
the supply of immigrant workers. Then, the surplus will be redistributed which will ensure a
better state for the residents. In addition, Bahcekapili & Cetin (2015) state that the immigration
with free trade and movement increases the return to scale theory, then in the hosting countries,
the growth rate is more than the normal rate. As a result, will be an increase in the return on
human capital as well as wages.
On the other hand, some economic theory supposes that the inflow of immigrants
considering them as new skills does not create change in the labor market. For instance, the
theory of Hecksher-Olin emphasizes that when the demand curve is stable, the supply of a new
group of skills by new human capital leads to expansions and contractions in a different
industrial sector without impact on the relative skills or wages (Card, 2009). Moreover, the
economic theory indicates the immigrants in the hosting countries increase the national income
of natives especially when the natives and immigrants differ in wealth. The added value of
immigrants increases when they are a complement to natives, not competitors. Generally, the
impact of immigrants in the hosting countries relies on the characteristics and labor skills of
both immigrants and natives. Besides, it relies on the policies of immigration in the hosting
countries (Borjas, 1999).
3. Literature Review
Many studies investigate the impact of immigration on the labor market including the
impact on the income of people living in the hosting countries either other immigrants or
natives. A group of these studies investigates the impact on the countries experienced high
waves of immigration for long years such as the United States of America (USA). Recently,
with the growth in the movement of SRs due to the internal conflict, many studies started
investigating their impact on the major hosting countries such as Turkey, Jordan, Lebanon, and
German. Those studies use different methods and concluded to different results depend on the
characteristics of immigrant refugees and the labor market in the hosting country.
3.1- On The Relevance of The Impact of Syrian Refugees on The Labor Market in The
Hosting Countries
Kuyumcu & Kösematoğlu (2017) assess the changes in economic performance in
Turkey before and after the influx of SRs in 2011 from a different perspective, including the
changes in unemployment rates, wages, formal and informal labor market, and impact on
children employment. The research uses variables of gross domestic product (GDP), poverty
rates, access to health, education and municipal services, and business and household
confidence to investigate the impact of SRs using different statistics. The study concludes that
the arrival of SRs made a positive supply shock and increasing in the competition for low
skilled Turkish workers in the informal markets due to the lack of work permits provided for
SRs especially before January 2016. However, the impact on the demand for formal labor is
unclear. The increase in the supply of informal workers reduces wages, increases the demand
for informal, not formal labor. Generally, Turkey's economy has affected negatively due to the
inflow of refugees. However, the GDP of Turkey grew in some years and declined in others.
Chatzichristou (2018) studies the impact of SRs on the German Labour Market using
data from the European Union Labor Force Survey (EU LFS) for the period between 2011 and
2016 using Ordinary Least Squares (OLS) and a novel Instrumental Variable of changes in the
population of refugees source countries in Germany over time, based on quarterly information
on labor and non-labor population for people with 15 to 64 years of age, and Syrian population.
The study uses the methodology of Ximena V. Del Caprio and Mathis Wagner that they use
for studying “The impact of Syrians Refugees on the Turkish Labor Market”. The paper uses
data for women and men ages 15-64 and divides them into two groups active and inactive
people. The results concluded that Syrian refugee has a positive and significant result on
German employment rates and national wages with only exception females of ages 15 to 24.
While there is a negative coefficient with the unemployment and retirement rates.
3.2- On The Relevance of Context and Data of This Study (The Impact of Syrian
Refugees on The Labor Market on Jordan)
Fallah et al. (2019) investigate the impact of the SRs on the labor market in Jordan. The
study uses a panel data set using two waves of the Jordanian Labour Market panel survey
(JLPM). The first wave before the influx of SRs into Jordan (2010) and the second after the
influx (2016). For the percentage of SRs in Jordan, it uses the data from the census of 2015.
The study limited to SRs and Jordanian men aged 15-64 since the participation rate in the labor
market is low for both SRs and Jordanian women. It examines how they affected in different
sectors. For instance, formal, informal, public, and private. Furthermore, it compared between
men based on the education level, basic or less vs. secondary and higher)". The study examines
the changes in weekly hours, hourly and monthly wages. Using the methodology of difference
in difference, the main findings of the study concluded that the SRs affected positively
significant in the formal sector. Also, a positive significant effect on hourly wages. While they
have an insignificant impact on employment and unemployment. Generally, the results
concluded the effects of SRs in Jordan is not negative. However, some changes occurred,
especially in the work sector.
