Trends in African Migration to Europe: Drivers Beyond Economic Motivations

Abstract

The current migration and refugee crisis in Europe requires an understanding of the different migration drivers beyond the well-known economic determinants. In this article, we view migration from a broader human security perspective and analyze the determinants of regular and asylum seeker migration flows from Africa to Europe for the period 1990 to 2014. Our results show that, in addition to economic determinants, a combination of push and pull factors influences migration decisions of individuals. In particular, rising political persecution, human rights violations, ethnic tensions, political instability, and civil conflicts in African source countries are all significantly associated with increased migration flows into European destination countries. Therefore, our results underscore the need for the European Union and European countries to collaborate with the source countries, not only in terms of supporting economic development in the source countries but also in promoting human security: human rights, democracy, peace, and social stability.

Keywords

international migration asylum seeker refugee crisis human security Poisson pseudomaximum likelihood

Introduction

Migration and the refugee crisis are high on the policy agendas of European countries, from economic, security, and social standpoints. In the current wave of migration, Europe is witnessing a mixed-migration phenomenon where a large number of economic migrants are joining asylum seekers in their journey to reach the European continent (Bertoli, Brücker, and Moraga 2013; Park 2015). Each year hundreds of thousands of immigrants flow into Europe mainly from Africa, the Middle East, and South Asia.¹ With such a large number of migration inflows, European countries are said to have reached a breaking point in their ability to meet the European Union (EU) standards for receiving immigrants and facilitating asylum applications (Banulescu-Bogdan and Fratzke 2015). Furthermore, the present-day migrant influx has imposed internal “political fatigue,” with nationalist parties gaining momentum in many EU member states and with security tensions rising due to terrorist-linked incidents in some countries (Park 2015). On the other hand, thousands of people perish every year while attempting to cross the Mediterranean Sea (Telschow 2014). These events have created mounting pressure on European governments, at least by groups supporting human rights, to work and spend more on rescue missions to enable acceptance of a more substantial number of immigrants. Therefore, migration is now a prominent feature in the economic, social, and political landscape of European countries (Kerr and Kerr 2011).

In response to these tense situations, European governments are actively working to cut the flows of migrants and asylum seekers across the Mediterranean sea in partnership with African governments (Garcia Andrade and Martin 2015). Important initiatives in this regard include the “EU-Africa Declaration on Migration and Mobility” of 2014 in which European and African governments pledge to combat human trafficking, and to facilitate the return and the readmission of migrants whose asylum applications have been refused (https://au.int/sites/default/files/pages/32899-file-5._the_eu_africa_declaration_on_migration_and_mobilty_2014.pdf). This declaration also emphasizes the importance of addressing the root causes of irregular migration such as by providing employment opportunities for the youth at regional level. In the same year, the so-called Khartoum Process and Rabat Process were established with a special focus on preventing and fighting migrant smuggling and human trafficking in the Horn of Africa and Central, Western, and Northern Africa, respectively. In November 2015, the Valletta Summit brought together the largest number of African and European heads of states and governments and concluded with the EU setting up an Emergency Trust Fund to promote development in Africa, in return for African countries to help European countries in the crisis (http://www.consilium.europa.eu/en/meetings/international-summit/2015/11/11-12/). In June 2016, the EU establishes a wide-ranging Migration Partnership Framework (MPF), which aims to coordinate collaborations with African and other third country governments (European Commission 2016).

These intensified cooperations of the EU with African governments, including those that are authoritarian and which are accused of severe human rights violations and political persecution, have caused significant controversy. For instance, critics argue that by basing development aid and foreign relations on countries agreement to cooperate with EU migration control objectives, the EU is making a significant policy change away from putting human rights as a central point of EU foreign policy (Castillejo 2017; Human Rights Watch 2018; Oette and Babiker 2017).²

A number of factors might be behind the EU’s apparent policy shift away from its long-held policy of defending human rights in Africa toward stronger border control and quicker returns of “illegal” migrants. Obviously, one of the reasons that the large number of migrant flows might have forced the EU to prioritize the short-term objective of reducing the number of immigrants above its principle of defending human rights. Moreover, it seems that the EU considers that economic factors, rather than human rights violations, ethnic tensions, and civil conflicts, are key reasons for the African migration flows to Europe. Indeed, for instance, the “migration compact”—an important contribution to the MPF document by the Italian government—explicitly state that while migrant flows through the Eastern Mediterranean route include both refugees and economic migrants, “flows through the Central/Western Mediterranean route are composed mainly by economic migrants” (Italian Government 2016, 1).

Despite the seemingly divergent views held by European governments and groups supporting human rights on the role of human security factors in the recent migration flows from Africa to Europe, there is surprisingly little empirical evidence to substantiate either of these stances. While a few extant studies have examined determinants of African migration flows, they have however either focused exclusively on intra-African migration flows or studied international migration from Africa together with intra-African flows. For instance, investigating trends of migration flows in sub-Saharan African (SSA) countries for the period 1965 to 2005, Naudé (2010) finds that armed conflict and lack of job opportunities are the most important determinants. However, he employs net migration flows data, which are obtained as the difference between emigration and immigration per 1,000 inhabitants. Consequently, Naudé (2010) does not examine determinants of bilateral migration flows from a given sending African country to specific host countries. Moreover, the net migration data include intra-African migration, which is the most prevalent form of African migration (Lucas 2015). As a result, it is unclear to what extent his findings could be used to explain trends in migration flows from Africa to Europe. Similarly, Ruyssen and Rayp (2014) examine intraregional migration in SSA during the period 1980 to 2000. However, they do not consider migration flows from Africa to Europe. Lucas (2015) provides an extensive literature survey and empirical study on the causes, patterns, and consequences of migration in Africa. In particular, by estimating a gravity model of African migration flows to 220 states and territories worldwide (including African states), he finds that violent conflicts, but not the level of democratization, are important drivers of African migration flows.

The present study aims to fill the aforementioned gap in the literature by providing a thorough empirical examination of the role played by human security factors in explaining trends in African migration to Europe during the period 1990 to 2014. We consider the most basic level of human security, that is, it includes freedom from fear (threats to the safety of people), freedom from want (threats to basic needs), and freedom to live in dignity (threats to human rights and by extension access to services and opportunities; see, for instance, Anand 1994; Gómez and Gasper 2013). We thus consider potential factors that provoke migration and displacement including wars, civil conflict, economic deprivation, violation of human rights, and oppressive regimes (Erdemir et al. 2008; Shrestha 2017).³

For our empirical investigation, we construct a new panel data set on bilateral migration flows for a large number of the destination (21 European) and source (51 African) countries for the period 1990 to 2014. While the choice of the period 1990 to 2014 is dictated by data availability, it is, however, the most relevant period for the current migration crises in Europe, since European countries have experienced a dramatic increase in the flow of African migrants in this period and African migrants have started to become a visible part of the migrant stock in Europe.

Besides the trends in the regular migration flows from Africa to Europe, this article studies the bilateral migration trends by additionally considering the flow of asylum seekers. It is noteworthy that this approach is typically important in the study of noneconomic determinants of migration and for dealing with the African forced-migration trends that occur in the absence of human security. In this regard, the existing empirical literature on bilateral migration largely focuses on the regular migration trends, either the flow of labor migrants or migrant stocks, by excluding refugees and asylum seekers (e.g., Ortega and Peri 2009; Mayda 2010; Fitzgerald, Leblang, and Teets 2014). However, given that the rising number of asylum seekers is at the heart of the recent refugee crisis in Europe, it is important to examine to what extent human security factors drive the flow of African asylum seekers to Europe.

To examine the determinants of bilateral migration flows from Africa to Europe, we estimate a gravity model that is similar to the model of Ortega and Peri (2013). Since the seminal work of Tinbergen (1962), the ordinary least squares (OLS) estimator has been widely used to estimate various versions of the gravity model, both in the trade literature and in migration studies. A well-known drawback of the OLS approach, however, is related to the fact that migration flows between pairs of countries may be zero in a substantial percentage of observations, and omitting those zero observations biases the regression results (Yotov et al. 2016). We address this methodological challenge in estimating gravity models by using the Poisson pseudomaximum likelihood (PPML) estimator, which is particularly suitable in regressions where the dependent variable has a significant proportion of zeroes (Beine and Parsons 2015).

Our results show that broader human security factors are significant determinants of the “South–North” migration flows. In accordance with the existing literature, income gaps between African and European countries remain a strong determinant of migration flows. However, income gaps are not the only important reason for the rise in the migrant flows to Europe: broader human security factors in Africa are equally important determinants of both regular and asylum seeker migration trends. Indeed, we find that poverty, violent civil conflicts, political persecution, human rights abuses, and ethnic tensions have a substantial influence on migration across our entire set of specifications. Similar results are also obtained when we consider the flow of asylum seekers as an alternative dependent variable. Our results are robust to using alternative model specifications, to excluding North African countries, and to employing OLS instead of the PPML estimator.

The rest of this article is organized as follows. The second section introduces the empirical specification of our model. The third section describes our data in detail. The fourth section provides the results and the fifth section concludes. Finally, the Online Appendixes present Supplemental Material.

Empirical Model Specification and Conceptual Foundations

Empirical Model Specification

Since the seminal work of Tinbergen (1962), the gravity model of trade has been widely used to study the effects of trade policies on dyadic trade flows (see Online Appendix A for a review of theoretical foundations). This model specifies international trade as a positive function of the attractive “mass” of two economies and a negative function of the distance between them. The migration literature has vastly implemented the gravity model (e.g., Lewer and Van den Berg 2008; Mayda 2010; Beine and Parson 2015; Figueiredo, Lima, and Orefice 2016). Beine, Bertoli, and Fernández-Huertas Moraga (2016) summarize this strand of literature and provide a practical guide on the empirical implementation of the gravity model in migration studies.

The current article closely follows the gravity model developed in Ortega and Peri (2013). Accordingly, taking as a starting point the random utility maximization theoretical models developed by Beine, Docquier, and Özden (2011), Grogger and Hanson (2011), and Ortega and Peri (2013), in which income maximization problems or wage differentials are a driving force to make a migration decision, we emphasize the broader human security conditions. Specifically, we analyze a number of political and sociocultural factors that may influence the individual’s decision to move from his or her current location. Furthermore, similar to Ortega and Peri (2013) and Beine, Bertoli, and Fernández-Huertas Moraga (2016), our empirical model specification considers multiple destinations.

