Next Article in Journal
Cost-Sharing Contracts for Energy Saving and Emissions Reduction of a Supply Chain under the Conditions of Government Subsidies and a Carbon Tax
Previous Article in Journal
Integrated Supply Network Maturity Model: Water Scarcity Perspective
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Migrants’ Role in Enhancing the Economic Development of Host Countries: Empirical Evidence from Europe

1
East European Center for Research in Economics and Business, Department of Marketing and International Economic Relations, Faculty of Economics and Business Administration, West University of Timisoara, 16 Pestalozzi Street, 300115 Timisoara, Romania
2
Center for Banking and Financial Research, Department of Finance, Banking and Economic Analysis, Faculty of Economics and Business Administration, University of Craiova, 13 A I Cuza Street, 200585 Craiova, Romania
3
Department of Tourism and Hospitality Management, Tourism Faculty, Adnan Menderes University, 5 Candan Tarhan Bulvari, 09400 Kusadasi Aydin, Turkey
4
Research Group in Social and Economic Complexity, Department of Sociology, Faculty of Sociology and Psychology, West University of Timisoara, 4 Vasile Parvan Street, 300223 Timisoara, Romania
*
Author to whom correspondence should be addressed.
Sustainability 2018, 10(3), 894; https://0-doi-org.brum.beds.ac.uk/10.3390/su10030894
Submission received: 24 February 2018 / Revised: 13 March 2018 / Accepted: 14 March 2018 / Published: 20 March 2018
(This article belongs to the Section Economic and Business Aspects of Sustainability)

Abstract

:
This research examines several modellers of immigration flows deployed within the European Union (EU), as well as their economic consequences upon the most targeted ten migrant receiving countries. The paper’s aim is to identify specific ways in which migrants can contribute to host countries’ sustainable development through positive spillover upon natives, labour market performance, and the overall economic activity. A set of methods and macro-econometric models, based on country fixed effects, spatial analysis, and structural equations modelling, was applied on a balanced panel formed by ten EU host economies. We analysed distinctly the labour and humanitarian (asylum seekers) migration flows, considered throughout two separate time periods, namely 2000–2015 and 2000–2019 (2019 being the deadline for Brexit negotiations). The results highlight that the immigration flows were mainly shaped by labour market outcomes, while the primary positive immigration impact was induced upon the gross domestic product (GDP) per capita and employment levels, both for natives and the foreign population.

1. Introduction

International migration flows continue to increase year-by-year, and they are currently above the peak level registered in 2007 before the global economic crisis, with important effects for migrant sending and receiving economies. Immigration leads to significant changes in labour market performances, both for natives and the foreign-born populations, which largely vary from one country to another.
An attempt to understand the impact of international migration on European Union (EU) countries is, however, a challenging analytical issue since the effects of migration may vary across time and place. Equally importantly, robust and timely migrant data are not always readily available. Additionally, the existing migration data have certain imperfections, and this opens a debate on the conclusiveness of the migration studies. There is, nevertheless, a need for continuous assessments of the economic effects induced by migration into host countries to offer new evidence on particular ways in which migrants can positively contribute and enhance sustainable economic development.
Thus far, a number of studies have been undertaken to depict the adverse and beneficial consequences of international immigration, particularly on the population of a single receiving country. Some of these studies concluded that migrants increased the wages of national workers [1,2,3], while others claimed that immigration negatively affected wages in the receiving country [4], yet others concluded that increased immigration was negatively associated with employment rates in European countries [5]. Studies based on regional data further enhanced the current knowledge on immigration effects [6,7,8]. Rodriguez-Pose and Vilalta-Bufi [9] (p. 1) highlight that “factors such as the matching of educational supply and local labour needs, job satisfaction, and migration may have a stronger connection to economic performance than the traditional measures for the European regions”. Different from previous studies, we processed the macro-econometric models on national data rather than on those collected by regions, for a better quantification of the effects induced by immigration upon natives and labour market outcomes (as suggested by some migration specialists) [10].
Hence, the present research aims to analyse the effects induced by economic (labour) and humanitarian (asylum seekers) migration upon labour markets and sustainable economic development of the most targeted EU-10 host countries according to Eurostat [11] and the United Nations High Commissioner for Refugees—The UN Refugee Agency (UNHCR) [12] data (namely Germany, France, the United Kingdom, Austria, Sweden, Italy, Spain, Belgium, Denmark, and Finland). We considered two distinct time periods for analysis, namely 2000–2015 and 2000–2019. These series were taken into account for two reasons: (i) first, to capture the long-run effects of immigration (2000–2015 period) to better assess the specific ways in which migrants contribute to sustainable economic development; (ii) second, to take into account the significant changes induced by the disintegration risk brought by the Brexit decision (2000–2019 timespan), 2019 being a milestone year for a possible outcome of the negotiations being held between the United Kingdom (UK) and EU-27.
In terms of the linkages between migration and sustainable economic development, this research examines the migration economic effects upon host countries through the underlying conditions that lead to employment creation, additional income, a poverty risk reduction, improved educational background, and living standards. Therefore, in our dataset, we included specific variables to capture these credentials in addition to the Gross Domestic Product (GDP) per capita and per person employed such as, for example, employment and unemployment rates (total, long term, foreign population); earnings and earnings’ dispersion; at-risk-of-poverty rate; educational attainment and participation rates; and life expectancy. However, the focus is primarily on the GDP per capita since it is a widely used key indicator of living standards and economic welfare. Moreover, we focused on a longer period of time, to better capture a possible trade-off between rapid economic growth and long term growth.
The paper is structured as follows: After the introduction, Section 2 covers the issues relating to basic migration concepts and main economic consequences, along with the recent developments of international migration globally and in Europe, to orient the reader about the current situation. Section 3 details the research methodology, with a focus on the macro-econometric models developed and the data used in the empirical analysis. Section 4 includes the research results and their implications, and the final section comprises the concluding remarks.

2. The Literature Review

2.1. Migration: A Conceptual Framework

Migration is often defined in a broad sense through “a temporary or permanent change in residence” [13] (p. 49), and more specifically as the “movement of a person or a group of persons from a certain geographical unit towards another, crossing political and administrative frontiers, in their will to establish in a different place than the origin one” [14] (p. 6). Within this framework, there are no restrictions to the geographical distance covered by migrants or to the voluntary or undesired nature of the migration act. However, there is a clear distinction between internal and international migration by relating these concepts to national borders [15]. Thus, in this paper, the focus is on international migration which implies a cross-border movement of persons (labour force, refugees, asylum seekers). The complexity of the migration phenomenon is significantly increased under the impact of globalization. In our analysis, we take into account two aspects of different migrants’ credentials: better labour conditions and living standards (known as labour migration or economic migration) and international protection (humanitarian migration). Economic migration (or labour migration) arises from various motives and is determined by new employment opportunities, improved working conditions, and wage differentials between migrant sending and receiving countries. Humanitarian migration includes “all recipients of protection—whether refugee status, temporary protection, subsidiary protection, etc.” [16] (p. 4). In this paper, we refer to refugees in lines with the Geneva Convention on the Status of Refugees (1951): “a person who is outside his or her country of nationality or habitual residence; has a well-founded fear of being persecuted because of his or her race, religion, nationality, membership of a particular social group or political opinion; and is unable or unwilling to avail him—or herself of the protection of that country, or to return there, for fear of persecution” [12] (p. 5). An asylum applicant is an individual seeking international protection that has formally submitted a request for asylum on the grounds of a refugee status, subsidiary, or humanitarian temporary protection, but has not yet completed the asylum procedure [16] (p. 4). By comparison, from an economic perspective, a migrant leaves his country of origin due to various reasons, mainly in search of better employment opportunities, living standards at destination, family reunification, or study, and continues to benefit from the protection of his government even when abroad.

2.2. Migration: Consequences

As the conventional economic theory suggests, the impact of migration on receiving countries depends on a number of factors, including the skill mix of the immigrants and the characteristics of the host country.
In the matter of welfare effects induced by international migration for EU Member States, Aubry et al. [17] quantified these economic consequences focusing on the impact upon natives’ living standard. Thus, the authors developed a model that combines the interactions between the labour market, fiscal aspects, and market size (such as the changes in the variety of goods available for consumers) under the impact of international migration, along with the trade relations between sending and receiving economies. Their results reveal that the immigration inflows had positive effects upon 69% of the native population for 34 Organization for Economic Cooperation and Development (OECD) members and on 83% of the natives from 22 wealthier OECD economies.
Immigration impacts upon a host country’s labour market also refer to the effects induced by the variations in the labour supply of a certain qualification, associated with a consequent change in the labour market’s equilibrium. The distributional effects of international migration are essential, one of its major repercussions consisting of a significant change in the size and structure (skills) of the labour force in both the sending and receiving economies. The effects are amplified when there is an associated change in skills distribution among the labour force [18]. For example, this is the case in a country confronted with large immigration inflows with an average qualification level higher (or lower) than that of native workers. Thus, the changes induced in the structure of the labour force have direct effects on inequality by changing the percentage of persons with high or low wages within the economy. Moreover, the immigration effects on natives’ wages and employment perspectives have been largely debated in terms of trying to answer the question of whether the foreign immigrants’ arrival in the host countries generates adverse consequences, including the emigration of natives from the affected areas [19]. Most studies highlighted a reduced impact in this regard [20,21,22]. Immigration flows affect the size and skills of the labour force when the educational background of the immigrants does not match those of the natives. This change in the educational level of the labour force leads to a disequilibrium between labour demand and supply of different types at the existing wage and output levels in the economy. Restoring the equilibrium would imply short-term variations in wage and employment levels according to the educational attainment and could (or not) induce additional long-term changes. The economic theory highlights that economic (labour) immigration does not generate significant negative effects upon wage levels and natives’ employment degree, at least in the long-run, even when the educational attainment improves [22]. Glitz [8] points to a displacement effect of around 3.1 unemployed resident workers for every 10 immigrants that find a job, but with no detrimental effect on relative wages.
Different from others, Betz and Simpson [23] found that the overall effects of migration on natives’ happiness are very low, while Longhi et al. [6] applied a meta-analytic technique to 165 estimates from 9 recent studies for various OECD countries and assessed whether immigration leads to job displacement among native workers. Ozgen et al. [7] report that immigration is a driving force for innovation when there is a distinct composition of immigrants coming from different backgrounds, rather than the size of the immigrant population in a certain region.
Analysing the social effects of immigrants into EU-15, Tsapenko [24] (p. 451) identified that “the inflow of migrants of another culture creates serious threats to national identity and the lifestyle established in Western societies”, which will increase the concerns of host economy’s population and drop their life contentment and deteriorate their emotional state. On the other hand, “the socio-economic situation of migrants is usually substantially worse than that of natives” [24] (p. 446), hence, accurate strategies (policies) for migrants’ integration becomes necessary. An important element used for analysing the effects of international migration are the active labour market policies (ALMPs), namely, policies with positive impact over the labour market performance and integration of the unemployed [25]. Another variable with different and controversial effects over the labour market outcomes are the passive labour market policies (PLMPs) [26]. The PLMPs refer mainly to (i) maintaining the income for the unemployment period through unemployment benefits, as well as (ii) the earlier retirement which facilitates total or partial retirement of elder workers with decreased possibilities in finding a job.