Malaeb & Wahba (2018) examine the impact of SRs on the other immigrants in the
Jordanian labor market since most of the SRs are unskilled, as a result, they compete for the
immigrants in the informal sector especially in the regions that high numbers of SRs. The study
focuses only on males aged between 15 and 59 years since in Jordan the labor force
participation of females is low. It investigates the impact using panel data of the Jordanian
Labor Market Panel Survey (JLMPS) before the Syrian influx (2010) and after the Syrian influx
(2016). The study focuses on identifying the impact on the immigrants' employment, hourly
and total wages, weekly hours of work, formal and informal work and participation of the labor
force. The study uses an instrumental variable approach to illustrate the impacts by using two
different instruments to make a good identification strategy. The first is using an instrumental
variable that utilizes the distance to borders and refugees' camps since about 80% live outside
camps. The second is using an instrument for identifying the settlement of immigrants before
the wave of Syrian influx which ensures measuring the impact on immigrants in areas that have
positive outcomes of the labor market. Based on the Ordinary least square analysis, the results
show a decrease in the labor force of immigrants and Jordanians over time, but it is not
significant. Moreover, the effect on the market outcomes is higher in the regions with a low
percentage of refugees compared to those with a high percentage. On the other hand, the
percentage of refugees makes immigrants work less than usual but without effect on their
hourly wages, which causes a negative significant effect on the total wages.
3.3- On The Relevance of The Research Question (Impact of Immigration on Wage)
Card (2009) illustrates the impact of immigration on the wage inequality in the United
States of America (USA) at different levels across cities. The study uses data about the
immigrant densities, mean salaries, and education outcomes in twelve largest cities with the
nation from two sources, census of 1980 and 2000 and American Community Survey (ACS)
for 2005 and 2006. The data illustrates significant characteristics of both immigrants and
natives who have years of experience up to 45 years and their ages over 18 years. One of these
levels illustrates the wage gap between immigrants and natives. Wage inequality is
significantly related to the densities of immigrants. Generally, the wage inequality within
groups of natives after controlling a group of characteristics such as education race, gender,
ethnicity, and age has a significant relation with the share of immigrants in the studied cities.
While the wage between groups of natives has no strong relation with the share of immigration
in the labor market since immigrants cause approximately 5% of the wage inequality in the
USA from 1980 to 2000. By gender, among men, wage inequality among native workers men
increased by 0.137 and by 0.142 among all workers, about 4 % difference. In contrast, the wage
inequality increased by 0.139 between native women, and by 0.148 between overall women
workers, about 6% difference.
Xu et al. (2016) investigate empirically the impacts of immigration on the income
inequality at the level of 50 states in the USA for the period from 1996 to 2008 based on the
data source of Guetzkow, Western, and Rosenfeld (2007) and using dynamic and static models
of pooled cross-sectional time-series data. The research studies the impact of four immigrant
groups including high-skills immigrants and low-skill immigrants. The dependent variable is
income inequality and is measured by using the Gini coefficient for state-level, While the
independent variables are the percentage of immigrants from the total population in each state.
The study uses two models. First, the static Ordinary Least Square (OLS) and the dynamic
model by Error Correction Model (ECM). The results conclude that both low-skills and high-
skills have a significant positive impact on income inequality. However, the impact of low-
skill immigrants is more than high-skills immigrants. Generally, the study concludes that
immigrants have a long term significant positive impact on income inequality at the state level
in the USA.
3.4- On The Relevance of The Methodology of Difference in Difference (DID)
Bahcekapili & Cetin (2015) investigate the impact of SRs on the Turkish economy. It
focuses on the southeastern Anatolia region which has the most population of SRs. It uses a
cross-section data from Turk Stat's statistics to obtain the effect of immigration on the
workforce, prices, internal migration, and foreign trade in selected regions by comparing the
differences before (the period from 2010 to 2012) and after immigration (the period from 2013
to 2014). The study uses a macro level for the variables of education level, inflation, food, and
non-alcoholic beverages, clothing and footwear, actual rentals for housing, transport inflation,
foreign trade, and internal migration data to investigate the effect on unemployment rates in
each selected region using the methodology of (DID). The results indicate a noticeable decline
in the unemployment rates in Gaziantep and Adıyaman and Kilis (GAK), and the decline
significantly occurred among the unskilled labor groups. Generally, the study shows that the
SRs contributes negatively to the economic growth in Turkey irrespective of positive change
in some regions such as GAK.
4. Data
The research uses micro-level panel data from different sources. The first source of data
is the (JLMPS) conducted by the Economic Research Forum (ERF) in collaboration with the
Jordanian Department of Statistics (DoS). The survey covers two years, the first is before the
Syrian influx (2010) and the second is after the Syrian influx (2016), the sample of the survey
consists of 7,229 households with 33,450 individuals (Krafft & Assaad, 2018) To investigate
the objective of the research question, this study is limited to 3,075 (2,416 males, 659 females)
employed native Jordanian aged between 16 to 64 and participated in the two waves of survey
2010 and 2016. The second source of data is the statistics of the (UNHCR) which is used to
obtain data about the percentage of the registered SRs in twelve governorates in Jordan, table
(5) in appendix (B) shows the percentage in each governorate.