Formally, out of the set of N global countries, the individual i from his or her source country $s \in S = s_{1}, s_{2} ...... s_{n}$ , where $S \subseteq N$ , makes a decision of whether to stay in his or her source country s or to migrate to the destination country $d \in D = d_{1}, d_{2} ... d_{n}$ , where $D \subseteq N$ . It is noteworthy that, as indicated in, among others, Kennan and Walker (2013), individuals are assumed to have rational expectations. Thus, they make an informed decision to migrate in seeking their maximum utility.⁴ Therefore, following Ortega and Peri (2013), we assume the following utility function:

V_{s d t} = α_{s} + α_{d} + α_{d t} + β_{1} W_{s d t} + β_{2} F_{s d t} + β_{3} X_{s d t} + θ Z_{s d t},

where $V_{s d t}$ represents the utility in country d for individuals from country s and it captures not only the average earnings but also the average security of individual i by leaving for destination country d. Moreover, $α_{s}$ and $α_{d}$ are time-invariant source-specific push factors fixed effects and time-invariant destination-specific pull factors fixed effects, respectively. In addition, $α_{d t}$ denotes destination-country fixed effects and it varies over time and across countries. Hence, the fixed effects capture the role of amenities, destination-specific cost heterogeneity, and variables that may change through time (essentially immigration policy, culture, and attitudes toward immigrant community) and affect the choices made by the immigrants. Although our primary focus is on human security factors at the source, in some specifications, we also consider particular pulling factors.

In equation (1), a vector $W_{s d t}$ represents the value of expected earnings in the source country s and the destination country d in year t. In this case, the income differential is measured using per capita gross domestic product (GDP) of countries s and d. The vector $F_{s d t}$ represents levels of freedom (free from all forms of violence or political persecution) in the source country s and the destination country d that vary over time t. To estimate the effect of freedom on migration decision, we consider a number of proxy variables, which include political stability (internal and external conflicts), institutional quality (government stability, law and order, control of corruption and bureaucratic quality), ethnic and religious tensions (fractionalization and polarization), democracy, autocracy, political rights, and civil liberties of the source countries. In particular, we follow Baudassé, Bazillier, and Issifou (2018) to frame the role of political institutions on migration flows using the exit and voice dichotomy of Hirschman (1970). Accordingly, if people have the freedom to express their dissatisfaction through the existing institutional frameworks, they choose the voice option and prefer to stay at home. Whereas violations of basic rights due to poor quality of institutions lead them to choose the exit option and leave their country. The vector $X_{s d t}$ represents other control variables, notably, the socioeconomic factors such as urbanization and socioeconomic conditions. The vector $Z_{s d t}$ stands for country-pair overall migration costs that affect migration decisions. These factors include time-varying factors such as migrant networks and time-invariant variables such as geographic distance, common language, common legislation, and colonial legacy.

Considering separability in migration costs and including an error term in equation (1), we estimate the following estimable model of the flow of regular immigrants ( $M_{t s d}$ ) from the source country s to the destination country d at time t:

l n (M_{t s d}) = β_{1} l n (W_{s d t - 1}) + β_{2} F_{s d t - 1} + β_{3} X_{s d t - 1} + θ Z_{s d t} + α_{s} + α_{d} + α_{d t} + ∊_{s d t} .

Equation (2) could be estimated using OLS.

However, migration between pairs of countries may be zero in a substantial percentage of observations, and omitting those zero observations biases the regression results. In particular, due to the fact that equation (2) is a pseudogravity model in a double log form, a large number of observations could be dropped because of the zero values in the dependent variables $M_{t s d}$ . Hence, standard regression methods, such as the OLS, require omitting observations with zero values, which leads to inconsistent estimates of the coefficients due to selection bias. Furthermore, a second source of bias is related to the fact that if the variance of $∊_{s d t}$ depends on one or more of the determinants of $M_{t s d}$ , then the expected value of $∊_{s d t}$ will also depend on some of the regressors in the presence of zeroes (Silva and Tenreyro 2006).

These methodological challenges in estimating gravity models can be addressed using the PPML estimator (Silva and Tenreyro 2006; Beine and Parsons 2015). Using PPML, we estimate the exponential of the gravity model as

M_{t s d} = exp [β_{1} l n (W_{s d t - 1}) + β_{2} F_{s d t - 1} + β_{3} X_{s d t - 1} + θ Z_{s d t} + α_{s} + α_{d} + α_{d t}] u_{s d t},

where $u_{s d t} = exp (∊_{s d t})$ is a multiplicative error term.

In addition to the regular migration flows ( $M_{t s d}$ ), we estimate a gravity model for the flow of asylum seekers ( $A S_{t s d}$ ) from the source country s to the destination country d in year t.

Conceptual Foundations

A growing body of theoretical and empirical literature on international migration has documented that individuals have complex and often overlapping motivations for leaving their places of origin, including income (Naudé 2010; Dutta and Roy 2011; Clemens, Özden, and Rapoport 2014; Docquier, Ruyssen, and Schiff 2018), political instability (Naudé 2010; Docquier, Ruyssen, and Schiff 2018; Clemens 2017; Shrestha 2017), migrant networks (Beine and Parsons 2015), institutions (Baudassé, Bazillier, and Issifou 2018), and climate change (Beine, Coulombe, and Vermeulen 2015). In the following, we discuss the conceptual foundations for the set of human security variables that we consider in the current study by categorizing them as economic factors, civil conflicts, ethnic tensions, and institutional quality of source countries.

Economic factors

A large body of literature documents that a significant difference in the average income in terms of average GDP per capita (GDPPC) between the origin and destination countries is a principal determinant of international migration. In other words, relatively lower main income at the source countries in comparison with per capita income at the destination countries motivates potential immigrants to decide to migrate. In our context, hence, the actual economic deprivation and abject poverty of most African countries will likely have an enormous push effect on the migrants and refugees of Africa. On the other hand, economic development and relatively high personal incomes in Europe attract immigrants.

Civil conflicts

There are enough evidences that show that individuals will have a greater incentive to migrate when there are civil conflicts than when there are no conflicts in their home countries (see, for instance, Weiner 1992; Schmeidl 2001; Ibáñez and Vélez 2008). In the absence of safety and security, the expected returns of labour, development projects, and investments are significantly decreased. Since civil conflicts directly affect both the security and livelihood of individuals’ people are likely to be forced to migrate in seeking for an alternative survival strategy. Controlling for economic and other determinants, on the average one may, therefore, expect to see a higher number of migrants from source countries devastated by civil conflicts. The African political condition, in general, is characterized by historical injustices and oppressive governance structures (Ongayo 2008). Since their independence, many states have witnessed civil wars, large-scale mass killings of civilians, and other forms of direct political violence for decades (Dunn 2009). United Nations Conference on Trade And Devolopment (2018) illustrated that in the African context, severe conflicts often lead to a significantly increased flow of internally displaced people or refugees, if they flee across borders. Moreover, conflicts can also be a driver of economic migration. Hence, conflict is one of the human security indicators that are explored in this article.

Institutions

Weak performance of institutions in the source countries may be a sufficient motive for emigrating in search of institutions which perform better (Baudassé et al. 2018). Very often totalitarian regimes are a push factor of migration. The lack of democracy, political rights, and civil liberties and endemic corruption act as push factors for migrants seeking greater freedoms (Solimano 2010). In line with the Hirschman’s (1970) “exit and voice” dichotomy, individuals may decide to migrate when institutions are not satisfactory and fundamental human rights are violated.

Mostly in countries where there are autocratic political systems and state sponsored persecution, harassment, discrimination and torture people who disagree with the policy or ideology of the government, and/or have minority religious beliefs or ethnic backgrounds (Solimano 2010) are pushed toward migration. In nondemocratic countries, even when individuals are not physically persecuted restraints of fundamental freedoms may ultimately motivate them to leave their country of origin. In sum, if the political environment is hostile, then the tendency is that the economic outcome is most likely to be poor. Therefore, such situations trigger migration for political and economic reasons. Relating to this Solimano (2005) argues that in nondemocratic cases “individuals who are dissatisfied with the prevailing political and economic conditions, may choose to exit their home country.” (p. 264) Hence, we assume that better institutional quality reduces migration flow while poor institutional performances motivate individuals to leave their home countries and relocate to more democratic and safer countries where they can pursue better freedom, protection, education, and careers. Accordingly, Naudé (2010), Ruyssen and Rayp (2014), and Lucas (2015), among others, show that the African migration flow—at least the intra-African migration—is profoundly influenced by the political setup of the continent. Furthermore, due to the overly repressive character of the regimes, the majority of African countries have been receiving the lowest rankings on political rights and civil liberties for decades (see https://freedomhouse.org/regions/sub-saharan-africa). These preceding events have made Africans vulnerable to displacement including migration within and emigration from the continent.

Ethnicity

Ruble (1989) ethnic identity “developed, displayed, manipulated or ignored in accordance with the demands of a particular situation.” (p. 401) In this context, it is noteworthy to consider that social identity serves as a structural foundation for potential group formation and social conflicts. When conflict arises, ethnic identities may result in suboptimal behavior (Constant and Zimmermann 2011). Existing literature has argued that ethnic tensions raises the likelihood of waging civil conflicts and engenders a kind of “structural violence.” Ethnic heterogeneity increases the probability of civil conflicts and civil wars (see, for instance, Reynal-Querol and Montalvo 2005; Esteban, Mayoral, and Ray 2012; Giménez-Gómez and Zergawn 2018). Fearon and Laitin (2003) further stipulate that between 1945 and 1999, about 51 percent of major civil wars originated by way of ethnic conflicts. Moreover, where there are ethnic tensions, women, children, and other vulnerable social groups are exposed to various forms of sexual, physical, and nonphysical violence in their relation to ethnic-national identities (Korac 1998). Hence, we assume that civil conflict and structural violence such as molestation, marginalization, ethnic tension, segregation, and the development of an underclass along the line of ethnic identities significantly increases the flow of international migration.

It is noteworthy that ethnic heterogeneity is measured by ethnic polarization and ethnic fractionalization. Ethnic polarization measures the existence of deep cleavages in society within a given country based on perceived distances between interethnic groups and group size of each ethnicity. Ethnic fractionalization, on the other hand, measures the likelihood that two randomly chosen people will be a part of different ethnic groups in a given country. Bang and Mitra (2013) show that ethnic-related conflicts increase the fraction of skilled labor migration.

In this context, the contemporary African political setup is profoundly influenced by ethnic identity. The interethnic relationships in Africa, especially in the political arena, are associated with competition, exclusiveness, the prevalence of genocidal violence, and conflicts among ethnic groups (Berman 1998; Daley 2006). On top of the political violence and instabilities, human and democratic rights violations are prevalent across the African continent (Mutua 2009).