2.3. Migration: Current Situation

The main statistics of international migration are offered by the United Nations, Department of Economic and Social Affairs (UN DESA) [27], the United Nations High Commissioner for Refugees—The UN Refugee Agency (UNCHR) [12], the Organization for Economic Cooperation and Development (OECD) [28,29], the International Organization for Migration (IOM) [30], and Eurostat [11]. Based on these migration databases, statistics show that, globally, the total number of migrants has increased from 94 million in 1970, to 153 million in 1990, reaching around 244 million in 2015, from the total world population of 7.3 billion (meaning 1 person in every 30 people). Still, as a share of the world’s population, the migration flow is relatively steady, representing 2.3% in 1970, and 3.3 in 2015 [30]. Around 75% of international migrants in 2015 (almost 157 million) targeted the “high-income economies…compared with 77 million foreign-born who resided in middle-income countries (about one third of the total migrant stock) and almost 9 million in low-income countries in the same year” [30] (p. 20).
Out of 244 million migrants in 2015, the largest share was for Europe (over 31%), Asia (almost 31%), and North America (over 22%) (Figure 1a). In Europe, the highest migrants’ allocation was for Western Europe (36%) and Eastern Europe (26%) with a total of 76 million migrants in 2015 (Figure 1b).
The most preferred European countries by migrants in 2015 were: Germany (12 million migrants), the United Kingdom (8.5 million), France (7.8 million), Spain (5.9 million), Italy (5.8 million), Sweden (1.6 million), Austria (1.5 million), Belgium (1.4 million), Denmark (572 thousand), and Finland (315 thousand) [27]. The main origin countries of migrants into European regions were the Russian Federation (over 6 million), Poland (3.6 million), Romania (almost 3 million), and Morocco (2.5 million) [27]. For 2016 compared to 2015, the migration flows increased by 30% in Germany, 4% in France, 7% in Belgium, 18% in Spain, 26% in Sweden, and decreased by 6% in the UK (mainly due to the BREXIT vote in June 2016), and by 5% in Denmark [28]. In the last few years, the main channel of migration was family migration, “which encompasses four main subcategories (family formation, accompanying family, family reunification and international adoption)” [28] (p. 2). The share of family migrants of the total migrant population in the most targeted 10 EU countries represents over 40% in Germany and Sweden, 50% in France, around 35% in the United Kingdom, and over 30% in Austria, Italy, Spain, Belgium, and Finland [28]. Considering the amplitude of this component, we will focus on family members’ migration in future research, based on the main findings on this paper.
The asylum seekers (considered through new asylum requests into OECD countries) in 2016 amounted to 2.8 million people, “Germany remained the top recipient, with over 720,000 applications, followed by the United States (262,000) and Italy (123,000)” [30] (p. 32). The new asylum applicants in 2016 compared with 2015 increased by 63% in Germany (coming mainly from Syria, Afghanistan, Iraq), 52% in the United States (El Salvador, Mexico, Guatemala), and 47% in Italy (mainly from Nigeria, Pakistan, Gambia) [29].
Regarding the refugees’ situation at the end of 2016, there were, registered, a total of 22.5 million displaced people, considered “the highest on record, although the annual rate of growth has slowed since 2012” [30] (p. 32). The highest number of refugees into the most targeted EU countries by migrants in 2016 are listed next. Germany received migrants as follows: Syria (375,122 people), Afghanistan (over 46,000 people), Eritrea (over 30,000), Iran (22,910 people), Turkey (19,136 people). The United Kingdom received migrants as follows: Iran (over 14,000 people), Eritrea (over 13,000), Afghanistan (over 9000 people), Syria (over 8000 people). France received migrants as follows: Sri Lanka (over 24,000) and Congo (over 14,000). Italy received migrants as follows: Eritrea (over 13,000) and Somalia (over 12,000). Spain received migrants as follows: Palestine (almost 400 people). Sweden received migrants as follows: Syria (over 34,000 people), Iraq (over 24,000), Eritrea (over 14,000), Afghanistan (over 12,000). Austria received migrants as follows: Russian Federation (over 19,000), Afghanistan (almost 12,000), Iran (over 3000). Belgium received migrants as follows: Afghanistan (over 4000) and Iraq (over 3000). Denmark received migrants as follows: Afghanistan (over 2000). Finland received migrants as follows: Somalia (almost 3000) [12]. According to the UNHCR estimations, more than half (51%) of the total number of refugees in 2016 were under the age of 18 [12]. Furthermore, around 100,000 persons risked their lives to arrive in Europe by the Mediterranean Sea (through Italy) migration corridor from January 2017 to October 2017 [31]. Migration in Europe through Italy is preferred because “… Italy has been one of the main entry points into Europe through North Africa as Libyan shores are only 290 miles away… and Italy has been seeking EU support for dealing with the migration crisis according to the Dublin Regulation” [32].
Based on these facts, our focus is on two components of the migration flows in Europe: (1st) the emigration flows from Central and Eastern Europe (CEE) countries, targeting better labour conditions at destination countries from Western and Southern Europe (economic migration); and (2nd) the refugees and asylum seekers found in Europe due to poverty, political instability, and armed assaults in Syria, Iraq, and other neighbouring countries (humanitarian migration).

3. Research Methodology and Data

By reviewing the current state of the literature on migration economic consequences and in compliance with our general research objective, we defined the following research hypotheses:
Hypothesis 1 (H1).
There is a positive direct relationship between the immigration flows (mainly economic/labour migration) and the host country’s economy/living standards (defined through the GDP per capita and per person employed).
Hypothesis 2 (H2).
There is a positive direct relationship between the immigration flows (mainly economic/labour migration) and host country’s labour market performance (defined through the employment rate (total and for the foreign population) and wage levels).
Hypothesis 3 (H3).
Education has positive effects both on attracting new migrants and enhancing migrants’ integration with positive spillovers on the economic development of host countries.
Hypothesis 4 (H4).
Labour market policies (particularly, ALMPs) have significant positive effects on migrants’ labour integration, with positive spillovers upon labour market outcomes and economic welfare.
The first step in our methodological endeavour was to determine the standardised indicators (through the standardisation method described in Equation (1)) in order to cover for data benchmarking between countries. This conversion allows developing a composite indicator based on the values accounted for each country by reference to the values of other countries included in the panel [33,34].
y i = x i m e a n s d ,
where: xi denotes crude values of the indicator and sd is the standard deviation.
The standardised indicators were further used for structural equations modelling (SEM) and macro-econometric models estimation.
Moreover, we applied linear extrapolation as a mathematical prediction of the values of considered variables during 2016–2019, based on the relationship given in the data for the same variables across time. This method was used in line with our main research objective and focused on capturing specific ways in which migrants can enhance sustainable economic development (long run effects in the Brexit framework). Therefore, the analysis was extended beyond the existing time series for 2000–2015 to capture the outcomes in the following years within the EU framework of integration or disintegration brought by the Brexit decision. Thus, we extrapolated our data for the 2016–2019 time period (2019 being the deadline for Brexit negotiations that started in 2017, and are expected/framed by the treaty to last for two years). Even though the extrapolation process is subject to more uncertainty, it can predict a future pattern in the statistical data since it is sampled periodically. Therefore, based on previous data history, we can approximate next data points. The extrapolation formula used is described by Equation (2).
y ( x ) = y 1 + x x 1 x 2 x 1 ( y 2 y 1 ) ,
where: x1, y1 and x2, y2 are the two endpoints of a linear graph and x represents the point which is to be extrapolated.
The immigration impact upon host economies, labour market outcomes, and natives was analysed through specific macro-econometric models. Looking into empirical methods applied by the economists for studying the international migration, Lozano and Steinberger [35] pointed out that the common methods used are the Ordinary Least Squares (OLS), Difference-in-Difference (DD) or First-Difference (FD), and Instrumental Variables (IV). In order to avoid the weaknesses, “the alternative methods for dealing with endogenous explanatory variables which are closely related to IV methods” (p. 18) are Generalized Methods of Moments (GMM) estimations and quantile or robust regressions. However, for a sensitivity analysis, it is recommended to apply and compare “various methods and/or various instruments” [36] (p. 10).
For this particular reason, in this empirical endeavour, we applied numerous econometric techniques, each trying to discard the other’s limits, so that the final estimations are accurate, robust, and correctly interpreted.
Furthermore, particular attention was given to a robustness check and validation in order to ensure unbiased results (numerous tests were applied to see if the selected variables properly suit the models developed and capture the linkages between migration and economic development—Appendix A, Table A5, Table A6, Table A7 and Table A8).
Migration effects on economic outcomes are extremely complex and multi-levelled, so in order to estimate a proper causal relationship, the developed models require specific attention to multiple sources of bias (endogeneity problem, potential sources, and correction methods). For example, possible immigration restrictions in Europe after Brexit and other legal credentials established by host countries could significantly impact migration economic outcomes, but these measures are difficult to quantify for longer time periods due to their unsteady nature.
At the same time, there is a large amount of empirical literature studying the migration effects on host economies that have addressed the problem of endogeneity with respect to the educational background [37] or income/wages [21,38]. One important alternative used to obtain unbiased estimators by coping with endogeneity are instrumental variables (IV estimators, Two-Stage Least Squares—2SLS) along with the Generalised Method of Moments (GMM) or robust regression, these methods were used in our econometric research design.
IV estimations (2SLS first difference) were performed by using the divergence in demand for labour, proxied through the unemployment rate of the foreign population as the instrumental variable, in line with the previous research of McKenzie and Rapoport [39]. However, the validity of the instrument is subject to uncertainty, since it is extremely difficult to find a well-grounded instrument. In this respect, Lozano and Steinberg [35] (p. 15) clearly state that “… difficulties in finding suitable instruments that can satisfy the significant restrictions of IV often limit application of this econometric tool”. At the same time, Lozano and Steinberger [35] (p. 17) mention that “utilizing IV regression techniques places high requirements on the proposed instrument; if the instrument is not highly correlated with the explanatory variable, IV estimates will be unreliable”.
In this respect, we firstly focused on within-country variation over time and developed several fixed effects (FE) models that follow the specific linear representation of panel data regression models as described by Baum [40] (p. 219) to properly analyse the immigration effects upon EU-10 receiving economies; this same configuration was also used in previous research with robust results [41,42]:
y i t = k = 1 k x k i t β k i t + ε i t ,
where: i = 1, …, N; t = 1, …, T; N is the number of panel units (countries); and T represents the number of time periods.
The fixed effects model or the Least Squares Dummy Variable (LSDV) model has the representation given into Equation (4) [40] (p. 220).
y i t = x i t β k + z i δ + u i + ε i t ,
where: xit is a 1 × k vector of variables varying between countries and in time; β represents a 1 × k vector of x coefficients; zi is a 1 × p vector of the variables that are constant in time, but vary between countries (as elements of the panel); δ represents a p × 1 vector of z coefficients; ui comprises the individual effects; and εit is the disturbance term.
These models were afterwards subject to IV estimations, dynamic GMM, and robust regression (general vs. bootstrap).
However, there was a keen focus to see how the migration performances of each of the ten EU migrant receiving countries could impact the overall outcome and change the migration patterns. For this purpose, we took into account spatial interference in the developed migration models. Therefore, to see if the observations are grouped together or randomly spread, we hereinafter applied numerous spatial analysis models (spatial lag, autoregressive) estimated through the Maximum Likelihood Estimator method (MLE), as described by Equation (5) [43].
y = λ W y + X β + u ,
where u is assumed to be classical.
At the core of our spatial models lies the inverse distance weights matrix (row–standardized) (W), with the following characteristics:
  • Period 2000–2015: dimension: 160 × 160; distance band: 0 < d ≤ 16,000; friction parameter: 1; minimum distance: 0.0; 1st quartile distance: 1.0; median distance: 1.7; 3rd quartile distance: 2.4; maximum distance: 5.3; largest minimum distance: 0.79; smallest maximum distance: 2.81;
  • Period 2000–2019: dimension: 200 × 200; distance band: 0 < d ≤ 16,000; friction parameter: 1; minimum distance: 0.0; 1st quartile distance: 1.1; median distance: 1.7; 3rd quartile distance: 2.5; maximum distance: 7.9; largest minimum distance: 0.79; smallest maximum distance: 4.42.
The existence (or absence) of spatial autocorrelation was verified with the Patrick Moran (Moran’s I) test, as given by Equation (6) [43] (p. 8).
I = R i   j ω i j i   j ω i j ( x i x ) ¯ ( x j x ) ¯ i ( x i x ) ¯ 2 ,
Further, we used structural equations modelling (SEM) in order to comprise and highlight the links (direct, indirect, total) between the determinants of immigration flows and their economic consequences. Simultaneous equations modelling represents an advanced technique of multivariate data analysis, used to design, test, and estimate causal relations between selected variables. The general representation of the SEM model is presented in Equation (7).
{ b 11 y 2 t + .... + b 1 m y m t + c 11 x 1 t + + c 1 n x n t = ε 1 t b 21 y 2 t + .... + b 2 m y m t + c 21 x 1 t + + c 2 n x n t = ε 2 t .......... .. b m 1 y 1 t + .... + b m m y m t + c m 1 x n t + + c m n x n t = ε m t ,
where t is the number of observed time periods; bij represents the yij endogenous variable’s parameters; cij are the xij exogenous variable’s parameters, I = 1, …, m; and j = 1, …, n.
The structural form of the migration model with simultaneous equations is, in fact, its initial configuration after the design stage and represents the structure of the economic process described in terms of components and linkages.
The developed SEM model provides an integrated framework of analysis for the overall migration process from a two-fold perspective (determinants-impact interactions) in line with our main research objective.
For our data, we selected a large scale of indicators as proxies for the variables of developed models that include both fundamental macro-economic aggregates as well as different measures used to quantify the labour market’s performance, educational attainment, inequality, and specific international migration indicators (economic versus humanitarian). Compared to previous studies, in our paper, we focus on analysing both economic (labour) and humanitarian migration indicators, with an impact quantified through national data compiled for ten EU migrant receiving countries. Business climate indicators of the European countries “provide an important class of information regarding the perception of current and further performance of economic activity” [44] (p. 139), with impact on the labour force dynamics, which suffered major changes [45], at the forefront of migration.
Hence, we analysed the following variables (Appendix A, Table A1):
  • Economic activity and other specific indicators: GDP per capita in Euro (GDP_cap), GDP per person employed, in USD (United States Dollars) (GDP_emp); employment rate (ER, ER_F) and unemployment rate (UR, UR_F, UR_LT) (%)—total, foreign population (F), long term (LT); part-time and temporary contracts employment rate (ER_part) (%); annual net earnings of a two-earner married couple with two children (EARN) (Euro) and earnings dispersion (Decile 9/Decile 5) (EARN_disp); expenditures on active (ALMPs) and passive (PLMPs) labour market policies (% of GDP); educational level (both general and vocational) reflected through the educational attainment for tertiary education (levels 5–8) (EDU_T), and the participation rate in education and training (EDU_part); at-risk-of-poverty rate (POV_R); life expectancy at birth (LE); business enterprise research and development expenditures for the business enterprise sector (BERD) (% of GDP); Gini coefficients (index);
  • Globalization indicators: KOF Index of Globalization—overall and economic (KOF_T, KOF_E); Export (X, mil. USD); inflows of foreign direct investment (FDI_I, mil. USD);
  • International migration indicators: flows and stock of immigrants and foreign population (IMIG), flows of asylum applicants (ASYL).
The data and associated macro-econometric models were processed in Stata.
To comprise the welfare effects and long term economic development in our macro-econometric models, we used, in some models, GDP per capita (and per person employed) as a dependent variable, while the other indicators were used as independent variables (considered for General Fixed Effects Models, and SEM Models). Moreover, we applied Multiple Fixed Effects Models (8 Models), considering the following as dependent variables: GDP_cap, GDP_emp, ER, ER_F, EDU_T, EARN, EARN_disp, and Gini.
On one side, we used IMIG and ASYL as independent variables for each model to try and capture the immigration effects in the absence of other explanatory variables. These models were processed in line with our first two research hypotheses and, thus, tried to better comprise the immigration impact upon host economies and labour markets. Also, to reveal the migration patterns and determinants/drivers towards the considered host economies, we used IMIG and ASYL, each one as dependent variables (considered for Spatial lag models).
The panel covers ten migrant receiving countries, EU-10 Member States, namely, Belgium, Denmark, Finland, France, the United Kingdom, Germany, Austria, Sweden, Italy, and Spain. We selected these countries as being the most targeted by immigrants and asylum seekers according to Eurostat [11] and OECD [28,29] data. Even though the panel considered is balanced, we must account for its relatively small size as a limit of our research, since robust and timely migrant data is not always readily available, while the existing migration data may have certain imperfections (also stated, for example, by Soon [46]).
Figure 2 graphically highlights the number of immigrants (a), asylum applicants (b), and GDP per capita levels (c) in 2015, and estimated for 2019, for the ten EU host countries. Thus, immigrants (IMIG) were at the highest level in Germany and the United Kingdom in 2015, these countries tending to be preferred even in 2019, under the Brexit negotiations deadline, even though the number slightly decreases. The asylum seekers (ASYL) were mostly found in Germany and Italy in 2015, maintaining at high levels in these countries for 2019, but especially decreasing for the UK after the Brexit. The GDP per capita is at the highest level in Sweden, for 2015, followed by an increase in Austria for the 2019.
The main databases used for collecting the data are Eurostat—European Commission [11], International Migration Database—OECD [28,29], World Development Indicators—World Bank [47], the United Nations Databases/United Nations High Commissioner for Refugees—UNHCR [12], UNU WIDER (The United Nations University World Institute for Development Economics Research)—World Income Inequality Database [48], and ETH Swiss Federal Institute of Technology Zurich [49], as described in the Appendix A, Table A1.