The governorates are divided into two groups, the treatment group that received high
percentages of SRs and the control group that did not receive SRs or received a small
percentage (less than 7%). The dependent variable is the income of natives and measured using
a proxy variable for the total monthly wages during the last months before the time of the
survey meeting. The independent variable is the percentage of SR in each governorate which
is calculated by dividing the number of SRs by the total population in each governorate in 2016.
A group of dummy variables is generated based on the concept of Difference in Difference
methodology which investigates the impact of treatment on the treated group, in this case, the
treatment is the inflow of SRs into Jordan. These variables include the intervention dummy
variable indicates 1 for the treated group and 0 for the control group, the time dummy variable
indicates 1 for the year after Syrian immigration (2016) and 0 for the year before the
immigration (2010). Besides, the study uses a group of control variables that could affect the
total monthly wages of native Jordanians since it is difficult to measure the impact of refugees
without taking into consideration the existence of other factors. Those variables include (age,
years of school, the total number of individuals in the household, marital status fixed effect,
type of job (formal/informal) fixed effect, and the place of living (urban/rural) fixed effect).
Moreover, the variable of governorate fixed effect is used to investigate the degree of impact
on wages between the governorates.
Table 1 shows the descriptive statistics of the dependent, the independent, and the
control variables. The statistics indicate a high variation in the logarithm of the total monthly
wages among Jordanians. In contrast, the percentages of SRs record medium variation since
the difference between the governorates that received the high percentages, and which received
small percentages.
Table 1 Data Descriptive
Variable
Unit
Obs.
Mean
Std. Dev
Min
Max
The logarithm of
total monthly
wages
Jordanian
Dinar (JD)
4,550
6.859688
0.6081402
2.890372
10.95867
Syrian refugees
Percentage
6,150
10.05691
9.721091
2
41
age
Number
6,150
34.7252
10.41959
17
64
Years of schooling
Number
6,109
11.86528
3.580676
0
26
Total individuals in the
household
Number
6,150
5.795772
2.277735
1
17
Marital status
Number
6,124
1.679131
.5125396
1
4
(0-4)
Urban/Rural
Number
(1-2)
6,150
1.294959
.4560616
1
2
Formal/informal job
Number
(0-1)
4,892
.800695
.3995186
0
1
Governorate
Number
6,150
19.7935
7.776389
11
34
On the other hand, Fig. 1 shows the normality result of the logarithm of the total
monthly wages using histogram, kdensity graph. It indicates the rates of total monthly log
wages are normally distributed. Fig. 2 shows the relationship between the percentage of the
SRs in the Jordanian governorates and the logarithm of total monthly wages. In addition, it
shows the average of logarithm wages in each governorate. It indicates a positive relationship
between the percentage of SRs and rates of wages. In governorates with a high percentage of
SRs, the rates of wages increased. Moreover, no high variation in the rates of wages between
governorates. As shown in Fig. 3 & Fig. 4, respectively, the rates of monthly wages set between
6 and 8% for both males and females it seems no differences between the two groups.
Fig. 1. Histogram-Kdensity dependent variable distribution
of the logarithm of total monthly wages (3-month)
Fig. 2. The relationship between the % of Syrian refugees
and Logarithm of total wages
G#: Governorate code: review Table 6 for details
Fig. 3. The rates of wages of males
Fig. 4. The rates of wages (Females)
5. Methodology
To investigate the impact of SRs on the income of natives, this study considers the
methodology of Difference in Difference to investigate the causal effect of exposure to the new
intervention on the treated groups. For instance, measuring the average difference in different
outcomes (economic, social, health, labor market, etc) because of the effect of immigration on
the hosting countries at different levels of cities, items, services, and individuals. DID is a
quasi-experimental method. It measures the percentage change in the outcome before and after
the intervention between the treated and the control groups. The DID estimator is determined
in the equation. 1 (Zeldow & Hatfield, 2019).
(1)
On the other hand, this approach is based on a group of assumptions that should be
tested for the selected data before applying the methodology. The first assumption is the
parallel trend assumption. It states that the average outcome for both control and treated groups
had the same trends across the time during the absence of the intervention. It is very essential
to ensure the internal validity of the DID model. (Zeldow & Hatfield, 2019). In this study, the
available data about the rates of income in all governorates before the intervention of SRs is
just for one year (2010). Therefore, it is difficult to provide accurate evidence about this
assumption. Accordingly, the study uses the average incomes of natives from employment in
all governorates in Jordan for three years before the crises of SRs (2006, 2008, 2010). The data
source is the Jordanian Department of Statistics. As shown in Fig. 8 in Appendix (B), control
governorates and treated governorates have the same trend of natives' income rates before the
intervention of SRs in Jordan.
The second assumption indicates the intervention unrelated to the outcome, which
means that the allocation of intervention is not determined by the outcome. it means that the
allocation of SRs in Jordanian governorates is not related to wages or economic returns in these
governorates or as a result of governmental procedures to distribute refugees in Jordan. It is
about the geographical factor since most SRs live in the governorates that close to the border
with Syria and due to the self-selection by SRs (Stave & Hillesund, 2015). On the contrary, in
Germany, for instance, the immigrants were allocated by the government to different regions
to ensure equal distribution over the country due to economic considerations (Glitz, 2011).