Data

We construct a new panel data set with information on migration flows and asylum seeking as well as on several macroeconomic, political, and institutional factors covering twenty-one European countries of destination and fifty-one African countries of origin from 1990 to 2014 (see the list of countries in Tables B.1 and B.2 of Online Appendix B). While the choice of the post–Cold War period is dictated by data availability, it has two main merits. First, it is the most relevant period with respect to the current migration crises in Europe. In the trends of international migration toward Western Europe since the Second World War, there are three distinct periods: (1) the labor migration from the 1950s till the beginning of the 1970s, (2) the family migration in the mid-1970s, and (3) the third wave of the international movement that emerged in the post–Cold War era (Geddes and Scholten 2016). It is noteworthy that there has been a marked surge in the number of immigrants, specially asylum seekers, to Europe since the early 1990s. Second, this period also allows us to consider some of the former Eastern European countries, where data are typically available after 1990.

In the following two subsections, we describe in detail the dependent and explanatory variables that we use in the current study. Specifically, we first present the sources and the construction of migration data, both regular flows and asylum seekers, which are our alternative dependent variables. Subsequently, we discuss the explanatory variables, which include several economic and political determinants of international migration (Table 1 reports summary statistics for these variables).

Table 1.

Summary Statistics.

Variables	N	Mean	SD	Min	Max
Bilateral migration flow	26,775	260.5	1,815	0	84,978
Bilateral asylum seekers	26,775	74.7	369.6	0	15,819
GDP per capita destination (in thousand)	26,723	31,624	21,429	1,445	116,664
GDP per capita source (in thousand)	26,670	1,624	2,709	64.81	23,348
Urban population at source (in thousand)	26,712	606,591	956,926	34,481	83,300
Political stability at source	26,775	50.4	5.88	42	61
Network (stock migration in thousand)	26,775	7,212	309	27	183
Common language	26,775	0.133	0.339	0	1
Colonial ties	26,775	0.0523	0.223	0	1
Distance	26,250	5,699	2,076	716.6	16,632
Common legislation	26,775	0.284	0.451	0	1
Ethnic fractionalization at source	24,150	0.615	0.267	0.05	0.959
Ethnic polarization at source	24,150	0.527	0.209	0.014	0.897
Civil conflict at source	26712	.210	.407	0	1
Political regime characteristics at source	26,775	0.231	5.456	−9	10
Political rights at source	26,775	0.392	0.488	0	1
Civil liberates at source	26,775	0.471	0.499	0	1
Government stability at source	19,425	2.756	1.867	0	1
Socioeconomic conditions at source	19,425	3.982	1.649	0	8
Corruption at source	19,425	2.406	0.962	0	5
Law and order at source	19,425	3.089	1.182	0	6
Democratic accountability at source	19, 425	3.021	1.286	0	6
Bureaucracy quality at source	19,425	1.379	0.892	0	4
Investment portfolio at source	19,425	7.16	1.46	4	9
External conflict at source	19,425	10.24	0.862	0	12
Religious tension	19,425	1.8	1.452	9	12

Note: N, mean, SD, min, and max represent number of observations, mean, standard deviation, minimum, and maximum, respectively. GDP = gross domestic product.

Migration Flows and Asylum Seeking

The main dependent variable in equation (2) is the annual migration inflows from the source country s to the destination country d, in a given year t $M_{t s d}$ . These migration data measure the yearly inflow of foreign-born population by nationality into the hosting countries. Note that the inflow migration data exclude temporary visitors with a tourist visa or people who travel for reasons of study, medical, and business purposes.⁵

To construct the migration inflow series, we use two complementary data sources, which help us to cover the entire sample period. The first source is the 2015 update of the international migration flows data of the International Migration Report (IMR) of the United Nations (United Nations 2015a). This database contains time series data on the flows of international migrants as recorded by forty-five destination countries.⁶

This database considers legal migration only and presents both inflows and outflows according to the place of birth, citizenship, place of previous, or next residence, both for foreigners and nationals, as reported by each country’s national agencies in charge of collecting migration data. For most African source countries, the database covers the period from the early 1990s until 2013, despite missing data for some bilateral countries. The second source of data is the Organization for Economic Cooperation and Development (OECD) “International Migration Database” (IMD), which comprises migration inflows data starting from the mid-1990s up to 2014.⁷ Similar to the IMR, IMD contains time-series data on the inflows of foreign populations into thirty-five OECD countries for which data are available. However, IMD has a broader coverage than IMR.

In order to merge these two databases, it is critical to ensure that the two databases have uniform definitions of migration. The majority of the destination countries report migration data that are collected from a population register or are based on the number of residence permits issued. We observe that in most cases these databases embrace overlapping figures when data are available. Hence, our final migration inflow series is constructed mainly by using IMD, which has a broader coverage of countries and periods. The IMR data are used to fill missing values. In rare cases, we fill missing data using simple averages between data of the previous year and the following year.⁸

The African migration trend map in Figure C.1 in Online Appendix C shows the trend in migration flows from Africa to Europe. In general, there has been a significant rise in the number of Africans migrating to the selected European destination countries. Closer observations of the data reveal that African immigrants are highly concentrated in a few Western European countries. In particular, the maps display that the major destinations of African migrants over the years are France, Italy, Spain, and the United Kingdom, although a considerable number of Africans have also migrated into Belgium, Germany, and Sweden. A large number of African immigrants in France and the United Kingdom may be partly linked to the fact that about 65 percent of the contemporary African nations are former colonies of these two countries. It is well-established in the literature that colonial ties increase migration flows by creating, for example, common official languages, cultural attachments, social networks, and business relations (Fawcett 1989). Southern European countries, such as Italy and Spain, did not have many African colonies and hence have weaker colonial ties with African countries. African migrants inflows into these Mediterranean countries might have been induced by their strong economic performance since the 1980s, as well as their growing economic integration with other European countries (Bonifazi et al. 2009; Ortega and Peri 2013). However, this might also reflect the fact that many immigrants use Southern European countries as a transition point to move to other Western European countries. Moreover, as Online Appendix Figure C.1 shows, the North African countries Algeria, Egypt, Morocco, and Tunisia have been sending a consistently high number of migrants to Europe in the two and a half decades under study. This might be related to the fact that geographical proximity as well as the presence of a large diaspora attract immigrants due to lower migration costs.

As an alternative dependent variable, we use data on yearly inflows of asylum seekers into the European hosting countries by African countries of origin from 1990 to 2014. Utilizing the asylum seeking data helps to address two crucial issues. First, the widely applied migration inflow data comprise the regular inflow of immigrants into the hosting countries only. As a result, the database omits the significant number of asylum seekers, which are the primary source of the refugee crisis in Europe.⁹ Second, we check the robustness of our results on the political determinants of extracontinental migration by using the asylum seeking data.

According to the 1986 definition of the United Nations High Commissioner for Refugees (UNHCR), an asylum seeker is a person who has sought protection as a refugee, but whose claim for refugee status has not yet been assessed (http://www.unhcr.org/excom/exconc/3ae68c43c0/detention-refugees-asylum-seekers.html). Data on the inflow of asylum seekers come from the IMD database, which in turn is based on data provided by the UNHCR. The UNHCR regularly produces complete statistics on refugees and asylum seekers in OECD countries and worldwide.¹⁰ In rare cases, we also use the original UNHCR database to complement missing data (downloadable at www.unhcr.org/figures-at-a-glance.html).

Figure C.2 in Online Appendix C exhibits the inflow of African asylum seekers into the European destination countries. The maps display an upward trend in the flow of asylum seekers from Africa into many European countries: France, Germany, Italy, Sweden, Switzerland, and the United Kingdom. In particular, the annual inflow of African asylum seekers has markedly risen in Germany since 2010. Additionally, Italy is a significant entry point for African refugees. Despite being stricken by the Euro-zone crisis (Mody and Sandri 2012), tens of thousands of asylum seeker migrants continue to board overcrowded and unsafe boats heading to Italy, putting their lives in grave danger.

Economic and Political Determinants of Migration

To substantiate the effects of the broad human security factors of international migration, we consider several economic, political, and social determinants of migration as explanatory variables, as aforementioned in Conceptual Foundations subsection.

Economic factors

To capture the impact of economic drivers in the Africa to Europe migration flows, we use the logarithm of GDPPC in the source and the destination countries.¹¹ Our primary source of the GDPPC data is the “National Accounts Main Aggregates Database” of the Economic Statistics Branch of the United Nations Statistics Division (see https://unstats.un.org/unsd/snaama/introduction.asp).

Political factors

To investigate the effects of political factors on migration flows from Africa to Europe, we employ several political indices. We measure political instability by means of the civil conflict incidence, which is an indicator variable that takes a value of 1 if there is a new or existing conflict in year t, and 0 otherwise. We obtain the data on conflict incidence from the Armed Conflict Database of the Uppsala Conflict Data Program and the Peace Research Institute of Oslo.¹² This database codes armed conflicts at a low threshold of twenty-five battle-related deaths per year in conflicts where there is the use of armed force between two parties, of which at least one is the government (Pettersson and Wallensteen 2015).

Additionally, while the index of ethnic fractionalization measures the probability that two randomly selected individuals in a country will belong to different ethnic groups, the ethnic polarization indexes measure the normalized distance of a particular distribution of ethnic groups from a bimodal distribution. Data for both ethnic fractionalization and polarization indices are obtained from Reynal-Querol and Montalvo (2005). Following Esteban, Mayoral, and Ray (2012), we use time-invariant versions of these variables, since short-run changes are likely to be correlated with the incidence of conflict.

To assess how the characteristics of the political regimes affect the bilateral migration flows, we use indicators for democratic and autocratic patterns of authority. In the Polity IV database, the polity series contains coded annual information on the level of democracy and autocracy, both ranging from 0 (low) to 10 (high; Marshall, Jaggers, and Gurr 2012).¹³ Following Esteban, Mayoral, and Ray (2012) and Giménez-Gómez and Zergawu (2018), we transform these indices into time-invariant dummy variables as short-run changes in these measures are likely to be correlated with the incidence of conflicts. Specifically, a country receives a time-invariant 1 (considered democratic) if it has received a democracy score higher than or equal to 4 for 40 percent of the years and 0 otherwise. The autocracy dummy is also computed in the same manner.

Furthermore, we test the effect on bilateral migration of civil liberty and political rights variables, which are measured on a scale from 1 to 7 (where 1 represents the highest levels of liberties and political rights and 7 indicates the lowest level). For ease of interpretation, we converted these indicators to time-varying dummy variables. Specifically, a country is considered to have a favorable rating for civil liberty or political rights in a specific year (dummy variable takes on 1) if it receives a rating less than or equal to 4, and 0 otherwise. The data source for these variables is Freedom House (2017; see https://freedomhouse.org/report-types/freedom-world).