4. Results and Discussion

4.1. Effects of Economic and Humanitarian Migration upon EU-10 Host Economies Revealed by Macro-econometric Models and Spatial Analysis Results

In order to verify each Hypothesis drawn in Section 3 (H1–H4), in the first stage of our empirical analysis we developed five macro-econometric spatial lag (autoregressive) models of the shaping factors of immigration and asylum seeker inflows for the EU-10 host economies, during 2000–2015. The estimation results are synthesised in the Appendix A, Table A2 (for IMIG variable), Table A3 (for ASYL variable), along with descriptive statistics in Table A4. We consider these models important for a better understanding of the migration motives, migration patterns, and determinants/drivers towards the considered host economies, in order to essentially capture the way in which migration could further impact the EU-10 countries by contributing to their sustainable economic development (measured through the GDP per capita).
The spatial lag model controls for spatial autocorrelation in the dependent variable (immigration flows) and includes migration outcomes in neighbouring locations as an additional explanatory variable.
The models have been tested for positive and/or negative spatial autocorrelation based on Moran’s I test that revealed the presence of global autocorrelation and, in some cases, local autocorrelation as well, according to data found in—Figure 3a,b. Therefore, the migration performances achieved by the neighbouring locations are also essential for impacting economic welfare in the considered host economies.
LM (Lagrange Multiplier) tests are significant for the entire sample (2000–2015), thus indicating the presence of spatial dependence. Rho reflects the spatial dependence inherent in the sample data, measuring the average influence on observations by their neighbouring observations. It has a positive effect and is highly significant for the 2000–2015 sample (Appendix A, Table A2 and Table A3). Moreover, we applied several panel unit-root tests that pointed out (in some cases) that non-stationarity might be an issue, especially for the 2000–2015 time series (Appendix A, Table A5a,b). As a result, for the study of immigration determinants, we used spatial lag models, along with SEM, estimated through the MLE, since the Ordinary Least Squares (OLS) estimators could be less efficient.
The overall panel results highlight that an increase in the number of immigrants is associated with an improvement in labour market outcomes, hereby reflected through an employment increase (both total and for the foreign population or in the case of part-time/temporary contracts), along with a reduction of long-term unemployment and enhanced active labour market measures. However, the paradox is that a slight increase in GDP per capita and per person employed and of annual net earnings induces a decrease in immigration inflows that could indicate that immigrants are searching for a stable job that provides income and employment security, work-life balance, and welfare. On the other hand, the international flows of goods, services, and capital, namely FDI (Foreign Direct Investment) inwards and exports, have a positive impact upon international migration, thus inducing additional immigration inflows.
There are important positive consequences generated by the immigration inflows, along with other socio-economic variables, on GDP per capita. We focused on GDP per capita in line with our first research hypothesis (as reflected by the results obtained and synthesised in Table 1a,b).
Within this context, for the migration impact analysis, we firstly focused on within country variation over time and configured a general fixed effects model in order to assess the effects of economic and humanitarian immigration upon the GDP per capita (Table 1a,b, model 1).
Moreover, to account for the endogeneity problem inherent in migration models (as described in the methodological endeavour) we also applied other econometric procedures, namely, the 2SLS with first difference estimation (IV by using the divergence in demand for labour, proxied through the unemployment rate of the foreign population, as the instrumental variable, used before by [39]) in model 2; Arellano-Bond dynamic panel data estimation (Dynamic GMM) in a two-step process, including instruments for differenced equations and robust standard errors to control for heteroscedasticity in model 3; robust regression—general (model 4) and bootstrap (50 replications) estimation (model 5). All the results obtained throughout different econometric procedures are slightly different in size, but consistent in sign in the presence of statistical significance (statistically significant coefficients did not change their sign throughout different procedures).
The fixed effects model results show that there is evidence to attest that labour immigration has significant positive economic effects, leading to important increases in GDP per capita levels, especially in the long term (0.278 estimated coefficient for IMIG, extremely statistically significant at 0.1% level in the 2000–2015 sample and 0.297 significant at 1% level for 2000–2019). IV estimations are in line with FE results. We tested the instrument relevance by regressing the outcome on the instrument and found an F-statistic of 9.96 for the 2000–2015 sample, and 10.24 in the 2000–2019 sample, which is above the 10 level that causes concern, so as we can say that the instrument’s weakness is low. Dynamic GMM estimations are consistent with FE and IV results (even though they have a lower statistical significance) and show the same positive reverberation of labour immigration upon host countries’ GDP per capita, while the asylum seekers tend to have a negative impact. The Arellano-Bond test for zero autocorrelation in first-differenced errors gives accuracy to these estimations since it reveals no autocorrelation (order z1 = 0.84885, p = 0.3960; z2 = 0.31133, p = 0.7555; chi2(94) = 5.45 × 10−22, p = 0.9998), while the Sargan test builds up additional robustness by confirming the null hypothesis and revealing that the over-identifying restrictions are valid.
Moreover, to further deal with endogeneity and to account for the fact that the dataset might be driven by larger economies (e.g., Germany) we applied the robust regressions (general and bootstrap). The estimated coefficients are consistent in sign even though are slightly different in size (lower magnitude) and less significant, and they highlight migrants’ important role in enhancing the economic development of host countries.
Education is essential for migrants’ integration and overall economic growth, hence an increase in the educational background of host countries’ labour force (comprising both natives and migrants) towards a tertiary level has significant positive economic consequences, leading to an increase in GDP per capita (as reflected by the 2000–2015 estimations). However, for the 2000–2019 sample, these coefficients are weaker and remain statistically insignificant. At the same time, large expenditures on research and development (BERD) could frame an innovative environment with important positive spillovers on welfare and living standards (statistically significant and extremely high, positive coefficients for the BERD variable for both samples FE and IV estimations). Nevertheless, efficient active labour market policies are also essential for the labour market performance and long-term economic development, as reflected by our estimations. Thus, H3 is partially confirmed and H4 is fulfilled.
Therefore, in the framework of new accurate EU migration policies, migrants can play a decisive role in enhancing economic welfare for host countries, even in the Brexit context that brings to the front the risk of reinforcing barriers to intra-EU labour mobility. This is the case for the unskilled workers that could further lead to negative spillover effects on intellectual capital movements and economic development of migrant sending and receiving countries. At the same time, failing to properly coordinate the current refugee crisis in Europe and to integrate the asylum seekers and refugees into host economies/labour markets might prove to be detrimental since our estimations reveal negative coefficients for the ASYL variable. Thus, H1 is partially fulfilled.
Nevertheless, there are important effects induced by immigration upon host economies, reflected not only through the GDP per capita (Table 1a,b). Based on our previous results, the research was continued through developing new macro-econometric models that account for an accurate assessment of the economic consequences of international migration on host economies and labour markets (Table 2a,b).
Therefore, we processed other models that try to capture the impact of increased labour immigration and asylum seeker inflows upon GDP per person employed (model 2), total employment rate (model 3), employment rate of the foreign population (model 4), tertiary education attainment (model 5), earnings (model 6), earnings dispersion (model 7), and income inequality (model 8).
For this second set of estimations, we focused on three econometric procedures: fixed effects (within country variation) presented in Table 2a; panel 2SLS (IV) estimation—first difference (FD) reported in Table 2b (model 3 and model 4 were processed in a first difference approach without the instrument for its connection to the dependent variable) and Arellano-Bond dynamic panel data estimation—GMM in a two-step process, including instruments for differenced equation and robust standard errors to control for heteroscedasticity in Table 2c. These methods comply with our research objectives and are widely used for processing migration models [50,51] since they cover for endogeneity and discard other problems such as reverse causality.
By carefully analysing the estimated coefficients as resulted from various econometric modelling procedures, we found evidence to attest strong interdependencies between the labour immigration process (IMIG variable) and EU-10 host countries’ socio-economic performances.
Thus, the specific ways in which migrants can contribute to host economies as reflected by the 2000–2019 sample estimations is through an active labour market participation (increased ER and ER_F as proved by highly statistically significant coefficients estimated through all three methods considered) that leads to slightly higher levels of household earnings (EARN, 0.257 significant at 5% in FE; 0.103 significant at 5% in IV; 0.0976 extremely statistically significant at 0.1% in GMM). Therefore, H2 is reconfirmed and fulfilled.
Overall, additional labour immigration flows are positively embedded into host economies, leading to an improvement in living standards and welfare (GDP_cap), even in the Brexit framework.
If up until now (2000–2015 timespan) the most important positive effects were induced mainly through significant improvements in labour market outcomes both for natives and especially for the foreign population (including increases in GDP per person employed), in the long-run Brexit framework (2000–2019), the GMM estimations point out important rises in GDP per capita (0.0429 significant at 5% level), total employment rate (0.132 significant at 5% level), and annual net earnings (0.0976 extremely significant at 0.1% level) for EU-10 under the compelling effects of labour immigration. Thus, H1 and H2 are reconfirmed. The lagged variables are also positive and statistically significant, thus capturing important dynamic effects.
Also in these GMM models, the Arellano-Bond test for zero autocorrelation in first-differenced errors gives accuracy to estimations since it reveals no autocorrelation for all eight models, while the Sargan test builds up additional robustness by confirming the null hypothesis and revealing that the over-identifying restrictions are valid.
On the other hand, humanitarian migration has ambiguous effects, due to its nature, the estimated coefficients being less statistically significant and often changing in sign from one period to another. However, the need for some extremely comprehensive and efficient ways to manage this refugee crisis in Europe is a certainty. Thus, H1 is partially fulfilled.
These final estimations show that the Brexit effects will induce a multitude of changes at different levels and require an integrated framework of analysis. All of these mechanisms should, therefore, be a major priority for policy makers and governments (particularly in Europe, but also worldwide, since the UK is an important trade and investment partner for other non-European countries, like USA, China, Japan, Singapore, Canada) in order to enhance the positive credentials of this process and significantly diminish potential negative spillovers.