6. Results and Empirical models
The study revolves around three sections to investigate to what extent the income of
natives affected by the inflow of SRs in Jordan, and to what extent that contributed to increasing
the inequality between them. The first section studies the impact of SRs on the income of
natives at governorate levels to investigate the variation of effects between the governorates.
The second section investigates the impact at gender level to measure the changes in income
between men and women among governorates. The third section studies the impact on the
individual fixed effect levels i.e among all employed Jordanians aged between 17 and 64 and
takes into consideration all fixed variables over time.
6.1- First section: The Impact of Syrian Refugees on The income at The Governorate
Level
This section investigates the impact of SRs on the income of natives between the
governorate to measure the level of changes between governorates in Jordan due to the inflow
of SRs by designing two models. The first without taking into consideration the effect of other
control variables as shown in the equation. 2. While the second model adds a group of control
variables and fixed effect variables to the regression to investigate their effects as shown in
equation (3).
(2)
• Yigt: The independent variable, income for individual i at time t, and it is measured by
the total wages.
• DT: The time dummy variable indicates 0 for the year before the intervention (2010)
and 1 for the year after the intervention (2016).
• DY: The intervention or treatment dummy variable that indicates 0 for the control
group and 1 for the treatment group.
• G: The governorate fixed effect.
• uit: The error term for individual i at time t.
(3)
• Xigt: A vector of control variables that include (years of schooling, number of
individuals in the household, and age).
• Zigt: A vector of fixed-effect variables that includes (urban/rural fixed effect,
formal/informal job fixed effect, and marital status fixed effect).
6.1.1- The Assumption of The Theory
By reference to Wooldridge (2016), the study uses the statistical hypothesis test to
define the relationship between the existence of SRs in Jordan and the income inequality of
natives. Moreover, understand how to interpret the findings of each specified model of the
results. the test is based on two hypotheses. First, the null hypothesis (H0) assumes no
relationship between the SRs and income inequality in Jordan, i.e. insignificant relationship
when (β1=0) Second, the alternative hypothesis (H1) means that there is a relationship between
the SRs and income inequality between natives. It considers the two-sided alternative (β1≠0,
β1>0, β1<0).
To define βo and β1, the relationship between the two variables is defined as the following:
Y= βo+ β1Xi
So, β1 measures the changes of the expected value of Y due to the partial effect of Xi when
other independent variables are controlled.
Ho=β1=0 the variables are insignificant (fail to reject H0 and reject H1) i.e; statistically,
the variables not fit this Model (P(value)>0.05), economically, there is no relation between
the dependent variable and other variables.
H1= β1≠0 the variables are significant (fail to reject H1 and reject H0) i.e; statistically,
the variables fit this Model (P (value)<=0.05), economically, there is a relation between the
dependent variable and other variables.
6.2.2- Results of The First Section
For the first model, as shown in table (3) in the first column of the model (1), the
coefficient of intervention (Treatment dummy) indicates differences in the rates of income
between both control and treated governorates before the inflow of SRs into Jordan. The rates
of income in the treated governorates are less than the control governorates. The coefficient of
time (Time dummy) indicates increasing in rates of wages over time in the control
governorates. On the other hand, the coefficient of interaction between time and treatment
indicates a positive significant impact for SRs on the rates of natives income in both control
and treated governorates at 5% confidence interval, since the value of DID estimator indicates
an increase in the rates of income by 8% in the treated governorates more than in the control
governorates. These results could be discussed based on related theories and studies. For
instance, Bahcekapili & Cetin (2015) concluded that immigration with free trade and
movement increases the return to scale in the hosting countries. Then, the growth rate is more
than the normal rate. As a result, the growth rate increases the return on human capital as well
as wages. Some hosting countries of SRs witnessed the same impact of SRs on the rates of
income. For instance, in Germany, the SRs have a positive and significant result on German
employment rates and national wages. In Jordan, only 10 % of SRs have permits to work in the
formal sector which is majorly fully by Jordanians who are highly educated. Therefore, the
competition between Jordanians persisted which makes them attain more wages (Stave &
Hillesund, 2015), In addition, the increase in the international aids and humanitarian work to
support SRs in Jordan contributed to increasing aid economy and creating more formal job
opportunities for Jordanians with higher wages. Besides, the existence of SRs in Jordan led to
an increase in the demand to meet the needs of SRs from the Jordanian market especially in
the sectors of services and housing (Carrion, 2015). Also, the percentage of unemployment
between Jordanians is mainly between youth less than 25 who still not joint the labor market.