To examine the effects of overall political stability of the countries on bilateral migration flows, we use the political risk index from the International Country Risk Guide (ICRG) data set. The political risk rating includes twelve weighted variables covering both political and social attributes. The risk components include government stability, socioeconomic conditions, investment profile, internal conflict, external conflict, corruption, military in politics, religious tensions, law and order, ethnic tensions, democratic accountability, and bureaucracy quality. In the data set, the minimum number of points that can be assigned to each component is zero, while the maximum depends on the indexed weight that component is given in the overall political risk assessment (Howell 2011).

Finally, we also consider proxies for migration costs using geographical and cultural distances. In the analysis of costs of international migration, Beine and Parsons (2015) note that migrant networks play a fundamental role in determining migration flows from two channels. In terms of the assimilation channel, networks affect private costs and benefits of migration, and from the point of view of the policy channel, they lower legal entry barriers through family reunification programs (Beine and Parsons 2015). Hence, we estimate both migration flows and asylum seeking trends by controlling for migrant networks using bilateral migration stocks. Our migrant network or diaspora variable, which is again taken from the IMD database, is defined as the bilateral migrant stock in the beginning of the year to which a flow corresponds. Additionally, we use bilateral geographical distances between the two capital cities (in kilometers) and dummy variables for the common language, common legal origin, and colonial ties (if a source country is a former colony of a destination). These variables are widely used in the migration literature as important determinants of migration decision (see, for instance, Taylor 1994; Leblang, Fitzgerald, and Teets 2009; Kim and Cohen 2010; Mayda 2010; Ortega and Peri 2013; Ruyssen and Rayp 2014; Beine, Coulombe, and Vermeulen 2015; Docquier, Ruyssen, and Schiff 2018).

Results

In this section, we discuss the empirical results on the determinants of migration flows from Africa to Europe using data on both regular and asylum seeker migration flows. First, we discuss determinants of the regular migration flows from African countries to Europe as specified in equation (3). Subsequently, we reestimate equation (3) using the flows of asylum seekers as an alternative dependent variable.

Baseline Results

In equation (3), our dependent variable is annual migration flows between bilateral countries. Our main hypothesis, in this case, is that human security conditions significantly influence the international migration flows in different ways. Specifically, GDPPC at the source and the destination countries are expected to have opposite effects on migration flows. While higher income per capita at the source country is expected to show a negative effect on migration flows, income per capita at the destination country is expected to display a positive impact. Improvements in democratization and human rights protections (both political and civil rights) at the source country are expected to reduce the outflow of migrants, whereas civil wars, institutional autocracy, and ethnic tensions are expected to increase the rate of migration outflow. Concerning variables such as proxy that measure migration costs, geographical distance is expected to have a negative effect while social networks (proxied by migrant stocks) are expected to increase migration flows by reducing psychological costs, facilitating integration into the host society, and increasing the possibility of obtaining a job. Having a colonial tie or cultural attachments (common official language and a common source of legislation) is also expected to reduce migration costs and, hence, increase migration flows.

Table 2 presents the baseline results on the determinants of the regular migration flows from Africa to Europe. We consider the economic and political indicators of human security together with indicators for the cost of migration that are discussed in Data section as explanatory variables. In all the specifications, source-country fixed effects $(α_{s})$ are included. Additionally, we add destination-country fixed effects $(α_{d})$ in columns 1 and 2, year fixed effects $(α_{t})$ in column 2, and time-variant fixed effects of the destination country $(α_{d t})$ in columns 3 to 8.

Table 2.

Determinants of African Migration Flows to Europe.

Dependent Variable Migration Flows (in Level) Specification	(1) PPML	(2) PPML	(3) PPML	(4) PPML	(5) PPML	(6) PPML	(7) PPML	(8) PPML
Estimation Method	(1) PPML	(2) PPML	(3) PPML	(4) PPML	(5) PPML	(6) PPML	(7) PPML	(8) PPML
Log GDP per capita destination _t−1	0.889***	1.707***
Log GDP per capita destination _t−1	(0.156)	(0.352)
Log GDP per capita source _t−1	−0.394***	−0.214***	−0.140***	−0.140***	−0.136***	−0.122*	−0.106*	−0.206*
Log GDP per capita source _t−1	(0.0655)	(0.0586)	(0.0510)	(0.0510)	(0.0520)	(0.0635)	(0.0560)	(0.116)
Civil conflict source _t−1	0.238***	0.290***	0.288***	0.288***	0.291***	0.241***	0.322***	0.305***
Civil conflict source _t−1	(0.0422)	(0.0356)	(0.0397)	(0.0397)	(0.0423)	(0.0482)	(0.0517)	(0.0831)
Autocracy source _t−1	0.0396	0.140***	0.0630*	0.0630*	0.0628*	0.115**	0.0943	0.0273
Autocracy source _t−1	(0.0633)	(0.0487)	(0.0349)	(0.0349)	(0.0348)	(0.0491)	(0.0703)	(0.0732)
Democracy source _t−1	−3.198***	−2.860***	−2.278***	−0.838***	−1.108*	−1.598***	−1.979***	−2.006***
Democracy source _t−1	(0.243)	(0.253)	(0.131)	(0.179)	(0.663)	(0.221)	(0.382)	(0.470)
Political rights source _t−1	−0.0390	−0.111*	−0.0967	−0.0967	−0.0944	−0.391***	−0.292***	−0.139
Political rights source _t−1	(0.0895)	(0.0649)	(0.0785)	(0.0785)	(0.0756)	(0.0993)	(0.0996)	(0.121)
Civil liberties source _t−1	−0.105	−0.0765*	−0.0610	−0.0610	−0.0626	−0.0220	0.000576	0.0147
Civil liberties source _t−1	(0.0730)	(0.0458)	(0.0401)	(0.0401)	(0.0403)	(0.0687)	(0.0540)	(0.0920)
Ethnic polarization source	3.229***	2.869***	2.391***		6.085***	1.366***	2.421***	2.404***
Ethnic polarization source	(0.203)	(0.276)	(0.230)		(0.897)	(0.364)	(0.791)	(0.748)
Ethnic fractionalization source				3.363***
Ethnic fractionalization source				(0.224)
Log urban population source _t−1					−0.0704
Log urban population source _t−1					(0.210)
Landlocked source					0.631***
Landlocked source					(0.220)
Network _t−1	0.0479***	0.0411***	0.0725***	0.0725***	0.0726***	0.128***	0.0946***	0.0697*
Network _t−1	(0.00811)	(0.00859)	(0.0124)	(0.0124)	(0.0127)	(0.0181)	(0.00985)	(0.0365)
Log distance	−0.634***	−0.620***	−0.625***	−0.625***	−0.625***	−0.751***	−0.316***	−0.755***
Log distance	(0.0549)	(0.0534)	(0.0569)	(0.0569)	(0.0568)	(0.0770)	(0.0548)	(0.171)
Common legislation	0.902***	0.910***	0.855***	0.855***	0.854***	1.184***	0.924***	0.614***
Common legislation	(0.0651)	(0.0672)	(0.0668)	(0.0668)	(0.0663)	(0.114)	(0.0817)	(0.179)
Common language	0.697***	0.730***	0.512***	0.512***	0.513***	0.983***	0.837***	0.663***
Common language	(0.0501)	(0.0519)	(0.0613)	(0.0613)	(0.0599)	(0.0478)	(0.0579)	(0.133)
Colonial ties	1.099***	1.115***	1.045***	1.045***	1.044***	0.575***	0.620***	0.952***
Colonial ties	(0.0432)	(0.0481)	(0.0504)	(0.0504)	(0.0526)	(0.0672)	(0.0640)	(0.139)
Constant	2.852**	−4.746*	3.617***	2.277***	8.738***	9.791***	0.0558	−12.49***
Constant	(1.183)	(2.626)	(0.510)	(0.727)	(2.688)	(0.722)	(0.868)	(0.845)
Observations	22,635	22,635	20,790	20,790	20,790	18,630	18,858	3,555
R ²	.709	.755	.846	.846	.846	.913	.902	.864
Fixed effects
$Destination (α_{d})$	Yes	Yes	No	No	No	No	No	No
${Source(α}_{s})$	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
$Destination - year (α_{d t})$	No	No	Yes	Yes	Yes	Yes	Yes	Yes
$Year (α_{t})$	No	Yes	No	No	No	No	No	No
Sample	All	All	All	All	All	Non − (UK + France)	SSA	Five years

Note: Results are obtained by using the PPML estimation method. The estimation period is 1990 to 2014. Column 8 uses only years 1990, 1995, 2000, 2005, and 2010. Column 6 omits France and the United Kingdom and Column 7 considers the sub-Saharan Africa subsample. Standard errors in parenthesis are heteroskedasticity robust and clustered by year. Significance at the 1 percent, 5 percent, and 10 percent level is indicated by ***, **, and *, respectively. GDP = gross domestic product; PPML = Poisson pseudomaximum likelihood; SSA = sub-Saharian African.

Generally, the coefficients for all explanatory variables are statistically significant and carry the expected signs in most of the specifications. Higher per capita income at the destination countries and civil conflicts, state autocracy, and ethnic tension at the source countries lead to higher migration flows, as expected. The control variables such as migrant networks, shared legal roots, common official language, and colonial legacy positively impact on migration flows. Conversely, higher GDPPC, democracy, political rights, civil liberties, and the landlockedness of the source country decrease migration flows. Moreover, the larger the distance between the source and the destination countries, the lower is the bilateral migration flows.

To put results into the context of the estimated elasticities, we analyze the variables across specific models. In Table 2, while column 1 controls for fixed effects of source countries $(α_{s})$ and destination countries $(α_{d})$ only, column 2 includes time effects ${(α}_{t})$ as well. The results show that a 10 percent increase in the average income earned per person at the destination countries leads to an increase of about 8.9 and 17.1 percent, respectively, in annual bilateral migration flows. Whereas, from the same columns, a 10 percent increase in per capita income at the source countries is associated with a 3.9 and 2.1 percent decrease in dyadic migration flows (from source to destination countries), respectively. The results for the per capita income are consistent with much of the literature (e.g., Bertoli, Brücker, and Moraga 2013; Ortega and Peri 2013). Hence, these results suggest that higher average income levels at the destination countries have strong pulling power on the migration flows compared to the potential reduction effects of income growth at the source countries. A possible explanation could be that the income levels at the source countries are substantially lower in absolute terms than at the destination countries. With respect to economic development in the source countries, the literature depicts controversial results between income at the origin countries and emigration. Clemens, Özden, and Rapoport (2014) summarize that across the literature, cross-sectional studies generally find either a positive or inverted-U relationship between income and emigration from low-income countries, while time-series studies do not find a consistent relationship between income and emigration.¹⁴ In accordance with the policy makers’ conventional wisdom (Clemens, Özden, and Rapoport 2014), our findings confirm that an increase in average income per capita in the origin countries is negatively correlated with migration flows, which coincide with several studies that use time-series macro studies and control for country fixed effects (see, for instance, Carlos 2002; Hatton and Williamson 2003; Clark, Hatton, and Williamson 2004, 2007; Bertoli, Brücker, and Moraga 2013; Ortega and Peri 2013).