4.2. Effects of Economic and Humanitarian Migration upon EU-10 Host Economies Revealed by Structural Equations Modelling (SEM) Results

We further developed a new complex model based on structural equations (SEM) that aims to analyse international migration (both economic and humanitarian) from a double perspective of the shaping factors and economic consequences upon host countries, as shown in Figure 4 (general model configuration—Figure 4a, estimations for the 2000–2015 period—Figure 4b and 2000–2019 period—Figure 4c).
SEM modelling allowed us to identify the direct, indirect, and total interlinkages between the migration process and the economic performances of host countries from an integrated framework that takes into account both the major determinants of the immigration inflows (economic and humanitarian), the mediated factors (policies/adopted measures/labour market performance of employees, both migrants and natives), and final output (GDP per capita).
We consider the SEM procedures essential because the Breusch-Pagan LM test (applied when T > N) showed some evidence of cross-sectional dependence, while unit-roots may still be problematic. The estimation of the SEM model has thus been performed through the MLE method.
Therefore, we were able to verify, in this case, every Hypothesis, H1–H4, through an integrated research framework. The results obtained have combined to five an adequate identification of the direct, indirect, and total effects of international migration upon the EU-10 host countries, by taking into account the inter-relationships between the variables (Table 3 and Table 4).
The assay of the pattern has been performed partly by applying the Wald test for each equation and through the correspondence tests applied to the pattern (such as Comparative Fit Index, Tucker-Lewis Index, Size of residuals, Information criteria: Akaike, Baesian).
We also calculated the Cronbach’s Alpha coefficients (per item and total scale) which underline the accuracy of the model. The results of all these tests applied are presented in the Appendix A, Table A5, Table A6, Table A7 and Table A8.
SEM model results reconfirm our previous estimations and highlight the positive ways in which migrants’ can contribute to the host country’s welfare. Thus, when the estimated coefficients are extremely significant from a statistical point of view at a 0.01 level, there is evidence to attest that both labour and humanitarian migration can lead to GDP per capita increases (total effects). Thus, H1 is reconfirmed and fulfilled.
As for the mediated impact, we can see that an improvement in the educational background has positive effects on enhancing international migration that leads to improvement in employment rates with extremely important further positive impact on economic welfare (increases in GDP per capita).
The asylum seekers (found in a smaller number than the economic immigrants) tend to also generate positive effects upon the labour market, at least in the short-term, slightly increasing the employment rate. Thus, H2 and H3 are reconfirmed and fulfilled.
The results of the empirical analysis complete the core elements of the host country’s macroeconomic environment in shaping the attitudes towards migration.
Estimations show that the additional immigrant flow has no negative effects on the labour market performance of natives and the foreign population already existing in the considered host economies—elements that were also considered by Card and DiNardo [20]—but, on the contrary, the increased flow positively determines an increase in the employment rates (aspects shown by the estimated coefficients that are statistically significant and can, therefore, be interpreted).
The negative coefficients associated with the KOF variable accounting for globalization (2000–2019) tends to highlight that, nowadays, international migration represents a major frontier of the globalization process. ALMPs and migration flows are jointly significant in positively influencing the employment rate, with further positive effects on living standards and economic welfare of the population residing in the ten countries considered. Thus, H4 is reconfirmed and fulfilled.

5. Concluding Remarks

Employment moves from one country to another lead to changes in the economic conditions for both the origin and destination countries. The empirical analysis performed in this context pursued the development of complex methods to evaluate the effects induced by the international migration (both economic and humanitarian) over the ten host-economies EU Member States that are primarily pursued by immigrants and asylum seekers, thus trying to answer the main questions and fears connected to migration. There was a keen interest in assessing the economic consequences of migratory flows form a two-fold perspective, focusing on its determinants and direct/indirect/total effects, to better capture and determine the most suitable ways to enhance migrants’ roles in supporting long-term economic development.
Based on these investigations, we could observe that short-term and long-term impacts of migration, which may vary in magnitude, are extensively determined by labour and capital flexibility and the ability of labour markets to adjust to various changes in the short and longer term. In line with what the theory predicts, international migration, one of the main globalization vectors, generates complex economic and social consequences upon migrant sending and receiving countries. For the receiving countries, migration helps fill job vacancies and skills, support economic growth, and bring energy, innovation, and cultural diversity. These credentials are essential for sustainable economic development.
Our estimations highlighted that the labour immigrant flows have important positive economic consequences embedded by a series of major changes in the labour market performance both for the native workers and for the foreign population (employment and wage level increases, intensified earnings dispersion) and improved levels of GDP per capita and per person employed. As Ortega and Peri [52] underlined, the “international migration flows are highly responsive to income per capita at destination… tightening of laws regulating immigrant entry reduce rapidly and significantly their flow” [52] (p. 47).
Overall, the analyses confirm the main economic effects of international migration in a globalised era, some of which have previously been highlighted in relevant published literature, but are comprehensively integrated in this paper. These effects refer to both positive and negative elements for migrant receiving countries, and they differ significantly from one country to another according to the specific measures and policies adopted to coordinate this extremely complex process.
The performed research is in line with the most recent initiatives launched globally, especially by the UN DESA and International Labor Organization (ILO) specialists, on the role of labour migration in achieving future sustainable development goals. They attest that “migration has the potential to lift millions out of poverty, provide access to decent work and foster sustainable development” and highlight that “greater efforts are required to strengthen the evidence-base on migration and development so that policies can be made in a more informed manner” [53] (p. 1).
Our research is not without limitations though, mainly induced by low data availability for longer time series that are accurate in revealing the amplitude of the migration phenomenon. These kinds of data limitations were also mentioned by other studies, for example, Soon [48]. Therefore, we intend to expand the immigration analysis to a bilateral matrix approach following the emigration–immigration–return migration axis, with a particular interest on brain drain–brain gain–brain waste coordinates, family members’ component, and the effects of destination countries’ entry laws for migrants, as Ortega and Peri [52] have previously proved.

Acknowledgments

Part of this work was supported by a mobility project of the Romanian Ministry of Research and Innovation, CNCS—UEFISCDI, project number PN-III-P1-1.1-MCD-2017-0114, within PNCDI III. We would like to thank the reviewers for their accurate and thoughtful comments that were extremely useful in framing the overall research.