Finally, it could say that the SRs impact the wages of natives indirectly since the SRs do not
affect the allocated of natives in the labor market. Consequently, no major change in the
percentage of natives' share from the Jordanian labor market especially in sectors of
administration, education, health, and services (Stave & Hillesund, 2015).
For the second model, when the control groups are added to the model. The results in
column one of the model (2) as shown in Table 3 indicate no changes in the significance of the
impact of SRs on the rates of income between the governorates. However, it shows a significant
impact from most control variables (both fixed and not fixed variables) on the rates of income.
For instance, education has a positive relationship with the rates of natives' income. Increasing
in years of schooling by one year increases the income rates of natives by 3.9%. The raising of
social investment in education increases the financial return of education and decreases the
disparities of income between communities since the inequality of school enrolment leads to
variations in earnings and income inequality (Winegarden, 1979). Also, Solow illustrates
income growth as a result of achieving convergence in physical capital. On the other hand, in
the mid of 1970s in Europe, the decrease in education rates was the reason behind declining in
income growth (O'Neill, 1995). In Jordan, about 64% of labor force participation is mainly
high educated persons (Stave & Hillesund, 2015).
Moreover, age has a low significant relationship with increases in income, an increase
in age by one year increases the income rates by only 9.3%. Since persons acquire more skills
and experience in the work overtime which normally increases their returns. In Jordan, the
percentage of labor force participation is higher for persons who are aged more than 24 (Stave
& Hillesund, 2015). In contrast, the family size has a negative relationship with rates of income
as increases in the number of individuals in the household by one person decreases the rates of
income by 12%. The theory states that the opportunities of individuals in households in the
labor market. If they obtain opportunities, then they contribute significantly to increasing the
earnings (Groesbeck & Israelsen, 1994). In Jordan, the increase in unemployment rates
contributes to decreasing the opportunities for work and earning as well.
Table 2 Results of The First Level, The Impact of Syrian Refugees on The Income of Natives at The
Governorate Level
Variables
(Model 1)
(Model 2)
The logarithm of total monthly
wages
The logarithm of total monthly
wages
Coefficient
Std. Err
Coefficient
Std.Err
Interaction between time
and treatment
0.0809**
0.0350
0.0919***
0.0332
Time dummy
0.282***
0.0237
0.209***
0.0231
Treatment dummy
-0.00242
0.0733
-0.0630
0.0697
Gov_Amman
0.0172
0.0614
0.0124
0.0590
Gov_Balqa
0.0493
0.0651
0.00445
0.00620
Gov_Zarqa
-0.0441
0.0492
-0.00676
0.0470
Gov_Madaba
Omitted
Omitted
Gov_Irbid
-0.0409
0.0469
-0.0314
0.0447
Gov_Mafraq
-0.0208
0.0508
-0.0287
0.0483
Gov_Jarash
0.0724
0.0669
0.0223
0.0637
Gov_Ajloun
0.0983
0.0712
0.0669
0.0678
Gov_Karak
0.0583
0.0639
-0.0111
0.0607
Gov_Tafileh
0.0622
0.0714
0.00169
0.0680
Gov_Maan
-0.0198
0.0575
0.00570
0.0545
Gov_Aqba
Omitted
Omitted
Years of schooling
0.0337***
0.00247
Number of individuals in
the household
-0.00539
0.00395
age
0.00548***
0.00111
Rural fixed effect
Omitted
Urban fixed effect
0.0285
0.0194
Informal job fixed effect
-0.290***
0.0243
Formal fixed effect
Omitted
Single
0.0492
0.0777
Married
0.160**
0.0759
Divorced
Omitted
Widow
0.179
0.185
Constant
6.700***
0.0590
6.071***
0.105
Observations
4,550
4,550
R-squared
0.072
0.169
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
For the informal sector, as shown in the results of both models (1) and (2) in Table 3.
The rates of income in the informal sector are less than the rates of income in the formal sector
by 2.9%. To investigate if this difference in the rates of income is related to the existence of
SRs widely in the informal sector or not. The study investigates the impact of SRs on the rates
of income of informal Jordanian employees in all governorates as the equation. 4 illustrates.
(4)
• F: A dummy variable indicates 1 for formal work and 0 for informal work.
As shown in Table 3, the SRs have a significant impact on the rates of income in both
sectors of work formal and informal. The rates of income increased in the formal sector by
3.9% more than the income rates in the informal sector. the reason behind that could be
attributed to the decreasing in the percentage of Jordanians in the informal sector who crowded
out by SRs. For instance, a 30% percentage of Jordanian labor in the construction and
agriculture sectors crowded out by 40% of SRs, contrary to the situation before the crises (Stave
& Hillesund, 2015). In Turkey, the same scenario has happened, and the SRs affected the
Turkish who work informal sector and irregular workplaces significantly negative (Del Carpio
& Wagner, 2015).