Since our main emphasis is on the human security factors at the source countries, we use appropriate fixed effects (i.e., time-invariant source-countries ${(α}_{s})$ and time-variant destination-countries $(α_{d t})$ ), as provided by column 3. Specifically, the results show that an increase in per capita income at the source countries leads also to a decrease in migration flows in this specification. Concerning political determinants of migration, most of the variables exert statistically significant results with the expected signs. In particular, the incidences of violent civil conflicts at the source countries are a significant pushing factor, which is in accordance with Naudé (2010), who finds that violent civil conflicts are important drivers of net migration from sub-Saharan Africa.

An increase in the government’s autocracy in the source countries, which is an indicator of the presence of political persecution or human rights violations, leads to an increase in international migration flows. Democratization processes at the source countries, on the contrary, have a strong reducing effect on bilateral migration flows. Specifically, the size of the impact of the democratization dummy variable is particularly large: using coefficients from the standardized variables, it is apparent that democratization has the largest impact of all the determinants of bilateral migration flows. Furthermore, measures of political and civil rights at the source countries, which proxy political freedom, often carry the expected negative signs, although estimated coefficients are rarely statistically significant. The results on the crucial roles of political institutions (democracy, autocracy, political rights, and civil liberties) on the flow of African migration to Europe are consistent with most of the literature on international migration and institutions (Baudassé et al. 2018).¹⁵

The other important noneconomic determinants of migration flows are ethnic polarization and fractionalization, which measure the level of ethnic diversity and tensions at the source countries. In this regard, increases in these social heterogeneity factors are found to strongly increase migration outflows.

The remaining variables, which mainly proxy migration costs, are also significant and carry the expected signs. Specifically, migrant networks have a strong positive impact on bilateral migration flows. Moreover, those African countries that are farther away from Europe have fewer migrants heading to Europe due to distance-related costs. On the contrary, variables of colonial ties, common legal roots, common official language, and other cultural attachments increase migration flows.

The rest of the columns of Table 2 contain robustness check results. In column 4, we employ ethnic fractionalization instead of ethnic polarization and find a qualitatively similar pushing effect of ethnic tensions on migration flows. In column 5, we add urban population and landlockedness of the source countries to investigate their potential effects on migration flows. The results show that having a large urban population has a negative but statistically insignificant effect on bilateral migration flows. The negative sign is consistent with the fact that large urban population represents socioeconomic improvements. However, the landlockedness of the source countries decreases migration flows, as expected. The effects of the rest of the explanatory variables remain qualitatively unaffected. In column 6, we estimate determinants of bilateral migration flows omitting the two main former colonial powers in Africa (the United Kingdom and France). Once more, results on the main and control variables remain robust and highly significant. Historically, due to geographical proximity, Europe is the main destination for North African migrants (Zlotnik 1991; Flahaux and De Haas 2016). However, in the last two decades, there has been a surge in the number of migrants from sub-Saharan Africa toward Europe. Accordingly, in column 7, we estimate a subsample of SSA migration trends by omitting the five North African countries from the full sample. The results show that the baseline estimation is still robust despite the fact that the number of observations is reduced by 9 percent. Finally, column 8 presents estimation results of a subsample where five-years period are considered (1990–1994, 1995–1999, 2000–2004, 2005–2009, and 2010–2014). These results show that while an increase in income per capita at the source is still a relevant factor for reducing international migration flows, violent civil conflicts and ethnic tensions at the source are also important pushing factors for African migrants. Notwithstanding, improvements of political rights at the source substantially decrease the rate of bilateral migration flows in the five-year intervals. These results confirm the robustness of our baseline findings.

In summary, our baseline results reveal that human security determinants are important factors in shaping the South–North migration trends in the past few decades. The estimation results suggest that African migration patterns toward Europe are significantly influenced by economic, political, social, and cultural conditions. In particular, African extracontinental migrations are caused by poverty, civil wars, ethnic tensions, and civil and political rights violations. The results also indicate that improvements in per capita income and political conditions at the source countries are negatively related to the rate of migration flows. Furthermore, social, cultural, geopolitical, and historical ties with European countries have a significant impact in influencing the trend in African migration toward Europe.

Determinants of the Flows of African Asylum Seekers to Europe

Each year thousands of migrants from Africa enter Europe after braving the perils of crossing the Mediterranean Sea using inadequate transport conditions. Although several factors could be listed as reasons, the EU’s tightened entry policies for African migrants, on the one hand, and lack of financial means and appropriate travel documents by the migrants, on the other hand, are thought to have forced African immigrants to choose the irregular pathways (Hansen and Jonsson 2011; Flahaux and De Haas 2016). To examine the extent to which human security factors determine the flow of African asylum seekers to Europe, we reestimate equation (3) using data on the flow of asylum seekers.

The estimation results of the determinants of the flows of asylum seekers from Africa to Europe are reported in Table 3. While we still use the specifications used in Table 2, the dependent variable here is the number of annual asylum seekers, which represents the flows of asylum seekers. The results obtained by using asylum seeker data are qualitatively similar to the baseline results using regular migration flow data. Specifically, an increase in the GDPPC of the destination countries is associated with an increase in the number of asylum seekers. As in Table 2, throughout the specifications, an increase in the GDPPC of the source countries leads to a decrease in the number of asylum seekers. Moreover, political turmoil, fear of being persecuted for reasons of ethnicity, or political opinion are found to drive African migrants to demand a refugee status in Europe. As expected, the source countries’ democratization leads to a decrease in bilateral flows of asylum seekers. Throughout the specifications, the institutional quality variable has a statistically significant impact with the expected sign. Furthermore, migration cost and culture-related control variables remain statistically significant with the expected signs. Unlike in Table 2, however, a large urban population has a significantly negative effect on bilateral flows of asylum seekers. It likely reflects the negative effect of socioeconomic improvements associated with urbanization in reducing the flows of asylum seekers from Africa to Europe.

Table 3.

Determinants of the Flows of African Asylum Seekers to Europe.

Asylum Seekers (in Level) Specification	(1) PPML	(2) PPML	(3) PPML	(4) PPML	(5) PPML	(6) PPML	(7) PPML	(8) PPML
Estimation Method	(1) PPML	(2) PPML	(3) PPML	(4) PPML	(5) PPML	(6) PPML	(7) PPML	(8) PPML
Log GDP per capita destination _t−1	0.286	1.003***
Log GDP per capita destination _t−1	(0.191)	(0.321)
Log GDP per capita source _t−1	−0.229***	−0.173***	−0.177***	−0.177***	−0.140**	−0.230***	−0.179***	−0.509***
Log GDP per capita source _t−1	(0.0764)	(0.0640)	(0.0652)	(0.0652)	(0.0664)	(0.0752)	(0.0677)	(0.0990)
Civil conflict source _t−1	0.351***	0.400***	0.393***	0.393***	0.403***	0.497***	0.406***	0.470***
Civil conflict source _t−1	(0.0724)	(0.0550)	(0.0594)	(0.0594)	(0.0586)	(0.0508)	(0.0608)	(0.132)
Autocracy source _t−1	−0.410***	−0.161	−0.111	−0.111	−0.108	−0.0702	−0.0879	−0.147
Autocracy source _t−1	(0.126)	(0.149)	(0.150)	(0.150)	(0.148)	(0.163)	(0.165)	(0.217)
Democracy source _t−1	−3.821***	−3.060***	−3.761***	−1.209***	−6.897***	−2.281***	−3.595***	−2.348***
Democracy source _t−1	(0.407)	(0.236)	(0.330)	(0.186)	(0.777)	(0.303)	(0.246)	(0.513)
Political rights source _t−1	−0.0456	−0.196*	−0.172	−0.172	−0.178	−0.129	−0.195*	−0.151
Political rights source _t−1	(0.105)	(0.109)	(0.108)	(0.108)	(0.108)	(0.123)	(0.107)	(0.175)
Civil liberties source _t−1	−0.461***	−0.357***	−0.347***	−0.347***	−0.337***	−0.359***	−0.343***	−0.322*
Civil liberties source _t−1	(0.0883)	(0.0712)	(0.0775)	(0.0775)	(0.0785)	(0.0904)	(0.0710)	(0.174)
Ethnic polarization source	3.887***	3.945***	3.725***		7.453***	2.929***	3.996***	4.101***
Ethnic polarization source	(0.501)	(0.450)	(0.471)		(1.263)	(0.517)	(0.676)	(1.083)
Ethnic fractionalization source				5.667***
Ethnic fractionalization source				(0.484)
Network _t−1	0.0332***	0.0238***	0.0613***	0.0613***	0.0622***	0.0985***	0.0762***	0.0539***
Network _t−1	(0.00441)	(0.00369)	(0.00736)	(0.00736)	(0.00776)	(0.00443)	(0.00752)	(0.0153)
Log urban population source _t−1					−0.937***
Log urban population source _t−1					(0.215)
Landlocked source					−0.498*
Landlocked source					(0.299)
Log distance	−0.411**	−0.412**	−0.397**	−0.397**	−0.390**	−0.562**	0.0378	−0.451
Log distance	(0.179)	(0.181)	(0.172)	(0.172)	(0.172)	(0.243)	(0.226)	(0.388)
Common legislation	0.337***	0.352***	0.311***	0.311***	0.311***	0.167**	0.401***	0.369***
Common legislation	(0.0381)	(0.0367)	(0.0531)	(0.0531)	(0.0526)	(0.0762)	(0.0602)	(0.112)
Common language	0.714***	0.743***	0.591***	0.591***	0.584***	0.890***	0.608***	0.710***
Common language	(0.0770)	(0.0715)	(0.0679)	(0.0679)	(0.0663)	(0.0637)	(0.0696)	(0.138)
Colonial	0.240***	0.277***	0.134	0.134	0.138	−0.0928	−0.0429	0.183
Colonial	(0.0816)	(0.0803)	(0.0936)	(0.0936)	(0.0932)	(0.174)	(0.0961)	(0.211)
Colonial ties	5.249***	−0.884	7.097***	4.665***	19.51***	3.650*	2.346	6.027*
Colonial ties	(1.979)	(2.419)	(1.333)	(1.569)	(2.994)	(1.934)	(1.827)	(3.282)
Observations	22,635	22,635	21,105	21,105	21,105	18,945	19,698	3,555
R ²	.283	.315	.468	.468	.467	.565	.500	.586
Fixed effects
${Destination(α}_{d})$	Yes	Yes	No	No	No	No	No	No
${Source(α}_{s})$	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
$Destination - {year(α}_{d t})$	No	No	Yes	Yes	Yes	Yes	Yes	Yes
${Year(α}_{t})$ )	No	Yes	No	No	No	No	No	No
Sample	All	All	All	All	All	Non − (UK + France)	SSA	Five years

Note: The estimation period is 1990 to 2014. Column 5 includes the post-1995 subsample. Column 7 uses only years 1990, 1995, 2000, 2005, and 2010. Column 5 omits France and the United Kingdom and Column 6 considers SSA subsample. Standard errors in parenthesis are heteroskedasticity robust and clustered by year. Significance at the 1 percent, 5 percent, and 10 percent level is indicated by ***, **, and *, respectively. PPML = Poisson pseudomaximum likelihood; SSA = sub-Saharian African.