Author Contributions

Gratiela Georgiana Noja established the general framework of the paper. All authors contributed overall in writing the paper and approved the final manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table A1. Variables used in the empirical research endeavour.
Table A1. Variables used in the empirical research endeavour.
AcronymDescriptionSource
IMIGFlows of immigrants and foreign population (number)
The inflows include status changes, namely persons in the country on a temporary status who obtained the right to stay on a longer-term basis
OECD, Eurostat
ASYLInflows of asylum applicants (number, new asylum requests into OECD)OECD, UNHCR
GDP_capGross domestic product per capita (Euro)UNCTADstat
GDP_empGross domestic product per person employed (USD)UNCTADstat
ER, ER_FEmployment rate (%)—total and for the foreign population (F)Eurostat
OECD
UR, UR_F, UR_LTUnemployment rate (%)—total, for the foreign population (F), long term (LT)Eurostat
OECD
ER_partPart-time and temporary contracts employment rate (%)Eurostat
EARNAnnual net earnings of a two-earner married couple with two children (Euro)Eurostat
EARN_dispEarnings dispersion (Decile 9/Decile 5)Eurostat
ALMPsExpenditures on active labour market policies (% of GDP)Eurostat
PLMPsExpenditures on passive labour market policies (% of GDP)Eurostat
EDU_TEducational level (both general and vocational) reflected through the educational attainment for tertiary education (levels 5–8)Eurostat
EDU_partThe participation rate in education and training (%)Eurostat
EARNAnnual net earnings of a two-earner married couple with two children (Euro) and earnings dispersion (Decile 9/Decile 5)Eurostat, World Bank
POV_RAt-risk-of-poverty-rate (%)Eurostat, World Bank
LELife expectancy at birthEurostat, World Bank
BERDBusiness enterprise R&D expenditures for the business enterprise sectorEurostat
GiniGini coefficients (index)UNU WIDER
KOF_TKOF index of Globalization (overall dimensions—economic, social, political)ETH Zurich
KOF_EKOF index of Economic GlobalizationETH Zurich
XExports of goods and services (mil. USD)UNCTADstat
FDI_IInflows/inward of foreign direct investment (mil. USD)UNCTADstat
Source: own process.
Table A2. Results of the spatial analysis autoregressive models developed for assessing the economic immigration shaping factors.
Table A2. Results of the spatial analysis autoregressive models developed for assessing the economic immigration shaping factors.
(1)(2)(3)(4)(5)
IMIGIMIGIMIGIMIGIMIG
GDP_cap−0.169 ***−0.134 ***
(0.0405)(0.0326)
ER0.221 ***0.159 ***
(0.0577)(0.0381)
POV_R−0.185 *−0.198 ***−0.185 **−0.245 ***
(0.0720)(0.0583)(0.0583)(0.0612)
ALMPs0.00425
(0.0366)
Gini0.355 ***0.315 ***0.328 ***0.378 ***
(0.0826)(0.0615)(0.0641)(0.0701)
EDU_part−0.127 ***−0.0805 *−0.1000 **−0.0672 **
(0.0386)(0.0318)(0.0308)(0.0245)
KOF_E−0.228 ***−0.265 ***−0.276 ***−0.262 ***−0.146 ***
(0.0488)(0.0517)(0.0502)(0.0460)(0.0430)
GDP_emp−0.128 ***−0.158 ***−0.118 ***
(0.0341)(0.0348)(0.0319)
ER_F0.0590
(0.0428)
LE−0.145 ***−0.117 **−0.128 ***
(0.0317)(0.0381)(0.0317)
UR_LT−0.0937 *
(0.0456)
ER_part0.186 ***
(0.0419)
EARN−0.128 *
(0.0567)
BERD−0.149 ***
(0.0429)
FDI_I0.0994 **
(0.0338)
Export0.259 ***
(0.0602)
_cons0.116 ***0.116 ***0.116 ***0.116 ***0.116 ***
(0.0310)(0.0304)(0.0302)(0.0293)(0.0294)
rho0.983 ***0.981 ***0.981 ***0.980 ***0.982 ***
_cons(0.0169)(0.0193)(0.0188)(0.0199)(0.0183)
sigma0.391 ***0.385 ***0.382 ***0.370 ***0.369 ***
_cons(0.0330)(0.0342)(0.0332)(0.0392)(0.0411)
N160160160160160
Note: Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: authors’ research.
Table A3. Results of the spatial analysis autoregressive models developed for assessing the shaping factors of asylum seekers inflows.
Table A3. Results of the spatial analysis autoregressive models developed for assessing the shaping factors of asylum seekers inflows.
(1)(2)(3)(4)(5)
ASYLASYLASYLASYLASYL
GDP_cap−0.155 *−0.161 *
(0.0637)(0.0627)
ER0.2520.225 *
(0.146)(0.0931)
POV_R−0.0654−0.242 *−0.102−0.0890
(0.114)(0.121)(0.117)(0.130)
ALMPs−0.192 **
(0.0599)
Gini−0.20000.0401−0.0429−0.0401
(0.129)(0.137)(0.133)(0.141)
EDU_part0.09480.181 ***0.1030.0753
(0.0919)(0.0432)(0.0617)(0.0741)
KOF_E−0.543 ***−0.580 ***−0.572 ***−0.521 ***−0.389 ***
(0.104)(0.106)(0.107)(0.0984)(0.105)
GDP_emp−0.0562−0.0279−0.0624
(0.0570)(0.0561)(0.0752)
ER_F−0.134*
(0.0600)
LE−0.127−0.0679−0.115
(0.0844)(0.0779)(0.0777)
UR_LT−0.148 *
(0.0744)
ER_part0.284 **
(0.100)
EARN−0.0130
(0.129)
BERD0.200 *
(0.0841)
FDI_I0.0119
(0.0839)
Export0.149
(0.173)
_cons0.03010.03200.03240.03160.0317
(0.0576)(0.0588)(0.0589)(0.0568)(0.0576)
rho0.655 **0.695 **0.705 **0.686 **0.689 **
_cons(0.222)(0.223)(0.214)(0.233)(0.239)
sigma0.777 ***0.796 ***0.796 ***0.764 ***0.771 ***
_cons(0.117)(0.130)(0.124)(0.118)(0.106)
N160160160160160
Note: Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: authors’ research.
Table A4. Descriptive statistics of the indicators used within the empirical analysis.
Table A4. Descriptive statistics of the indicators used within the empirical analysis.
Summary Statistics for the 2000–2015 Sample
NMeansdMinMax
IMIG160279,129.1245,104.416,8951,077,073
ASYL16029,641.1931,141.131405236,560
GDP_cap16034,920.4710,571.9915,60064,161.97
GDP_emp16064,445.9121,031.7538,09599,363.52
ER16066.849386.16707653.777.9
ER_F16060.868755.42855545.270.4
ER_part16019.49755.3631637.827.8
UR1607.268751.8306683.413.3
UR_F16013.556875.5759556.435.4
UR_LT1602.9143752.1884160.513
EARN16051,394.210,763.1228,41278,469
EARN_disp1601.7983440.15983161.532.18
ALMPs1600.65953120.35455530.0351.691
PLMPs1601.4000060.65213320.12800013.053
EDU_T16024.71256.5345988.137.6
EDU_part16018.1843712.47486−6.795.3
POV_R16020.70254.5118439.130.29999
LE16080.261881.62967276.284.59999
BERD1601.2641870.60956550.382.985
Gini16028.493753.31474520.635.7
KOF_T16085.505884.09135277.1800192.63
KOF_E16078.78058.74508660.1999994.8
X160439,763.1392,694.652,077.391,849,760
FDI_I16027,102.9633,621.13−12,272.31198,276.5
Summary Statistics for the 2000–2019 Sample
NMeanSdMinMax
IMIG200305,259.6308,797.916,8951,845,793
ASYL20048,116.5567,278.91405490,520
GDP_cap20033,923.3310,840.048735.14164,161.97
GDP_emp20064,963.9721,269.7438,095102,240.9
ER20067.12356.34443553.777.9
ER_F20061.0715.59130445.274.4
ER_part20020.0635.3842817.831.4
UR2007.4012.2381562.10000115.7
UR_F20013.66255.5086453.60000135.4
UR_LT2002.97752.196558−0.800000113
EARN20054,223.5313,303.7928,412113,749
EARN_disp2001.7830750.16833481.322.18
ALMPs2000.6630750.38014460.0351.973
PLMPs2001.3422350.6662019−0.00399993.053
EDU_T20026.628.9374718.175.89999
EDU_part20018.462511.81902−6.795.3
POV_R20020.494.6461429.130.30001
LE20080.83752.01148576.287.39998
BERD2001.285850.59702490.382.985
Gini20028.853.58804920.639.7
KOF_T20084.68484.52408474.9400292.63
KOF_E20076.475810.0874349.0800294.8
X200419,140.9380,639.213,830.831,849,760
FDI_I20028,915.0840,379.77−45,351.04198,276.5
Source: own process in Stata.
Table A5. (a) Unit Root Tests of the Residuals from the GDP-Trade-FDI Impact Models, 2000–2015. (b) Unit Root Tests of the Residuals from the GDP-Trade-FDI Impact Models, 2000−2019.
Table A5. (a) Unit Root Tests of the Residuals from the GDP-Trade-FDI Impact Models, 2000–2015. (b) Unit Root Tests of the Residuals from the GDP-Trade-FDI Impact Models, 2000−2019.
(a)
Resid
LLC (Levin-Lin-Chu)p-value0.0021
t-statistic−2.8573
ADF (Augmented Dickey-Fuller) regressions: 1 lag
LR (Likelihood Ratio) variance: Bartlett kernel, 8.00 lags average
Im-Pesaran-Shinp-value0.3648
t-statistic−1.695
Test critical values: 1%−2.210
Test critical values: 5%−1.990
Test critical values: 10%−1.890
ADF regressions: No lags included
AR (Autoregressive) parameter: Panel-specific
Harris-Tzavalisp-value0.4693
Statistic0.8191
Z−0.0771
Fisher-type
Based on Augmented Dickey-Fuller tests
p-value0.05930
Inverse chi-squared (56)30.6959
Modified inv. chi-squared1.6912
(b)
Resid
LLC (Levin-Lin-Chu)p-value0.3872
t-statistic−0.2866
ADF regressions: 1 lag
LR variance: Bartlett kernel, 8.00 lags average
Im-Pesaran-Shinp-value0.9957
t-statistic2.6264
Test critical values: 1%−2.210
Test critical values: 5%−1.990
Test critical values: 10%−1.890
ADF regressions: No lags included
AR (Autoregressive) parameter: Panel-specific
Harris-Tzavalisp-value0.9990
Statistic1.0020
Z3.0961
Fisher-type
Based on Augmented Dickey-Fuller tests
p-value0.0451
Inverse chi-squared (56)31.8336
Modified inv. chi-squared1.8711
Ho: (All) Panels contain unit roots. Ha: Panels are stationary/At least one panel is stationary. Source: authors’ research.
Table A6. (a) Results for Cronbach’s alpha, 2000–2015; (b) Results for Cronbach’s alpha, 2000–2019.
Table A6. (a) Results for Cronbach’s alpha, 2000–2015; (b) Results for Cronbach’s alpha, 2000–2019.
(a)
Test Scale = Mean (Standardized Items)
Average
ItemObsSignItem-Test CorrelationItem-Rest CorrelationInter-Item CorrelationAlpha
GDP_cap160+0.76270.67770.15320.6656
IMIG1600.55030.41750.17640.7020
ASYL1600.20580.04000.21400.7496
EARN160+0.65400.54160.16510.6850
EDU_T160+0.44730.29940.18760.7176
UR_F1600.33500.17590.19990.7332
KOF_T160+0.61710.49680.16910.6912
POV_R1600.63220.51510.16740.6887
LE1600.37320.21730.19570.7280
ER160+0.82280.75590.14670.6540
ALMPs160+0.46960.32450.18520.7143
PLMPs1600.1306−0.03630.22220.7586
Total scale0.18190.7273
(b)
Test Scale = Mean (Standardized Items)
Average
ItemObsSignItem-Test CorrelationItem-Rest CorrelationInter-Item CorrelationAlpha
GDP_cap200+0.68200.57290.14800.6565
IMIG2000.40840.25190.17710.7031
ASYL2000.29430.12810.18930.7198
EARN200+0.56130.42670.16090.6783
EDU_T200+0.42290.26800.17560.7009
UR_F2000.36050.19920.18220.7103
KOF_T200+0.58770.45800.15810.6737
POV_R2000.63580.51600.15290.6651
LE2000.38780.22910.17930.7062
ER200+0.79280.71420.13620.6344
ALMPs200+0.50910.36570.16640.6871
PLMPs2000.21130.04140.19810.7310
Total scale0.16870.7089
Source: authors’ research.
Table A7. Wald tests for equations.
Table A7. Wald tests for equations.
VariableChi2dfp-ValueChi2dfp-Value
GDP_cap 2000–2015GDP_cap 2000–2019
Observed
IMIG217.2140.0000208.5140.0000
ER107.4040.0000136.7340.0000
ASYL34.7030.000083.2230.0000
GDP_cap70.4810.000080.4910.0000
H0: all coefficients excluding the intercepts are 0. We can thus reject that null hypothesis for each equation. Source: authors’ research.
Table A8. (a) Goodness-of-fit tests for SEM models, 2000–2015; (b) Goodness-of-fit tests for SEM models, 2000–2019.
Table A8. (a) Goodness-of-fit tests for SEM models, 2000–2015; (b) Goodness-of-fit tests for SEM models, 2000–2019.
(a)
Fit StatisticValueDescription
Likelihood ratio
chi2_ms(26)438.896model vs. saturated
p > chi20.000
chi2_bs(38)748.095baseline vs. saturated
p > chi20.000
Information criteria
AIC4697.528Akaike’s information criterion
BIC4759.031Bayesian information criterion
Baseline comparison
CFI0.419Comparative fit index
TLI0.150Tucker-Lewis index
Size of residuals
SRMR0.146Standardized root mean squared residual
CD0.754Coefficient of determination
(b)
Fit StatisticValueDescription
Likelihood ratio
chi2_ms(26)344.142model vs. saturated
p > chi20.000
chi2_bs(38)748.398baseline vs. saturated
p > chi20.000
Information criteria
AIC6686.990Akaike’s information criterion
BIC6752.957Bayesian information criterion
Baseline comparison
CFI0.539Comparative fit index
TLI0.327Tucker-Lewis index
Size of residuals
SRMR0.118Standardized root mean squared residual
CD0.764Coefficient of determination
Source: authors’ research.