Table 3 Impact of Syrian Refugees on The Work Sector (formal/ informal)
Variables
The Logarithm of total monthly wages
Coefficient
Std. Err
Interaction between time and
treatment
-0.065**
0.03460
Time dummy
0.2816**
0.02338
Treatment dummy
-0.0339**
0.07874
Interaction between time,
treatment, and formality dummy
0.3945**
0.03639
6.2- Second Section: The Impact of Syrian Refugees on The Income at Gender Level
This section investigates the impact of SRs on the income at the gender level in the
Jordanian governorates. It includes two models. The first does not take into consideration the
effect of control variables as shown in the equation. 5. While the second model investigates the
effect of control variables both fixed and not fixed over time as shown in the equation. 6.
(5)
• M: A dummy variable of males that indicates 1 for male and 0 for females
(6)
Table 4 Results of The Second Level, The Impact of Syrian Refugees on The Incomes of
Natives at gender
(Model 1)
(Model 2)
Variables
The logarithm of total
monthly wages
The logarithm of total
monthly wages
Coefficient
Std. Err
Coefficient
Std.Err
Interaction between time,
treatment, and Males
0.0628**
0.0320
0.0622**
0.0314
Time dummy
0.233***
0.0209
0.233**
0.0207
Treatment dummy
-0.0666
0.0687
-0.0410
0.0676
Gov_Amman
-0.0502
0.0582
-0.00949
0.0579
Gov_Balqa
-0.00953
0.0618
-0.0138
0.0609
Gov_Zarqa
-0.0552
0.0467
-0.0347
0.0462
Gov_Madaba
Omitted
Omitted
Gov_Irbid
-0.0439
0.0444
-0.0447
0.0439
Gov_Mafraq
-0.00342
0.0481
-0.0309
0.0473
Gov_Jarash
0.000664
0.0635
-0.0119
0.0625
Gov_Ajloun
0.0471
0.0675
0.0572
0.0665
Gov_Karak
0.00384
0.0606
0.0117
0.0596
Gov_Tafileh
0.0249
0.0677
0.0155
0.0667
Gov_Maan
0.0249
0.0544
0.0112
0.0535
Gov_Aqba
Omitted
Omitted
Males
0.277***
0.0224
0.275***
0.0223
Years of schooling
0.0482***
0.00248
0.0423***
0.00251
Number of individuals in
the household
-0.0125***
0.00371
-0.00674*
0.00388
age
0.01000***
0.000916
0.00682***
0.00109
Rural fixed effect
Omitted
Urban fixed effect
0.0284
0.0190
Informal job fixed effect
-0.296***
0.0239
Formal fixed effect
Omitted
Single
-0.0494
0.0765
Married
0.0347
0.0750
Divorced
Omitted
Widow
0.257
0.181
Constant
5.711***
0.0773
5.853***
0.105
Observations
4550
4550
R-squared
0.171
0.201
As shown in Table 4 in column 1 of model 1 and model 2, the results indicate a
significant impact for the SRs on the native males more than females with differences of 6.3%
between the treated and control groups. Accordingly, the influx of SRs contributed to an
increase in the gap between males and females. However, it could not be considered as the
major reason behind increasing the gap of income between males and females. After the influx
of the SRs, the participation rate of the labor force did not change between Jordanians, it stayed
around 66%. In addition, the labor force of Jordanian females remained about 18% before and
after the crisis. Also, the participation of female refugees in the Jordanian labor market is not
high, it does not exceed (7%). (Stave & Hillesund, 2015). The studies about the impact of
immigration on income inequality indicate the immigrants contributed to creating a gap of
income between males and females especially when the labor market in the hosting is highly
competitive. In Jordan, the competition of employment is more in the low-skills labor market
(Lumpe & Weigert, 2010). Therefore, the impact of SRs is more on other immigrants and low-
skill Jordanian workers since the Jordanian who works in the informal sector have less income
compared to who work in the formal sector by 2.9% as shown again in the results of column 1
of model 2 in table 4 (Carrion, 2005). Thus, the SRs have a low impact on income inequality
at the gender level. Also, other factors such as education, age, and the number of individuals
in the household increase the gap of income between males and females.
On the other hand, the results of section 3 have been added to the appendix A for further
analysis and information as Table 5 illustrates.
7. Conclusion
This paper deals with one of the significant movements of immigrants in the modern
era. It emphasizes the impact of SRs on the gender income inequality between natives in
Jordan. The study uses panel data for two years, a year before the influx of SRs into Jordan
2010 and year after 2016. The dependent variable is the income of individuals and the
independent variable is the percentage of SRs in each governorate.