The Role of Political Determinants on Migration Flows

As shown by Tables 2 and 3, African migration to Europe is not only driven by economic factors but also by noneconomic motivations such as civil conflicts, human rights violations, and political stability. In the following, we emphasize the role of political factors in extracontinental migration flows. In doing so, Table 4 reports the results from regressions that include broader determinants of international (regular and asylum seeker) migration flows between Africa and Europe.¹⁶ Specifically, we consider measures of political instability and specific measures of political risks at the source countries, in addition to the aforementioned political and distributional indexes in Tables 2 and 3. In order to facilitate the comparison, column 1 of Table 4 replicates column 3 of the baseline specification in Tables 2. Column 2 of Table 4 estimates the impacts of political variables focusing on political stability and institutions as the key determinants of international migration. It is noteworthy that in this specification, economic determinants are omitted and the coefficients remain with the expected sign, except for institutionalized democracy. Once more, civil conflict, institutionalized autocracy, and ethnic polarization are positively correlated with migration flows. In column 3, we consider alternative and disaggregated measures of both economic and political institutions, which are taken from the ICRG (Howell 2011). The set of indexes that we use in this analysis includes socioeconomic conditions, investment profile, democratic accountability, government stability, control of corruption, law and order, bureaucractic quality, external conflict, and religious tension of the source countries. With the exception of external conflict and religious tension, the rest of the institutional quality indicators are supposed to have a decreasing effect on the migration outflows.

Table 4.

Political Factors as Determinants of Migration Flows.

Estimation Method	(1)	(2)	(3)	(4)	(5)	(6)
Estimation Method	PPML	PPML	PPML	PPML	PPML	PPML
Dependent Variables (in Level)	M. Flows	M. Flows	M. Flows	A. Seekers	A. Seekers	A. Seekers
Log GDP per capita source _t−1	−0.140***			−0.177***
Log GDP per capita source _t−1	(0.0510)			(0.0652)
Civil conflict source _t−1	0.288***		0.280***	0.393***		0.272***
Civil conflict source _t−1	(0.0397)		(0.0386)	(0.0594)		(0.0696)
Autocracy source _t−1	0.0630*		0.0731**	−0.113		−0.234
Autocracy source _t−1	(0.0349)		(0.0345)	(0.150)		(0.166)
Democracy source _t−1	−2.278***			−3.700***
Democracy source _t−1	(0.131)			(0.334)
Political rights source _t−1	−0.0967		−0.154**	−0.175		−0.203***
Political rights source _t−1	(0.0785)		(0.0688)	(0.108)		(0.0784)
Civil liberties source _t−1	−0.0610		0.00905	−0.348***		−0.199**
Civil liberties source _t−1	(0.0401)		(0.0491)	(0.0773)		(0.0825)
Ethnic polarization source	2.391***		0.723***	3.882***		5.096***
Ethnic polarization source	(0.230)		(0.247)	(0.464)		(0.408)
Socioeconomic conditions source _t−1		−0.0398*	−0.0442**		0.0477	0.0294
Socioeconomic conditions source _t−1		(0.0209)	(0.0183)		(0.0427)	(0.0419)
Investment profile source _t−1		0.0135	0.0102		−0.0501**	−0.0460**
Investment profile source _t−1		(0.0110)	(0.0117)		(0.0241)	(0.0233)
Democratic accountability _t−1		−0.0654***	−0.0737***		−0.103***	−0.0630**
Democratic accountability _t−1		(0.0151)	(0.0130)		(0.0298)	(0.0287)
Government stability _t−1		0.0211	0.0331		−0.0627	−0.0588
Government stability _t−1		(0.0232)	(0.0264)		(0.0480)	(0.0505)
Control of corruption _t−1		0.0126	0.00195		0.0423	0.0611
Control of corruption _t−1		(0.0230)	(0.0276)		(0.0711)	(0.0715)
Law and order source _t−1		−0.0290	−0.0414		−0.0808	−0.0631
Law and order source _t−1		(0.0537)	(0.0459)		(0.0755)	(0.0611)
Bureaucracy quality source _t−1		−0.0442	−0.0326		−0.0608	−0.00706
Bureaucracy quality source _t−1		(0.0499)	(0.0531)		(0.0627)	(0.0686)
External conflict _t−1		−0.0310	−0.0127		−0.0868**	−0.0623*
External conflict _t−1		(0.0217)	(0.0188)		(0.0352)	(0.0325)
Religious tension _t−1		0.0383	0.0301		−0.0640*	−0.0804**
Religious tension _t−1		(0.0272)	(0.0255)		(0.0364)	(0.0378)
Network _t−1	0.0725***	0.0702***	0.0679***	0.0619***	0.0626***	0.0620***
Network _t−1	(0.0124)	(0.0122)	(0.0120)	(0.00723)	(0.00736)	(0.00725)
Log distance	−0.625***	−0.542***	−0.494***	−0.390**	−0.271	−0.316*
Log distance	(0.0569)	(0.0529)	(0.0516)	(0.171)	(0.174)	(0.177)
Common legislation	0.855***	0.985***	1.150***	0.336***	0.207***	0.239***
Common legislation	(0.0668)	(0.0693)	(0.0622)	(0.0534)	(0.0506)	(0.0515)
Common language	0.512***	0.276***	0.266***	0.636***	0.615***	0.595***
Common language	(0.0613)	(0.0480)	(0.0472)	(0.0740)	(0.0740)	(0.0762)
Colonial ties	1.045***	1.073***	1.234***	0.408***	0.207	0.230*
Colonial ties	(0.0504)	(0.0715)	(0.0675)	(0.0582)	(0.133)	(0.124)
Constant	3.617***	9.408***	8.847***	6.946***	8.448***	5.874***
Constant	(0.510)	(0.620)	(0.662)	(1.321)	(1.381)	(1.508)
Observations	20,790	16,704	15,708	21,105	16,884	15,946
R ²	.846	.845	.856	.470	.486	.489
Fixed effects
${Destination(α}_{d})$	Yes	Yes	Yes	Yes	Yes	Yes
$Destination - year (α_{d t})$	Yes	Yes	Yes	Yes	Yes	Yes
Sample	All	All	All	All	All	All

Note: Standard errors in parenthesis are heteroskedasticity robust and clustered by year. Significance at the 1 percent, 5 percent, and 10 percent level is indicated by ***, **, and *, respectively.

In a similar way to increases in GDPPC, improvements in socioeconomic conditions at the source countries have a negative impact on migration flows. The result implies that better economic institutions at the origin have a reducing effect on migration outflows. Among the measures of the quality of political institutions, democratic accountability significantly determines migration outflows. This result is consistent with the effect of the institutionalized democracy indicator that is used in our baseline results (see column 1). Although some of the variables carry the expected signs, across specifications, the impacts of most of the remaining economic and political institutions variables are not statistically significant.¹⁷ Column 3 combines variables used in columns 1 and 2, but the democracy variable of column 1 is omitted to avoid redundancy with democratic accountability. The results remain generally similar to those of columns 1 and 2.

The results in columns 4 to 6 of Table 4 are obtained from the same specifications as results in columns 1 to 3, but using the flow of asylum seekers instead of regular migration flows as the dependent variable. The results on the impact of political determinants on regular migration flows remain valid for the flow of asylum seekers.

Robustness Checks

Given that our estimation model is a pseudogravity model, we use the PPML estimator from Tables 2 to 4, which flexibly accounts for a significant proportion of zero observations in the dependent variables. In this section, we check for the robustness of the benchmark results by performing OLS estimation on equation (2) using the positive migration flows only, as in Ortega and Peri (2013). These results are documented in Tables D.1 and D.2 of Online Appendix D. For each specification, both the source-country fixed effects $(α_{s})$ and time-destination fixed effects $(α_{d t})$ are included.

Online Appendix Tables D.1 and D.2 report, respectively, the estimation results obtained when zero bilateral regular and asylum seeker migration flows are omitted. In both cases, GDPPC at the source and the destination countries, civil conflict, institutional autocracy and ethnic distributional indexes, political stability (both at the source and the destination countries), socioeconomic conditions, law and order, bureaucracy quality as well as the control variables display effects that are qualitatively similar to our baseline results in Tables 2 and 3. Furthermore, in the linear estimation, the civil liberty parameter of the source country significantly affects migration flows.

Another robustness check involves using the PPML estimator as in Tables 2 and 3 but reestimating the models by gradually increasing the number of political determinants of migration. This is meant to check whether our main estimation results are affected by the fairly large number of explanatory variables we considered. Respective results are documented in Online Appendix Tables E.1 and E.2. In general, the results show a remarkable degree of robustness to progressively including political drivers of migration. For instance, GDPPC of the source country has a significantly negative impact on migration flows from Africa to Europe although the respective coefficient decreases in absolute terms from −0.185 in the most parsimonious specification to −1.40 in the most comprehensive model in Online Appendix Table E.1. Similarly, the incidence of civil conflicts has a robust positive effect on migration flows, and the magnitude of its impact remains largely unaffected by the inclusion of additional explanatory variables. The same can be said of the effects of democracy score, distance, common legislation, common language, colonial ties, and migrant networks. The only exception is the measure of civil liberties at the source country, which always carry a negative coefficient, but becomes statistically significant in only one of the specifications in Online Appendix Table E.1. Still the indicator for civil liberties in African countries has a significantly negative effect on the flow of asylum seekers as documented in Online Appendix Table E.2.

Conclusions

The current migration and refugee crisis in Europe requires an understanding of the different migration drivers beyond the well-known economic determinants. While a few extant studies have examined determinants of African migration flows, they have however either focused exclusively on intra-African migration flows or studied international migration from Africa together with intra-African flows. The present article aims at filling this gap in the literature by providing a thorough empirical examination of the role played by human security factors in explaining trends in African migration to Europe during the period 1990 to 2014. To estimate the pseudogravity model of bilateral migration flows, we employ the PPML estimator, which is particularly suitable in regressions where the dependent variable has a significant proportion of zero values (Beine and Parsons 2015). This article contributes to the literature on the determinants of international migration flows in two ways.