References

  1. Gavosto, A.; Venturini, A.; Villosio, C. Do immigrants compete with natives? Labour 1999, 13, 603–621. [Google Scholar] [CrossRef]
  2. Dustmann, C.; Casanova, M.; Fertig, M.; Preston, I.; Schmidt, C.M. The Impact of EU Enlargement on Migration Flows; Home Office Online Report 25/03; Research Development and Statistics Directorate, Home Office: London, UK, 2003; pp. 1–76. [Google Scholar]
  3. Staffolani, S.; Valentini, E. Does immigration raise blue and white collar wages of natives? The case of Italy. Labour 2010, 24, 295–310. [Google Scholar] [CrossRef]
  4. Falzoni, A.M.; Venturini, A.; Villosio, C. Skilled and Unskilled Wage Dynamics in Italy in the 90s: Changes in the Individual Characteristics Institutions Trade and Technology; Working Paper Series 61; Laboratorio Ricardo Revelli (LABOR), Centre for Employment Studies: Moncalieri, Italy, 2007. [Google Scholar]
  5. Angrist, J.D.; Kugler, A.D. Protective or counter-productive? Labor market institutions and the effect of immigration on EU natives. Econ. J. 2003, 113, 302–331. [Google Scholar] [CrossRef]
  6. Longhi, S.; Nijkamp, P.; Poot, J. The Impact of Immigration on the Employment of Natives in Regional Labour Markets: A Meta-Analysis; ISER Working Papers Series 10; Institute for Social and Economic Research (ISER): Essex, UK, 2006. [Google Scholar]
  7. Ozgen, C.; Nijkamp, P.; Poot, J. Immigration and Innovation in European Regions; Norface Migration Discussion Paper 8; NORFACE Research Programme on Migration: London, UK, 2011. [Google Scholar]
  8. Glitz, A. The Labor Market Impact of Immigration: A Quasi-Experiment Exploiting Immigrant Location Rules in Germany. J. Labor Econ. 2011, 30, 175–213. [Google Scholar] [CrossRef]
  9. Rodriguez-Pose, A.; Vilalta-Bufi, M. Education, migration, and job satisfaction: The regional returns of human capital in the EU. J. Econ. Geogr. 2005, 5, 545–566. [Google Scholar] [CrossRef]
  10. Borjas, G.J. The Demand Curve is Downwards Loping: Reexamining the Impact of Immigration on the Labor Market. Q. J. Econ. 2003, 118, 1335–1374. [Google Scholar] [CrossRef]
  11. European Commission. Eurostat. Database. 2017. Available online: http://ec.europa.eu/eurostat/data/database (accessed on 15 March 2017).
  12. UNHCR (The UN Refugee Agency). Figures at a Glance. 2017. Available online: http://www.unhcr.org/figures-at-a-glance.html (accessed on 20 November 2017).
  13. Lee, E.S. A Theory of Migration. Demography 1966, 3, 47–57. [Google Scholar] [CrossRef]
  14. Bauer, T.K.; Haisken-DeNew, J.P.; Schmidt, C.M. International Labor Migration, Economic Growth and Labor Markets—The Current State of Affairs; RWI Discussion Paper 20; Leibniz Institute for Economic Research: Essen, Germany, 2004. [Google Scholar] [CrossRef]
  15. Zimmermann, K.F. European Migration: What Do We Know? Oxford University Press: Oxford, UK, 2005; pp. 237–253. ISBN 978-0-19-925735-5. [Google Scholar]
  16. Organization for Economic Cooperation and Development (OECD). Is This Humanitarian Migration Crisis Different? Migration Policy Debates 7; Organization for Economic Cooperation and Development: Paris, France, 2015. [Google Scholar]
  17. Aubry, A.; Burzynski, M.; Docquier, F. The welfare impact of global migration in OECD countries. J. Int. Econ. 2016, 101, 1–21. [Google Scholar] [CrossRef]
  18. Kahanec, M.; Zimmermann, K.F. Migration, the Quality of the Labour Force and Economic Inequality; IZA Discussion Paper Series 3560; Institute for the Study of Labor: Bonn, Germany, 2008. [Google Scholar]
  19. Brucker, H.; Fachin, S.; Venturini, A. Do foreigners replace native immigrants? A panel cointegration analysis of internal migration in Italy. Econ. Model. 2011, 28, 1078–1089. [Google Scholar] [CrossRef]
  20. Card, D.; DiNardo, J.E. Do immigrants inflows lead to native outflows? Am. Econ. Rev. 2000, 90, 360–373. [Google Scholar] [CrossRef]
  21. Card, D. Immigrant inflows, native outflows, and the local labor market impacts of higher immigration. J. Labor Econ. 2001, 19, 22–64. [Google Scholar] [CrossRef]
  22. Dustmann, C.; Fabbri, F.; Preston, I. The impact of immigration on the British labour market. Econ. J. 2005, 115, F324–F341. [Google Scholar] [CrossRef]
  23. Betz, W.; Simpson, N.B. The effects of international migration on the well-being of native populations in Europe. J. Migr. 2013, 2, 12. [Google Scholar] [CrossRef] [Green Version]
  24. Tsapenko, I.P. Social effects of immigration. Her. Russ. Acad. Sci. 2015, 85, 443–452. [Google Scholar] [CrossRef]
  25. Boone, J.; van Ours, J.C. Effective Active Labor Market Policies; IZA Discussion Paper Series 1335; Institute for the Study of Labor: Bonn, Germany, 2004. [Google Scholar]
  26. Card, D.; Kluve, J.; Weber, A. Active Labour Market Policy Evaluations: A Meta-Analysis. Econ. J. 2010, 120, F452–F477. [Google Scholar] [CrossRef]
  27. UN DESA (United Nations, Department of Economic and Social Affairs). Trends in International Migrant Stock: The 2015 Revision; UN DESA (United Nations, Department of Economic and Social Affairs): New York, NY, USA, 2015. [Google Scholar]
  28. Organization for Economic Cooperation and Development (OECD). International Migration Outlook. Trends in Migration Flows, 2016; Organization for Economic Cooperation and Development (OECD): Paris, France, 2017. [Google Scholar] [CrossRef]
  29. Organization for Economic Cooperation and Development (OECD). International Migration Outlook 2017 New Asylum Applications by Country of Destination, 2012-16; Organization for Economic Cooperation and Development (OECD): Paris, France, 2017. [Google Scholar]
  30. IOM (International Organization for Migration). World Migration Report 2018; The UN Migration Agency: Geneva, Switzerland, 2018; pp. 18–32. ISBN 978-92-9068-742-9. [Google Scholar]
  31. UNHCR (The UN Refugee Agency). Mediterranean Situation. 2017. Available online: http://data2.unhcr.org/en/situations/mediterranean (accessed on 17 December 2017).
  32. Tănasie, A.; Drăcea, R.; Lădaru, G.R. A Chaos Theory Perspective on International Migration. Sustainability 2017, 9, 2355. [Google Scholar] [CrossRef]
  33. Organization for Economic Cooperation and Development (OECD). Handbook on Constructing Composite Indicators: Methodology and User Guide; OECD Publishing: Paris, France, 2005; pp. 83–88. ISBN 978-92-64-04345-9. [Google Scholar]
  34. Chung, H. Measuring Flexicurity: Precautionary Notes, a New Framework and an Empirical Example. Soc. Indic. Res. 2012, 106, 153–171. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  35. Lozano, F.A.; Steinberger, M.D. Empirical Methods in the Economics of International Immigration; IZA Discussion Paper Series 5328; Institute for the Study of Labor: Bonn, Germany, 2010. [Google Scholar]
  36. McKenzie, D.; Sasin, M.J. Migration, Remittances, Poverty, and Human Capital: Conceptual and Empirical Challenges; Policy Research Working Paper 4272; World Bank: Washington, DC, USA, 2007. [Google Scholar]
  37. Vasco, C. The Impact of International Migration and Remittances on Agricultural Production Patterns, Labor Relationships and Entrepreneurship: The Case of Rural Ecuador; Kassel University Press GmbH: Munchen, Germany, 2011; pp. 100–125. ISBN 978-3-86219-086-7. [Google Scholar]
  38. Miluka, J.; Carletto, C.; Davis, B.; Zezza, A. The Vanishing Farms. The Impact of International Migration on Albanian Family Farming; Policy Research Working Paper 4367; World Bank: Washington, DC, USA, 2007. [Google Scholar]
  39. McKenzie, D.; Rapoport, H. Network Effects and the Dynamics of Migration and Inequality: Theory and Evidence from Mexico; World Bank: Washington, DC, USA, 2006. [Google Scholar]
  40. Baum, C.F. An Introduction to Modern Econometrics Using Stata; Stata Press: College Station, TX, USA, 2001; pp. 90–127. ISBN 978-1-59718-013-9. [Google Scholar]
  41. Son, L.; Noja, G.G. A macroeconometric panel data analysis of the shaping factors of labour emigration within the European Union. Theor. Appl. Econ. 2012, 11, 15–30. [Google Scholar]
  42. Noja, G.G.; Son, L. Challenges of International Migration in a Globalized World: Implications for Europe. Int. J. Innov. Econ. Dev. 2016, 2, 7–17. [Google Scholar] [CrossRef]
  43. Viton, P.A. Notes on Spatial Econometric Models. City Reg. Plan. 2010, 870, 9–10. [Google Scholar]
  44. Dascălu, E.D.; Marcu, N.; Pete, S.; Ulici, M.L.; Dumitraşcu, V. Dependent business climate. A network-based analysis. Rom. J. Econ. Forecast. 2016, 19, 138–152. [Google Scholar]