However, other studies concluded no effect for SRs in the Jordanian labor market such
as (Malaeb & Wahba (2018); Fakih & Ibrahim (2016). This paper concluded a positive effect
same to some kinds of literature and reports such as (Fallah et al. (2019); (Tiltnes et al. (2019);
(Stave & Hillesund (2015). Using the methodology of Difference in Difference, the results of
fixed-effect regression at the governorate level showed a positive significant impact of SRs on
the income of natives, and the value of the DID estimator indicates differences between
governorates up 8% increase in rates of income in the treaded governorates more than the
control groups. Thus, the positive impact on income rates increases the rates of income
inequality between the treated and control governorates. By gender, the results also showed
that the SRs increased the rates of income of males rather than females by 6.3%. At the
individual fixed effect level, the impact of SRs is insignificant at 5%. On the other hand, the
results showed that the SRs contributed to an increase in the rates of income in the formal sector
by 3.9% more than the rates of income in the informal sector. Furthermore, males work in the
informal sector have rates of income less than those who work in the formal sector by 2.9%.
Based on the results, the SRs affected positively in the Jordanian labor market through
the increase in the rates of income. However, this increasing widened the gap of income of
native between the governorates from a side and between males and females from another side.
Therefore, they increase income inequality. However, the impact of SRs on income inequality
whether by governorates or by gender is minor since the results did not show a huge difference
in rates of income between governorates and between gender. Therefore, gender income
inequality could be related to the distribution of jobs and the huge gap in the labor participation
force between males and females. Regarding the work sector, the positive impact for the SRs
on the rates of income in the formal sector is indirect impact related to increasing the demand
to meet the needs of SRs in Jordan as well as the increase in the economy of aid which created
formal jobs opportunities for Jordanian. Also, most SRs work in the informal sector, while 10%
work formally, so their direct impact is on the Jordanian in the informal sector (Carrion, 2015).
Overall, the income inequality of natives will not be affected drastically by immigrants
(Card, 2009). If both immigrants and natives in the labor market have the same characteristics,
then immigrants contribute significantly to increase income inequality (Xu et al., 2016).
Finally, collecting further information about SRs in Jordan. Then, study their impacts
on other economic factors will help significantly to understand their impacts. Therefore, that
would support finding suitable solutions for related problems, these solutions could achieve
the best utilization from the existence of SRs in Jordan and benefit from their skills in the labor
market. Also, encouraging pass work policies by the government which could support the
existence of SRs in the country.
8. References
Bahcekapili, C., & Cetin, B. (2015). The impacts of forced migration on regional economies:
The case of Syrian refugees in Turkey. International Business Research, 8(9), 1.
Borjas, G. J. (1999). The economic analysis of immigration. In Handbook of labor
economics (Vol. 3, pp. 1697-1760). Elsevier.
Card, D. (2009). Immigration and inequality. American Economic Review, 99(2), 1-21.
Card, D., Dustmann, C., & Preston, I. (2009). Immigration, Wages, and Compositional
Amenities NBER Working Paper No. w15521.
Chatzichristou, M. (2018). The impact of Syrian refugees on the German Labour Market.
Erasmus School of Economics.
Carrion, D. (2015). Syrian refugees in Jordan: confronting difficult truths. London, UK:
Chatham House, The Royal Institute of International Affairs.
Columbia University Mailman School of Public Health, (2019). Difference in Difference
estimation. Retrieved from https://www.mailman.columbia.edu/research/population-
health-methods/difference-difference-estimation
Del Carpio, X. V., & Wagner, M. (2015). The impact of Syrians refugees on the Turkish
labor market. The World Bank.
Esen, O., & Oğuş Binatlı, A. (2017). The impact of Syrian refugees on the Turkish economy:
Regional labor market effects. Social Sciences, 6(4), 129.
Fakih, A., & Ibrahim, M. (2016). The impact of Syrian refugees on the labor market in
neighboring countries: empirical evidence from Jordan. Defence and Peace
Economics, 27(1), 64-86.
Fallah, B., Krafft, C., & Wahba, J. (2019). The impact of refugees on employment and wages
in Jordan. Journal of Development Economics, 139, 203-216.
Glitz, A. (2012). The labor market impact of immigration: A quasi-experiment exploiting
immigrant location rules in Germany. Journal of Labor Economics, 30(1), 175-213.
Green, A. G., & Green, D. A. (2016). Immigration and the Canadian earnings distribution in
the first half of the twentieth century. The Journal of Economic History, 76(2), 387-
426. doi:http://dx.doi.org/10.1017/S0022050716000541
Groesbeck, J. D., & Israelsen, L. D. (1994). The relationship between household size, real
wages, and labor force participation rates of men and women. Economics Research
Institute Study Paper, 94(13), 1.
Krafft, C., & Assaad, R. (2018, April). Introducing the Jordan Labor Market Panel Survey
2016. In Economic Research Forum Working Paper Series No (Vol. 1186).
Kuhn, P., & Wooton, I. (1991). Immigration, international trade, and the wages of native
workers. In Immigration, trade, and the Labor Market (pp. 285-304). University of
Chicago Press.
Kuyumcu, I, & Kösematoğlu, H. (2017). The impacts of the syrian refugees on turkey’s
economy. Journal of Turkish Social Sciences Research, 2(1).