First, as aforementioned, we empirically analyze the specific Africa to Europe trends of migration flows. From the perspective of European destination countries, African migration flows show a dramatic rise and curbing this flow is among the top policy issues. From the view of African source countries, Europe continues to be an important destination for extracontinental African migrants, thanks to the historical and geopolitical ties between the two continents. Hence, this emphasis helps us to highlight essential determinants of migration flows from Africa to Europe.

Second, although the importance of noneconomic factors such as civil wars and conflicts in international migration is well-documented in the literature (Naudé 2010; Docquier, Ruyssen, and Schiff 2018), the present article additionally highlights the roles of institutions and social heterogeneity indicators such as ethnic polarization and religious tension in the decisions of individuals to emigrate. Extracontinental migration is not only costly but it is also risky. Hence, despite abject poverty, wars, civil conflicts, severe political persecution, and human rights abuses, the majority of the world population stay at home or move mainly to neighboring countries in their search for safety and protection. Our analysis underscores that, in addition to the obvious economic determinants and violent conflicts, a combination of several push and pull factors including political conditions (ongoing violence and instability, low institutional quality) and preexisting sociocultural structures influence the migration choice of individuals.

We find several notable results. First, most of the human security indicators significantly determine annual migration flows from Africa into European countries. Per capita income growth at a given European destination is associated with an increase in immigrant flows while per capita income growth at a given source country is negatively related with emigration from Africa. Rising political persecution, human right violations, political instability, and civil conflicts in source countries are also associated with increased migration flows into European destination countries. Second, in conjunction with the regular trends of migration flows, asylum seekers from Africa also have a combination of political and economic motivations to claim refugee status. Hence, categorizing African immigrants as the bogus asylum seeker in general terms would be highly misleading and could result in misguided migration policies. Third, cultural and migration cost-related factors, such as migrant networks, colonial ties, common languages, physical distance, and living in a landlocked country have shown significant effects on the trends in African migration to Europe.

The aforementioned findings have significant policy implications for managing the recent migration and refugee crisis in Europe. The African migration flows to Europe are complex and driven by mixed pushing and pulling factors. The central point of this discussion is acknowledging the heterogeneity of the flows, since a valid response will need to be grounded in a sound understanding of fundamental causes of the flows. Further, the African migrants’ motives, patterns, and trends should be seen from the broader human security point of view. Overlooking the political factors, which significantly influence international migration, and attempting to address only the economic causes through investing in Africa, may have counterproductive consequences. Therefore, the collaboration among the African source countries and all the European countries and institutions should be based not only on the economic development in the source countries but also on the promotion of human security: peace, human rights, democracy, and social stability.

Finally, it is noteworthy that, although our main result—that broader human security factors drive international migration—is based on data on trends in African migration to Europe in the post–Cold War period, it appears to be generalizable to the rest of the world for three main reasons. First, there are already studies on other parts of the world that have found violent conflicts as a significant driver of emigration (Clemens 2017; Shrestha 2017). Second, the frequency and severity of violent conflicts and human rights abuses in some parts of the world, especially the Middle East, are as high as, if not higher than, in Africa. This also makes human security factors potentially important drivers of emigration from other conflict-ridden regions such as the Middle East. Third, although the post–Cold War period is a period when Africa has witnessed some of the most violent conflicts in the world (e.g., Rwanda, Congo, Liberia), this is at the same time a period during which several African countries experienced peaceful transfers of power (e.g., Liberia, Ghana, Nigeria, Tanzania). Hence, our results may be also applied to politically stable emerging economies in Asia and Latin America.

Supplemental Material

Supplemental Material, Online_Appendix_JCR - Trends in African Migration to Europe: Drivers Beyond Economic Motivations

Supplemental Material, Online_Appendix_JCR for Trends in African Migration to Europe: Drivers Beyond Economic Motivations by José-Manuel Giménez-Gómez, Yabibal M. Walle and Yitagesu Zewdu Zergawu in Journal of Conflict Resolution

Supplemental Material

Supplemental Material, Replicationdat_dofile - Trends in African Migration to Europe: Drivers Beyond Economic Motivations

Supplemental Material, Replicationdat_dofile for Trends in African Migration to Europe: Drivers Beyond Economic Motivations by José-Manuel Giménez-Gómez, Yabibal M. Walle and Yitagesu Zewdu Zergawu in Journal of Conflict Resolution

Supplemental Material

Supplemetal_Material - Trends in African Migration to Europe: Drivers Beyond Economic Motivations

Supplemetal_Material for Trends in African Migration to Europe: Drivers Beyond Economic Motivations by José-Manuel Giménez-Gómez, Yabibal M. Walle and Yitagesu Zewdu Zergawu in Journal of Conflict Resolution

Footnotes

Acknowledgments

We are particularly grateful to two anonymous referees and the editor of this journal. We also thanks to Yuliya Lovcha and Carles Méndez-Ortega their comments and support.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Ministerio de Economía y Competitividad (ECO2016-75410-P (AEI/FEDER, UE) and Universitat Rovira i Virgili and Generalitat de Catalunya (2017PFR-URV-B2-53 and 2017 SGR 770).

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Anand

1994. “Human Development Index: Methodology and Measurement.” Technical Report. Human Development Report Office (HDRO), United Nations Development Programme (UNDP). https://ora.ox.ac.uk/objects/uuid:98d15918-dca9-4df1-8653-60df6d0289dd/download_file?file_format=application/pdf&safe_filename=HDI_methodology.pdf&type_of_work=Report.

Bang

J. T.

Mitra

. 2013. “Civil War, Ethnicity, and the Migration of Skilled Labor.” Eastern Economic Journal 39:387–401.

Banulescu-Bogdan

Fratzke

. 2015. Europes Migration Crisis in Context: Why Now and What Next. Migration Policy Institute 24. https://www.migrationpolicy.org/article/europe%E2%80%99s-migration-crisis-context-why-now-and-what-next.

Baudassé

Bazillier

Issifou

. 2018. “Migration and Institutions: Exit and Voice (from Abroad)?” Journal of Economic Surveys 32:727–66.

Beine

Bertoli

Fernández-Huertas Moraga

. 2016. “A Practitioners Guide to Gravity Models of International Migration.” The World Economy 39:496–512.

Beine

Coulombe

Vermeulen

W. N.

. 2015. “Dutch Disease and the Mitigation Effect of Migration: Evidence from Canadian Provinces.” The Economic Journal 125:1574–615.

Beine

Docquier

Özden

Ç.

. 2011. “Diasporas.” Journal of Development Economics 95:30–41.

Beine

Parsons

. 2015. “Climatic Factors as Determinants of International Migration.” The Scandinavian Journal of Economics 117:723–67.

Belloc

2015. “International Economic Assistance and Migration: The Case of Sub-Saharan Countries.” International Migration 53:187–201.

10.

Berman

B. J.

1998. “Ethnicity, Patronage and the African State: The Politics of Uncivil Nationalism.” African Affairs 97:305–41.

11.

Bertoli

Brücker

Moraga

J. F. H.

. 2013. “The European Crisis and Migration to Germany: Expectations and the Diversion of Migration Flows.” IZA Discussion Paper No.7170. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2210830.

12.

Bonifazi

Heins

Strozza

Vitiello

. 2009. The Italian Transition from an Emigration to Immigration country, Citizens and Governance in Knowledge-based Society. Italy, Europe: EU.

13.

Carlos

R. D.

2002. “On the Determinants of International Migration in the Philippines: An Empirical Analysis.” International Migration Review 36:81–102.

14.

Castillejo

2017. “The eu Migration Partnership Framework: Time for a Rethink?” German Development Institute Discussion Paper. https://www.die-gdi.de/discussion-paper/article/the-eu-migration-partnership-framework-time-for-a-rethink/.

15.

Castles

2004. “Why Migration Policies Fail?” Ethnic and Racial Studies 27:205–27.

16.

Clark

Hatton

T. J.

Williamson

J. G.

. 2004. “What Explains Emigration Out of Latin America?” World Development 32:1871–90.

17.

Clark

Hatton

T. J.

Williamson

J. G.

. 2007. “Explaining US Immigration, 1971–1998.” The Review of Economics and Statistics 89:359–73.

18.

Clemens

M. A.

2017. “Violence, Development and Migration Waves: Evidence from Central American Child Migrant Apprehensions.” IZA Discussion Paper No. 10928. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3029798.

19.

Clemens

M. A.

Özden

Ç.

Rapoport

. 2014. “Migration and Development Research Is Moving Far Beyond Remittances.” World Development 64:121–24.

20.

Constant

A. F.

Zimmermann

K. F.

. 2011. “Migration, Ethnicity and Economic Integration.” In International Handbook of Economic Integration 145–68. https://books.google.es/books?hl=ca&lr=&id=APAYeZRoJNgC&oi=fnd&pg=PA145&dq=Constant,+A.+F.,+and+K.+F.+Zimmermann.+2011.+%E2%80%9CMigration,+Ethnicity+and+Economic+Integration.%E2%80%9D+In+International+Handbook+of+Economic+Integration+145-68&ots=nTfr3O_PJi&sig=9PJQtknBnclNkYwO52NUroFND40#v=onepage&q=Constant%2C%20A.%20F.%2C%20and%20K.%20F.%20Zimmermann.%202011.%20%E2%80%9CMigration%2C%20Ethnicity%20and%20Economic%20Integration.%E2%80%9D%20In%20International%20Handbook%20of%20Economic%20Integration%20145-68&f=false.

21.

Czaika

2012. “Internal versus International Migration and the Role of Multiple Deprivation: Some Evidence from India.” Asian Population Studies 8:125–49.

22.

Daley

2006. “Ethnicity and Political Violence in Africa: The Challenge to the Burundi State.” Political Geography 25:657–79.

23.

Dao

T. H.

Docquier

Parsons

Peri

. 2018. “Migration and Development: Dissecting the Anatomy of the Mobility Transition.” Journal of Development Economics 132:88–101.

24.

De Haas

2007. “Turning the Tide? Why Development Will Not Stop Migration.” Development and Change 38:819–41.

25.

De Haas

2011. “The Determinants of International Migration.” IMI Working Paper 24. https://www.imi.ox.ac.uk/publications/wp-32-11.

26.

Docquier

Ruyssen

Schiff

M. W.

. 2018. “International Migration: Pacifier or Trigger for Military Conflicts?” The Journal of Development Studies 54:1657–79.

27.

Dunn

K. C.

2009. “Sons of the Soiland Contemporary State Making: Autochthony, Uncertainty and Political Violence in Africa.” Third World Quarterly 30:113–27.

28.

Dutta

Roy

. 2011. “Do Potential Skilled Emigrants Care about Political Stability at Home?” Review of Development Economics 15:442–57.

29.