  45. Marcu, N.; Meghişan, G.M.; Ciobanu, M.C. Research on Romanian Labour Market Dynamics. Rev. Chim. 2015, 66, 1540–1544. [Google Scholar]
  46. Soon, J.J. Effects of EU Expansion on Migrants’ Employment and Income: A Natural Experiment. Prague Econ. Pap. 2017, 1–16. [Google Scholar] [CrossRef]
  47. World Bank. World Development Indicators. Available online: http://data.worldbank.org/data-catalog/world-development-indicators (accessed on 4 September 2017).
  48. UNU WIDER (United Nations University World Institute for Development Economics Research). World Income Inequality Database; United Nations University: Helsinki, Finland, 2017. [Google Scholar]
  49. ETH Swiss Federal Institute of Technology Zurich. KOF Index of Globalization; KOF Swiss Economic Institute: Zurich, Switzerland, 2017. [Google Scholar]
  50. Chiswick, B.; Miller, P.W. Handbook of the Economics of International Migration: The Impact; Elsevier: Oxford, UK, 2014; pp. 793–844. ISBN 978-0-444-53768-3. [Google Scholar]
  51. Razin, A.; Sadka, E.; Suwankiri, B. Migration and the Welfare State: Political-Economy Policy Formation; MIT Press: Cambridge, MA, USA, 2011; pp. 13–116. ISBN 978-0-262-29926-8. [Google Scholar]
  52. Ortega, F.; Peri, G. The Effect of Income and Immigration Policies on International Migration. Migr. Stud. 2013, 1, 47–74. [Google Scholar] [CrossRef]
  53. UN DESA (United Nations, Department of Economic and Social Affairs). Examining the role of labour migration for sustainable development/Achieving the future sustainable development goals: The role of labour migration. In Proceedings of the Side event to the Second United Nations Conference on Landlocked Developing Countries (LLDCs), Vienna, Austria, 3–5 November 2014. [Google Scholar]
Figure 1. International migrant stock allocation by regions (a) and in Europe (b), 2015. Source: Own process based on United Nations, Department of Economic and Social Affairs (UN DESA) [27] data.
Figure 1. International migrant stock allocation by regions (a) and in Europe (b), 2015. Source: Own process based on United Nations, Department of Economic and Social Affairs (UN DESA) [27] data.
Sustainability 10 00894 g001
Figure 2. Number of immigrants (a); asylum applicants (b); and GDP per capita (c) levels in 2015 and 2019 (estimated) in European Union (EU)-10 receiving countries. Source: Own process of Eurostat [11], Organization for Economic Cooperation and Development (OECD) [28,29] and United Nations Databases/United Nations High Commissioner for Refugees (UNHCR) [12] data in Stata.
Figure 2. Number of immigrants (a); asylum applicants (b); and GDP per capita (c) levels in 2015 and 2019 (estimated) in European Union (EU)-10 receiving countries. Source: Own process of Eurostat [11], Organization for Economic Cooperation and Development (OECD) [28,29] and United Nations Databases/United Nations High Commissioner for Refugees (UNHCR) [12] data in Stata.
Sustainability 10 00894 g002
Figure 3. Moran’s I test for local (a) and global (b) spatial autocorrelation. Source: authors’ research.
Figure 3. Moran’s I test for local (a) and global (b) spatial autocorrelation. Source: authors’ research.
Sustainability 10 00894 g003
Figure 4. Structural Equations Model (SEM) developed for the analysis of international migration (both economic and humanitarian) from the host country’s perspective: general model configuration (a); 2000–2015 (b); and 2000–2019 (c). Source: authors’ research.
Figure 4. Structural Equations Model (SEM) developed for the analysis of international migration (both economic and humanitarian) from the host country’s perspective: general model configuration (a); 2000–2015 (b); and 2000–2019 (c). Source: authors’ research.
Sustainability 10 00894 g004
Table 1. (a) Results of the models developed for assessing the economic consequences of immigration upon host economies, 2000–2015 sample—General model; (b) Results of the models developed for assessing the economic consequences of immigration upon host economies, 2000–2019 sample—General model.
Table 1. (a) Results of the models developed for assessing the economic consequences of immigration upon host economies, 2000–2015 sample—General model; (b) Results of the models developed for assessing the economic consequences of immigration upon host economies, 2000–2019 sample—General model.
(a)
Dependent Variable: GDP_Cap_st (D.GDP_Cap_st in Model 2)(1)(2)(3)(4)(5)
Fixed Effects (FE)IV FD (2SLS)
D. Indep_Var
Dynamic GMM (Arellano-Bond)Robust RegressionRobust Reg. Bootstrap
IMIG_st0.278 ***0.02030.2770.03420.0342
(0.0775)(0.0783)(0.550)(0.0581)(0.0785)
ASYL_st−0.214 ***−0.131 **−0.603 *−0.213 ***−0.213 *
(0.0394)(0.0490)(0.260)(0.0497)(0.0983)
ER_st−0.1520.28915.47−0.340 ***−0.340 **
(0.113)(0.179)(10.00)(0.0767)(0.125)
EDU_T_st0.0626−0.007101.6660.1000.100
(0.0832)(0.0994)(1.287)(0.0558)(0.0525)
BERD_st0.365 **0.260−2.4420.07230.0723
(0.114)(0.174)(7.999)(0.0673)(0.0556)
ALMP_st0.262 ***0.263 **3.993*0.193 **0.193 *
(0.0709)(0.0976)(2.021)(0.0587)(0.0761)
PLMP_st−0.153 *−0.1055.250−0.0528−0.0528
(0.0603)(0.0897)(3.168)(0.0583)(0.0895)
Gini_st−0.0668−0.02801.053−0.143 *−0.143
(0.0782)(0.0615)(0.867)(0.0703)(0.0998)
EARN_st0.774 ***0.391 ***4.4490.900 ***0.900 ***
(0.0728)(0.117)(2.288)(0.0718)(0.0920)
L.GDP_cap_st−14.80
(8.311)
_cons−0.0000008330.0274−0.0460−0.0460
(0.0266)(0.0263)(0.0389)(0.0570)
Test parameters = 0F(9, 141) = 47.27
p = 0.0000
F(9, 150) = 52.94
p = 0.0000
Breusch-Pagan LM test of independenceChi2(45) = 290.444
p = 0.0000
Hausman test (RE vs. FE)Chi2(9) = 195.15
p = 0.0000
Sargan statistic (override test of all instruments)Chi2(1) = 0.320
p = 0.5719
Arellano-Bond test (zero autocorrelation in FD errors)z1 = 0.84885
p = 0.3960
N160150140160160
R20.7510.2280.7610.761
(b)
Dependent Variable: GDP_Cap_st (D.GDP_Cap_st
in Model 2)
(1)(2)(3)(4)(5)
Fixed Effects (FE)IV FD (2SLS)
D. Indep_Var
Dynamic GMM (Arellano-Bond)Robust RegressionRobust Reg. Bootstrap
IMIG_st0.297 **0.1050.00781−0.0215−0.0215
(0.0990)(0.0894)(0.266)(0.0547)(0.0838)
ASYL_st−0.234 ***−0.196 ***−0.309 *−0.0680 *−0.0680
(0.0344)(0.0450)(0.153)(0.0265)(0.0428)
ER_st0.02420.3222.729−0.149−0.149
(0.206)(0.219)(5.839)(0.0808)(0.199)
EDU_T_st0.07820.00540−6.906−0.0319−0.0319
(0.0586)(0.0826)(6.682)(0.0428)(0.0586)
BERD_e_st0.481 **0.555 **6.3880.1300.130
(0.155)(0.205)(4.806)(0.0748)(0.106)
ALMPs_st0.922 ***0.532 ***−1.017 **0.09660.0966
(0.113)(0.113)(0.328)(0.0535)(0.150)
PLMPs_st0.1090.0321−0.3920.06110.0611
(0.0864)(0.102)(1.805)(0.0599)(0.0879)
Gini_st0.1840.08930.298 **−0.0877−0.0877
(0.113)(0.0747)(0.109)(0.0729)(0.119)
EARN_st0.192 *0.1452.5770.808 ***0.808 ***
(0.0801)(0.120)(2.491)(0.0608)(0.213)
L.GDP_cap_st−1.657
(0.926)
_cons−0.0793−0.0112−0.0814−0.0814
(0.0497)(0.0299)(0.0451)(0.0648)
Test parameters = 0F(9, 181) = 13.32
p = 0.0000
F(9, 190) = 52.40
p = 0.0000
Breusch-Pagan LM test of independenceChi2(45) = 344.376
p = 0.0000
Hausman test (RE vs. FE)Chi2(9) = 426.40
p = 0.0000
Sargan statistic (override test of all instruments)Chi2(1) = 0.004
p = 0.9480
Arellano-Bond test (zero autocorrelation in FD errors)z1 = 0.7534
p = 0.4512
N200190180200200
R20.3990.2200.7130.713
Note: Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: authors’ research.
Table 2. (a) Results of the Fixed Effects models (within country variation) developed for assessing the economic consequences of immigration upon host economies, both samples (2000–2015 and 2000–2019)—Multiple models; (b) Results of the Panel 2SLS (IV) estimation (first difference) developed for assessing the economic consequences of immigration upon host economies, both samples (2000–2015 and 2000–2019)—Multiple models; (c) Results of the Arellano Bond Dynamic GMM estimations for both samples (2000–2015 and 2000–2019).
Table 2. (a) Results of the Fixed Effects models (within country variation) developed for assessing the economic consequences of immigration upon host economies, both samples (2000–2015 and 2000–2019)—Multiple models; (b) Results of the Panel 2SLS (IV) estimation (first difference) developed for assessing the economic consequences of immigration upon host economies, both samples (2000–2015 and 2000–2019)—Multiple models; (c) Results of the Arellano Bond Dynamic GMM estimations for both samples (2000–2015 and 2000–2019).
(a)
Fixed Effects (FE)
2000–2015(1)(2)(3)(4)(5)(6)(7)(8)
GDP_capGDP_empERER_FEDU_TEARNEARN_dispGini
IMIG0.2270.02330.241 ***0.749***0.0258−0.03840.217 **−0.209 *
(0.134)(0.0233)(0.0587)(0.124)(0.106)(0.128)(0.0739)(0.0867)
ASYL−0.0269−0.02230.0655 *−0.01050.09680.198 **−0.06960.178 ***