Lumpe, C., & Weigert, B. (2010). Immigration, education and wage inequality. Journal of
Economic Integration, 592-612.
Malaeb, B., & Wahba, J. (2018, April). Impact of refugees on immigrants’ labor market
outcomes. In Economic Research Forum Working Paper Series No (Vol. 1194).
O'Neill, D. (1995). Education and income growth: Implications for cross-country
inequality. Journal of Political Economy, 103(6), 1289-1301.
Stave, S., & Hillesund, S. (2015). The impact of Syrian Refugees on The Jordanian Labour
Market. International labour Organization.
Tiltnes, A., Zhang, H., & Pedersen, J. (2019, April). The living conditions of Syrian Refugees
in Jordan. FaFo Research Foundation.
UNHCR, Syria Regional Refugee Response. (2020, March). Retrieved from
https://data2.unhcr.org/en/situations/syria/location/36
Winegarden, C. R. (1979). Schooling and income distribution: Evidence from international
data. Economica, 83-87.
Wooldridge, J. M. (2016). Introductory econometrics: A modern approach. Nelson
Education.
World Food Program USA, 10 facts about Syrian refugees Crises in Jordan. (2019,
November). Retrieved from https://www.wfpusa.org/stories/10-facts-about-the-
syrian-refugee-crisis-in-jordan/
Xu, P., Garand, J. C., & Zhu, L. (2016). Imported inequality? Immigration and income
inequality in the american states. State Politics & Policy Quarterly, 16(2), 147-171.
Zeldow, B., & Hatfield, L.A. (2019). Confounding and Regression Adjustment in Difference-
in-Differences.
9. Appendices
9.1- Appendix (A)
9.2-1. Third Section of The Results: The Impact of Syrian Refugees at Level of
Native Individuals Fixed Effect.
This section illustrates the impact through three models at the level of individual fixed effect,
the first regression studies the relationship between the independent and dependent variables
without considers the effect of other control variables as shown in the equation. 7.
(7)
The second model examines the impact of SR on income with the existence of a group of
control variables as shown in equation. 8.
(8)
The third model adds fixed effect variables to the regression as shown in equation. 9.
(9)
Table 5 Results of The First Level, The Impact of Syrian Refugees on The Incomes of Natives
(Model 1)
(Model 2)
(Model 3)
Variables
The logarithm of total
monthly wages
The logarithm of total
monthly wages
The logarithm of total
monthly wages
Coefficient
Std. Err
Coefficient
Std. Err
Coefficient
Std. Err
Time dummy
0.307***
(0.0219)
0.411***
(0.0820)
0.209***
(0.0230)
Treatment dummy
-0.0913***
(0.0228)
Interaction between
time and treatment
0.0635*
(0.0321)
0.0650*
(0.0322)
0.0902***
(0.0331)
age
-0.0158
(0.0115)
0.00540***
(0.00110)
Years of schooling
-0.000619
(0.00937)
0.0336***
(0.00246)
Number of
individuals in the
household
-0.00816
0.00635
-0.00553***
(0.00392)
Rural fixed effect
Omitted
Urban fixed effect
0.0309*
(0.0184)
Informal job fixed
effect
-0.289***
(0.0239)
Formal fixed effect
Omitted
Single
0.0485
(0.0775)
Married
0.158**
(0.0757)
Divorced
Omitted
Widow
0.180
(0.184)
Constant
6.701***
(0.00990)
7.256***
(0.381)
6.176***
(0.0910)
Observations
4,550
4,550
4,550
Number of
individuals
3,075
3,075
3,075
R-squared
0.233
0.235
0.168
As shown in Table (5) results indicate to the SR have insignificant impact on the income of
natives Jordanian at 5% confidence interval.
9.2- Appendix (B)
Table 6 The Distribution of Syrian Refugees Over Governorates in Jordan (2016)
Governorate
Code
Total
Population
Syrian refugees
Population (UNHCR)
% SR
(UNHCR)
Amman
11
2816200
180026
6%
Balqa
12
345500
19147
6%
Zarqa
13
959100
109862
11%
Madaba
14
132900
11024
8%
Irbid
21
1243900
136496
11%
Mafraq
22
386500
158739
41%
Jarash
23
166600
9706
6%
Ajlun
24
123700
7930
6%
Karak
31
222500
8704
4%
Tafiela
32
67700
1501
2%
Ma'an
33
101200
7595
8%
Aqaba
34
132200
3372
3%
Total
6698000
654102
10%
Fig. 5. The labor participation rate for men by different
communities
Fig. 6. The labor participation rate for women by different
communities
Fig. 7. The percentage of employment by sector
Fig. 8. Rates of Individual's Income in Control and traeted group
(Parallel Trend Asuumption)
0
5000
10000
15000
20000
25000
30000
35000
2006 2008 2010
JD
Years
Rates of Individula's Income in Control and Treated
Governorates
Control Treated