Erdemir

Rittersberger-Tiliç

Ergun

Kahveci

. 2008. Rethinking Global Migration: Practices, Policies, and Discourses in the European Neighbourhood. Ankara, Turkey: KORA.

30.

Esteban

Mayoral

Ray

. 2012. “Ethnicity and Conflict: An Empirical Study.” The American Economic Review 102:1310–42.

31.

European Commission. 2016. “Communication on Establishing a New Partnership Framework with Third Countries under the European Agenda on Migration.” Accessed November 15, 2018. http://www.europarl.europa.eu/RegData/docs_autres_institutions/commission_europeenne/com/2016/0385/COM_COM(2016)0385_EN.pdf.

32.

Fawcett

J. T.

1989. “Networks, Linkages, and Migration Systems.” International Migration Review 23:671–80.

33.

Fearon

J. D.

Laitin

D. D.

. 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97:75–90.

34.

Figueiredo

Lima

L. R.

Orefice

. 2016. “Migration and Regional Trade Agreements: A (New) Gravity Estimation.” Review of International Economics 24:99–125.

35.

Fitzgerald

Leblang

Teets

J. C.

. 2014. “Defying the Law of Gravity: The Political Economy of International Migration.” World Politics 66:406–45.

36.

Flahaux

M. L.

De Haas

. 2016. “African Migration: Trends, Patterns, Drivers.” Comparative Migration Studies 4:1.

37.

Garcia Andrade

Martin

. 2015. “EU Cooperation with Third Countries in the Field of Migration.” Technical Report. http://www.europarl.europa.eu/RegData/etudes/STUD/2015/536469/IPOL_STU(2015)536469_EN.pdf.

38.

Geddes

Scholten

. 2016. The Politics of Migration and Immigration in Europe. London, UK: Sage.

39.

Giménez-Gómez

J. M.

Zergawu

Y. Z.

. 2018. “The Impact of Social Heterogeneity and Commodity Price Shocks on Civil Conflicts.” Journal of Policy Modeling, 40, 959–97.

40.

Gómez

Gasper

. 2013. Human Security: A Thematic Guidance Note for Regional and National Human Development Report Teams. New York: United Nations Development Programme.

41.

Grogger

Hanson

G. H.

. 2011. “Income Maximization and the Selection and Sorting of International Migrants.” Journal of Development Economics 95:42–57.

42.

Hansen

Jonsson

. 2011. “Demographic Colonialism: EU–African Migration Management and the Legacy of Eurafrica.” Globalizations 8:261–76.

43.

Hatton

T. J.

Williamson

J. G.

. 2003. “Demographic and Economic Pressure on Emigration Out of Africa.” The Scandinavian Journal of Economics 105:465–86.

44.

Hatton

T. J.

Williamson

J. G.

. 2005. “What Fundamentals Drive World Migration?” In Poverty, International Migration and Asylum, edited by Borjas

G. J.

Crisp

, 15–38. Berlin, Germany: Springer.

45.

Hirschman

A. O.

1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States, Vol. 25. Cambridge, UK: Harvard University Press.

46.

House

(2017). About freedom in the world: An annual study of political rights and civil liberties. Freedom House.

47.

Howell

L. D.

2011. International Country Risk Guide Methodology. East Syracuse, NY: PRS Group.

48.

Human Rights Watch. 2018. “Towards an Effective and Principled EU Migration Policy: Recommendations for Reform.” Accessed November 15, 2018. https://www.hrw.org/news/2018/06/18/towards-effective-and-principled-eu-migration-policy.

49.

Ibáñez

A. M.

Vélez

C. E.

. 2008. “Civil Conflict and Forced Migration: The Micro Determinants and Welfare Losses of Displacement in Colombia.” World Development 36:659–76.

50.

Italian Government. 2016. “Italian Non-paper. Migration Compact, Contribution to an EU Strategy for External Action on Migration.” Accessed November 15, 2018. http://www.governo.it/sites/governo.it/files/immigrazione_0.pdf.

51.

Kennan

Walker

J. R.

. 2013. “Modeling Individual Migration Decisions.” In International Handbook on the Economics of Migration, edited by Constant

A. F.

Zimmermann

K. F.

, 39–54. Cheltenham, UK: Edward Elgar.

52.

Kerr

S. P.

Kerr

W. R.

. 2011. “Economic Impacts of Immigration: A Survey.” Technical Report. National Bureau of Economic Research. https://econpapers.repec.org/paper/nbrnberwo/16736.htm.

53.

Kim

Cohen

J. E.

. 2010. “Determinants of International Migration Flows to and from Industrialized Countries: A Panel Data Approach Beyond Gravity.” International Migration Review 44:899–932.

54.

Korac

1998. “Ethnic-nationalism, Wars and the Patterns of Social, Political and Sexual Violence against Women: The Case of Post-Yugoslav Countries.” Identities Global Studies in Culture and Power 5:153–81.

55.

Leblang

D. A.

Fitzgerald

Teets

. 2009. Defying the Law of Gravity: The Political Economy of International Migration. Boulder: University of Colorado.

56.

Lewer

J. J.

Van den Berg

. 2008. “A Gravity Model of Immigration.” Economics Letters 99:164–67.

57.

Lucas

R. E.

2015. “African Migration.” In Handbook of the Economics of International Migration, Vol. 1, 1445–596. Elsevier. https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/B9780444537683000266.

58.

Marshall

Jaggers

Gurr

. 2012. “Political Regime Characteristics and Transitions, 1800–2008, Polity IV Project.” http://www.columbia.edu/acis/eds/data_search/1080.html.

59.

Mayda

A. M.

2010. “International Migration: A Panel Data Analysis of the Determinants of Bilateral Flows.” Journal of Population Economics 23:1249–74.

60.

Mody

Sandri

. 2012. “The Eurozone Crisis: How Banks and Sovereigns Came to be Joined at the Hip.” Economic Policy 27:199–230.

61.

Mutua

2009. Human Rights NGOs in East Africa: Political and Normative Tensions. Philadelphia: University of Pennsylvania Press.

62.

Naudé

2010. “The Determinants of Migration from Sub-Saharan African Countries.” Journal of African Economies 19:330–56.

63.

Oette

Babiker

M. A.

. 2017. “Migration Control à la Khartoum: Eu External Engagement and Human Rights Protection in the Horn of Africa.” Refugee Survey Quarterly 36:64–89.

64.

Ongayo

A. O

. 2008. Political Instability in Africa: Where the Problem Lies and Alternative Perspectives. Amsterdam, the Netherlands: The African Diaspora Policy Centre. Accessed October 15, 2009. www.diasporacentre.org/DOCS/Political_Instabil.pdf.

65.

Ortega

Peri

. 2009. “The Causes and Effects of International Migrations: Evidence from OECD Countries 1980–2005.” Technical Report. National Bureau of Economic Research. https://www.nber.org/papers/w14833.

66.

Ortega

Peri

. 2013. “The Effect of Income and Immigration Policies on International Migration.” Migration Studies 1:47–74.

67.

Park

2015. Europes Migration Crisis, 311–25. New York: Council of Foreign Relations.

68.

Pettersson

Wallensteen

. 2015. “Armed Conflicts, 1946–2014.” Journal of Peace Research 52:536–50.

69.

Reynal-Querol

Montalvo

J. G.

. 2005. “Ethnic Polarization, Potential Conflict and Civil War.” The American Economic Review 95:796–816.

70.

Ruble

B. A.

1989. “Ethnicity and Soviet Cities.” Soviet Studies 41:401–14.

71.

Ruyssen

Rayp

. 2014. “Determinants of Intraregional Migration in Sub-Saharan Africa 1980–2000.” Journal of Development Studies 50:426–43.

72.

Schmeidl

2001. “Conflict and Forced Migration: A Quantitative Review, 1964–1995.” In Global Migrants, Global Refugees: Problems and Solutions, edited by Zolberg

A. R.

Benda

P. M.

, 62–94. New York: Berghahn Books.

73.

Shrestha

2017. “Push and Pull: A Study of International Migration from Nepal.” Policy Research Working Paper 7965. Washington, DC: World Bank.

74.

Silva

J. S.

Tenreyro

. 2006. “The Log of Gravity.” The Review of Economics and Statistics 88:641–58.

75.

Solimano

2005. “Development Cycles, Political Regimes and International Migration: Argentina in the Twentieth Century.” In Poverty, International Migration and Asylum, edited by Borjas

Crisp

, 251–75. Dordrecht, the Netherlands: Springer.

76.

Solimano

2010. International Migration in the Age of Crisis and Globalization: Historical and Recent Experiences. Cambridge, MA: Cambridge University Press.

77.

Taylor

A. M.

1994. “Mass Migration to Distant Southern Shores.” In Migration and the International Labor Market, 1850–1939, edited by Hatton

T. J.

Williamson

J. G.

, 91–115. New York: Routledge.

78.

Telschow

2014. Technical Report November. A Policy Paper from the European Network of Political Foundations (ENoP). A new Approach to Migration in the Light of Africa-EU Relations. https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=2ahUKEwiznY3I3cTfAhVIYxoKHUsTBUoQFjABegQIARAC&url=http%3A%2F%2Fwww.enop.eu%2F%3Fmedia_dl%3D4482&usg=AOvVaw1sQNACZ8vYLNvS3RsnDge9.

79.

Tinbergen

1962. “An Analysis of World Trade Flows.” In Shaping the World Economy, edited by Tinbergen

, 1–117. New York: Twentieth Century Fund.

80.

United Nations. 1998. “Recommendations on Statistics on International Migration, Revision 1.” Sales No. E.98.XVII.14. https://unstats.un.org/unsd/publication/seriesm/seriesm_58rev1e.pdf.

81.

United Nations. 2015a. International Migration Report. Technical Report. http://www.un.org/en/development/desa/population/migration/publications/migrationreport/docs/MigrationReport2015_Highlights.pdf.

82.

United Nations. 2015b. “World at War: Global Trends, Forced displacement in 2014.” United Nations High Commissioner for Refugees. Tech. rep. https://www.unhcr.org/556725e69.pdf.

83.

United Nations Conference on Trade and Devolopment. 2018. Migration for Structural Transformation: Economic Development in Africa Report 2018. New York: United Nations Publications.

84.

Weiner

1992. “Security, Stability, and International Migration.” International Security 17:91–126.

85.

Yotov

Y. V.

Piermartini

Monteiro

J. A.

Larch

. 2016. An Advanced Guide to Trade Policy Analysis: The Structural Gravity Model. Geneva, Switzerland: World Trade Organization.

86.

Zlotnik

1991. “South-to-North Migration since 1960: The View from the North.” Population Bulletin of the United Nations 31/32:17–37.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

17.03 MB

0.00 MB

13.65 MB

0.25 MB