(0.0713)(0.0125)(0.0313)(0.0662)(0.0568)(0.0684)(0.0394)(0.0463)
_cons−0.0000004920.000000165−0.0000008−9.65 × 10−8−1.09 × 10−81.93 × 10−8−0.000001636.62 × 10−8
(0.0516)(0.00902)(0.0227)(0.0479)(0.0411)(0.0495)(0.0286)(0.0335)
N160160160160160160160160
R20.0190.0230.1540.2050.0220.0550.0620.103
2000–2019(1)(2)(3)(4)(5)(6)(7)(8)
GDP_capGDP_empERER_FEDU_TEARNEARN_dispGini
IMIG0.238 *0.0454 **0.223 ***0.591 ***−0.1090.257 *0.183 *0.115
(0.107)(0.0162)(0.0354)(0.0781)(0.143)(0.101)(0.0744)(0.0662)
ASYL−0.07030.001080.0248 *−0.0623 *0.123 *0.139 ***−0.0835 **0.102 ***
(0.0373)(0.00565)(0.0124)(0.0273)(0.0501)(0.0354)(0.0260)(0.0231)
_cons−0.07800.0191 *0.005910.01120.231 **0.153 **−0.06550.0347
(0.0601)(0.00911)(0.0199)(0.0440)(0.0807)(0.0570)(0.0419)(0.0373)
N200200200200200200200200
R20.0290.0620.3090.2500.0340.2020.0550.193
(b)
Panel 2SLS (IV) Estimation—First Difference (FD)
2000–2015(1)(2)(3)(4)(5)(6)(7)(8)
D.GDP_capD.GDP_empD.ERD.ER_FD.EDU_TD.EARND.EARN_dispD.Gini
D.IMIG0.1130.02490.179 ***0.461 ***−0.0417−0.0008690.0656−0.0357
(0.0807)(0.0160)(0.0427)(0.130)(0.0610)(0.0515)(0.0519)(0.0979)
D.ASYL−0.119 *−0.0239 *0.000262−0.04330.02690.0259−0.03860.0737
(0.0541)(0.0107)(0.0286)(0.0871)(0.0409)(0.0345)(0.0348)(0.0656)
_cons0.0897 ***0.0183 ***0.01710.05020.0958 ***0.119 ***−0.01660.0384
(0.0220)(0.00437)(0.0116)(0.0354)(0.0167)(0.0140)(0.0141)(0.0267)
Sargan test (override of all instruments)Chi2(1) = 4.607
p = 0.0318
Chi2(1) = 1.122
p = 0.2895
Chi2(1) = 1.187
p = 0.2759
Chi2(1) = 0.842
p = 0.3587
Chi2(1) = 0.188
p = 0.6646
Chi2(1) = 0.857
p = 0.3545
N150150150150150150150150
R20.0370.0390.1100.0780.0050.0040.0150.008
2000−2019(1)(2)(3)(4)(5)(6)(7)(8)
D.GDP_capD.GDP_empD.ERD.ER_FD.EDU_TD.EARND.EARN_dispD.Gini
D.IMIG0.1580.0279 *0.172 ***0.428 ***−0.08830.103 *0.03210.0516
(0.0904)(0.0133)(0.0362)(0.109)(0.0764)(0.0516)(0.0471)(0.0826)
D.ASYL−0.184 ***−0.0177 *−0.0165−0.0976−0.0143−0.0191−0.03790.0457
(0.0484)(0.00715)(0.0194)(0.0585)(0.0410)(0.0277)(0.0252)(0.0442)
_cons0.03080.0183 ***0.0227 *0.05350.128 ***0.129 ***−0.0266 *0.0418
(0.0251)(0.00371)(0.0101)(0.0303)(0.0213)(0.0144)(0.0131)(0.0230)
Sargan test (override of all instruments)Chi2(1) = 2.640
p = 0.1042
Chi2(1) = 0.964
p = 0.3261
Chi2(1) = 0.894
p = 0.3445
Chi2(1) = 0.193
p = 0.6601
Chi2(1) = 0.929
p = 0.3352
Chi2(1) = 1.298
p = 0.2545
N190190190190190190190190
R20.0720.0400.1080.0760.0100.0200.0120.011
(c)
Arellano Bond Dynamic GMM Estimations
2000–2015(1)(2)(3)(4)(5)(6)(7)(8)
GDP_capGDP_empERER_FEDU_TEARNEARN_dispGini_st
L.GDP_cap0.716 ***
(0.116)
IMIG0.0673 ***−0.001840.122 ***0.176−0.07190.0568 ***0.227−0.146 *
(0.00823)(0.0101)(0.0241)(0.124)(0.0539)(0.0124)(0.133)(0.0695)
ASYL−0.0355 **0.000909−0.0142−0.0176−0.000256−0.00710 *−0.03550.117 ***
(0.0138)(0.00448)(0.0165)(0.0477)(0.0244)(0.00311)(0.0226)(0.0351)
L.GDP_empl0.628 ***
(0.116)
L.ER0.723 ***
(0.0721)
L.ER_F0.665 ***
(0.100)
L.EDU_T1.014 ***
(0.143)
L.EARN0.976 ***
(0.0525)
L.EARN_disp1.000 ***
(0.163)
L.Gini0.175
(0.225)
Sargan test (overrides are valid)Chi2 = 9.698
p = 1.000
Chi2 = 8.320
p = 1.000
Chi2 = 8.861
p = 1.000
Chi2 = 7.185
p = 1.000
Chi2 = 6.616
p = 1.000
Chi2 = 8.555
p = 1.000
Chi2 = 6.756
p = 1.000
Chi2 = 3.743
p = 1.000
Arellano-Bond test (zero autocorrelation in FD errors)z1 = −1.826
p = 0.0678
z2 = −2.297
p = 0.0216
z1 = −2.018
p = 0.043
z2 = −1.953
p = 0.050
z1 = −1.987
p = 0.046
z2 = −1.045
p = 0.2959
z1 = −2.158
p = 0.030
z2 = −1.372
p = 0.169
z1 = −1.287
p = 0.198
z2 = 0.425
p = 0.670
z1 = −1.284
p = 0.198
z2 = −1.055
p = 0.2914
z1 = −2.545
p = 0.010
z2 = 2.227
p = 0.025
z1 = −1.114
p = 0.265
z2 = −0.779
p = 0.435
N140140140140140140140140
2000–2019(1)(2)(3)(4)(5)(6)(7)(8)
GDP_capGDP_empERER_FEDU_TEARNEARN_dispGini_st
L.GDP_capita1.009 ***
(0.0456)
IMIG0.0429 *0.004090.132 *0.129−0.07820.0976 ***0.0210−0.0401
(0.0173)(0.00976)(0.0579)(0.131)(0.0576)(0.0189)(0.0172)(0.118)
ASYL−0.0565 ***0.00215−0.0334 *−0.02410.00163−0.0343 ***−0.01150.0629 *
(0.0117)(0.00207)(0.0155)(0.0390)(0.0128)(0.00378)(0.0118)(0.0270)
L.GDP_emp0.766 ***
(0.118)
L.ER0.866 ***
(0.0693)
L.ER_F0.836 ***
(0.133)
L.EDU_T0.982 ***
(0.140)
L.EARN1.055 ***
(0.0199)
L.EARN_disp0.964 ***
(0.244)
L.Gini0.452 **
(0.170)
Sargan test (overrides are valid)Chi2 = 9.742
p = 1.000
Chi2 = 8.680
p = 1.000
Chi2 = 6.724
p = 1.000
Chi2 = 6.159
p = 1.000
Chi2 = 6.849
p = 1.000
Chi2 = 7.461
p = 1.000
Chi2 = 4.926
p = 1.000
Chi2 = 3.567
p = 1.000
Arellano-Bond test (zero autocorrelation in FD errors)z1 = −2.262
p = 0.0237
z2 = −2.288
p = 0.0221
z1 = −2.255
p = 0.024
z2 = −1.951
p = 0.051
z1 = −2.522
p = 0.011
z2 = −1.055 p = 0.2914
z1 = −1.995
p = 0.045
z2 = −1.237
p = 0.215
z1 = −1.125
p = 0.260
z2 = 0.299
p = 0.7644
z1 = −1.269
p = 0.204
z2 = −1.126
p = 0.2598
z1 = −1.971
p = 0.048
z2 = 2.029
p = 0.0424
z1 = −2.367
p = 0.017
z2 = −0.144
p = 0.8851
N180180180180180180180180
Note: Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: authors’ research.
Table 3. Results of the SEM models developed for the comprehensive analysis of the immigration shaping factors and their economic consequences upon host countries—Overall model estimation.
Table 3. Results of the SEM models developed for the comprehensive analysis of the immigration shaping factors and their economic consequences upon host countries—Overall model estimation.
(1) 2000–2015(2) 2000–2019
IMIG_st
EARN_st−0.315 ***−0.0525
(0.0761)(0.0725)
EDU_T_st0.326 ***0.0526
(0.0785)(0.0642)
UR_F_st−0.396 ***−0.409 ***
(0.0672)(0.0774)
KOF_T_st−0.724 ***−0.747 ***
(0.0529)(0.0586)
_cons−0.000000410−0.0371
(0.0513)(0.0652)
ER_st
IMIG_st−0.180 *−0.0251
(0.0729)(0.0576)
ASYL_st0.159 *0.0897 **
(0.0698)(0.0307)
ALMPs_st0.534 ***0.511 ***
(0.0724)(0.0587)
PLMPs_st−0.527 ***−0.508 ***
(0.0702)(0.0594)
_cons−0.000000641−0.0562
(0.0610)(0.0583)
ASYL_st
KOF_T_st−0.459 ***−1.057 ***
(0.0801)(0.134)
POV_r_st−0.187 *−0.533 ***
(0.0762)(0.131)
LE_st−0.04940.0758
(0.0767)(0.117)
_cons−0.0000007820.329 *
(0.0714)(0.134)
GDP_cap_st
ER_st0.553 ***0.534 ***
(0.0659)(0.0595)
_cons−2.09 × 10−8−0.118
(0.0657)(0.0611)
var(e.IMIG_st)
_cons0.422 ***0.773 ***
(0.0471)(0.0773)
var(e.ER_st)
_cons0.595 ***0.625 ***
(0.0665)(0.0625)
var(e.ASYL_st)
_cons0.817 ***3.280 ***
(0.0913)(0.328)
var(e.GDP_cap_st)
_cons0.690 ***0.746 ***
(0.0771)(0.0746)
N160200
Note: Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: authors’ research.
Table 4. Results of the SEM models developed for the comprehensive analysis of the immigration shaping factors and their economic consequences upon host countries—Total Effects upon GDP per capita (dependent variable).
Table 4. Results of the SEM models developed for the comprehensive analysis of the immigration shaping factors and their economic consequences upon host countries—Total Effects upon GDP per capita (dependent variable).
Total Effects upon GDP per Capita (Dependent Variable)
2000−20152000−2019
GDP_capCoefficients
(Standard Errors)
p > |z|Coefficients
(Standard Errors)
p > |z|
IMIG−0.09966310.013−0.01338780.664
(0.0403125)(0.030773)
ASYL0.55299240.0230.04788070.003
(0.0385967)(0.0163852)
ER0.08802770.0000.53391140.000
(0.0658692)(0.0595118)
UR_F0.03945460.0280.00546910.665
(0.0179352)(0.0126287)
ALMPs0.29541180.0000.27257240.000
(0.0533060)(0.0436363)
PLMPs−0.29145270.000−0.27139880.000
(0.0520867)(0.0438204)
EARN0.03134550.0400.00070270.709
(0.015237)(0.0018859)
EDU_T−0.03246260.040−0.00070380.701
(0.0157682)(0.0018335)
KOF_T0.03173360.281−0.04062890.073
(0.0294489)(0.0226649)
POV_R−0.01647230.101−0.0255150.022
(0.0100517)(0.0111169)
LE−0.0043450.5370.0036310.526
(0.0070383)(0.005732)
Source: authors’ research.

Share and Cite

MDPI and ACS Style

Noja, G.G.; Cristea, S.M.; Yüksel, A.; Pânzaru, C.; Drăcea, R.M. Migrants’ Role in Enhancing the Economic Development of Host Countries: Empirical Evidence from Europe. Sustainability 2018, 10, 894. https://0-doi-org.brum.beds.ac.uk/10.3390/su10030894

AMA Style

Noja GG, Cristea SM, Yüksel A, Pânzaru C, Drăcea RM. Migrants’ Role in Enhancing the Economic Development of Host Countries: Empirical Evidence from Europe. Sustainability. 2018; 10(3):894. https://0-doi-org.brum.beds.ac.uk/10.3390/su10030894

Chicago/Turabian Style

Noja, Graţiela Georgiana, Simona Mirela Cristea, Atila Yüksel, Ciprian Pânzaru, and Raluca Mihaela Drăcea. 2018. "Migrants’ Role in Enhancing the Economic Development of Host Countries: Empirical Evidence from Europe" Sustainability 10, no. 3: 894. https://0-doi-org.brum.beds.ac.uk/10.3390/su10030894

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop