Do Emigrants Self-Select Along Cultural Traits? Evidence from the MENA Countries

Abstract

This article empirically investigates whether emigrants from MENA (Middle East and North Africa) countries self-select along two cultural traits: religiosity and gender-egalitarian attitudes. Using Gallup World Poll data on individual opinions and beliefs and migration aspirations, we find that individuals who intend to emigrate to high-income countries exhibit significantly lower levels of religiosity than the rest of the population. They also share more gender-egalitarian views, although this effect holds only among the young (aged 15 to 30), among single women, and in countries with a Sunni minority. For countries most affected by the Arab Spring, the intensity of cultural selection has decreased since 2011. Still, the aggregate effects of cultural selection should not be overestimated. Self-selection along cultural traits has statistically significant but limited effects on the cultural distance between people (i.e., between migrants and natives at destination or between non-migrants in origin and destination countries). Emigration could even reverse the selection effect and lead to cultural convergence if migrants abroad transfer more progressive norms and beliefs to their home country, a mechanism that deserves more attention in future research.

Keywords

F22 015 J61 Z10

1. Introduction

Scholars in sociology, psychology, and anthropology have long emphasized that human decisions are affected by cultural norms, values, and beliefs that differ across human societies (Shweder and Levine 1984; Just and Monaghan 2000; Alexander 2008). More recent studies in demography, economics, and political science are rehabilitating culture as a proximate driver of modernization, human and economic development, and democracy (Hammel 1990; Young 2008; Tabellini 2010; Gorodnichenko and Roland 2017; Collier 2017). As these works show, factors that affect the distribution of cultural traits in human societies are likely to have persistent effects on demographic, economic, and political outcomes. In particular, if emigrants and those left behind share different norms and values, international migration can potentially influence cultural traits distribution.

Focusing on MENA (Middle-East and North Africa) countries, this article empirically tests this hypothesis concerning whether migrants and stayers hold a different set of cultural traits. Although the existing literature has long emphasized that migrants self-select along educational levels (Docquier, Lohest, Marfouk 2007; Grogger and Hanson 2011; Belot and Hatton 2012), migrants’ selection along cultural traits has been understudied. Using Gallup World Poll microdata, we construct proxies for religiosity and attitudes toward women’s rights, two cultural traits that correlate with economic development. We then empirically investigate (1) whether these two cultural traits affected the aspirations and plans of MENA natives to emigrate between 2007 and 2016, (2) whether selection on cultural traits varies across groups of respondents and with country-specific characteristics, and (3) whether the selection intensity changed after the Arab Spring (i.e., after the wave of anti-government protests across the Middle East that started in December 2010 in Tunisia (Haas and Lesch 2016). It is worth stressing that our analysis does not make any value judgment about specific cultural traits or argue that cultural differences should be combatted or that one set of traits dominates others. Instead, our goal is to shed light on the relationship between migration aspirations, preferred destination choices, and cultural traits.

As is pointed out by the rapidly growing literature on the relationship between culture and economics (Guiso, Sapienza, and Zingales 2009; Duflo 2012; Rapoport, Sardoschau, and Silve 2018), there are two important reasons to focus on cultural selection in general and on selection by religiosity and gender-egalitarian views in particular. First, cultural selection is a main mechanism through which emigration affects the distribution of cultural traits in the population left behind (Rapoport, Sardoschau, and Silve 2018). If not compensated for by ex-post transfers of norms and beliefs from destination to origin countries, selection on gender-egalitarian attitudes is likely to impact effective gender discrimination in the origin countries. In addition, cultural selection may increase the cultural distance between poor and rich countries, and such distance has been seen as a brake on knowledge diffusion and the transmission of democratic values (Desmet et al. 2011; Spolaore and Wacziarg 2016). Second, cultural selection is a key determinant of the cultural distance between migrants and host-country citizens and, therefore, shapes the level of cultural diversity, opinions toward immigration, and migrants’ capacity to assimilate at destination (Card, Dustmann, and Preston 2006, 2012). Below, we provide a more extensive discussion of these issues, followed by a review of the scarce literature on migrants’ selection along cultural traits.

From an origin-country perspective, the idea of culture as an important determinant of human, political, and economic development probably starts with Weber’s theory of the Protestant work ethic (Weber 1946; Guiso, Sapienza, and Zingales 2006; Spolaore and Wacziarg 2013). Culture here is seen as shaping individual effort and the overall quality of the institutions that support market-friendly exchange and openness to innovation (Spolaore and Wacziarg 2012). It is also understood to play a role in the formation of trust within and between countries (Guiso, Sapienza, and Zingales 2009). Recent works have produced sound empirical tests of the link between development outcomes and culture, often using opinion survey data to measure cultural elements such as economic beliefs (Piketty 1995; Di Tella and MacCulloch 2014) and trust (Knack and Keefer 1997; Algan and Cahuc 2010; Felbermayr and Toubal 2010). The effect of cultural distance between countries on the speed of knowledge diffusion (Spolaore and Wacziarg 2012) and on democracy transitions (Murtin and Wacziarg 2014) is another aspect on which the literature has increasingly focused.¹

The relationship between culture and economic performance has been recently investigated in an Arab context, showing the link between cultural beliefs and economic outcomes (Kuran 2012; Kostenko et al. 2017; Diwan, Tzannatos, and Akin 2018). In particular, a series of recent studies presented below have focused on the connection between views on gender inequalities, religiosity, and economic development. Within such work, gender-egalitarian attitudes have been shown to play key roles in explaining effective discriminatory behaviors toward women (Baxter and Kane 1995; Bergh 2007), and resulting gender inequalities in health, education, political empowerment, and employment have long been seen as major barriers to development (Duflo 2012; UN 2015). Concerning religiosity, Benabou, Ticchi, and Vindigni (2015) and Chase (2014) identify a negative association between individual religiosity and openness to innovation and effective patents per capita.

Analyzing the determinants of religiosity and gender-egalitarian attitudes in the MENA context is particularly relevant, since MENA countries hold significantly less egalitarian attitudes toward women’s employment and higher levels of religiosity, as compared to other regions (Price 2015). Moreover, the renewed upturn of patriarchal views among MENA countries, due to the rise of political Islam, has intensified gender inequality (Tzannatos 1999; Stetter 2008; Alexander and Welzel 2011; Norris and Inglehart 2011; El Mikawy, Mohieddin, and El Ashmaouy 2017). For all these reasons, the MENA region displays distinctive cultural traits whose impacts on migration aspirations merit further attention.

From a destination-country perspective, migrants’ selection on culture (we define culture in Section 2.2.2) is likely to influence the level of cultural diversity in the host country. Although diversity induces beneficial effects for the host country (Ottaviano and Peri 2006; Alesina, Harnoss, and Rapoport 2016), several empirical studies show that immigrants’ economic and socio-demographic outcomes at destination also depend on the distance between their identity and dominant norms of the host society (Pendakur and Pendakur 2005; Fernandez and Fogli 2009; Alesina and Giuliano 2010; Battu and Zenou 2010; Casey and Dustmann 2010; Bisin et al. 2011; Islam and Raschky 2013). Perceived cultural distance between natives and immigrants can be a source of negative attitudes toward immigrants (Card, Dustmann, and Preston 2006), leading to discrimination and marginalization. In particular, Islamophobia has been increasing in Western societies, and around 70 percent of Western natives think that tensions between the Muslim and Western worlds originate in cultural and religious differences (Gallup 2010).

In this context, migrant selection on culture has potentially important effects on many outcomes. Much research has emphasized the effect of the average cultural distance between countries on dyadic migration flows (Mayda 2006; Belot and Ederveen 2012; Krieger, Renner, and Ruhose 2018). However, to the best of our knowledge, very few studies investigate the link between individual cultural traits and migration aspirations, using micro-data. Berlinschi and Harutyunyan (2016), examining the Life in Transition Survey, identify a positive correlation between migration aspirations from Eastern European countries and opinions about home-country governance, political participation, and trust in other people. More related to our analysis, Myers (2000) finds that US citizens’ migration aspirations negatively correlate with involvement in social activities related to religion. Hoffman, Marsiglia, and Ayers (2015) find that external religiosity (e.g., participating in religious activities) and internal religiosity (e.g., spirituality) induce different effects on the migration aspirations of Roman-Catholic Mexican students. Falco and Rotondi (2016), using the Arab Barometer on nine Islamic countries, show a negative correlation between individuals’ attachment to Islamic law and migration aspiration. Our article uses a similar specification but focuses on different origin countries and on specific cultural traits often thought to cause cultural tensions in western societies. Taking advantage of our database’s coverage, our article sheds light on the heterogeneity — across groups of individuals and across countries — of cultural selection patterns in the MENA region.

The rest of this article is organized in the following way. Section 2 explains our research objectives and theoretical approach, while describing and validating the data sources that proxy for migration aspirations and culture in our study. Section 3 discusses the empirical specification and econometric issues related to our analysis. Estimation results are presented in Section 3, including our full set of robustness checks, heterogeneity analysis, and Arab Spring analysis. The social and policy implications of our results are discussed in Section 5.

2. From Theory to Data

2.1. Objectives and Theoretical Framework

There is a large literature in demography and economics investigating the determinants of country- or group-specific aspirations to migrate. To mention just a few, Becerra et al. (2010) and Becerra (2012) show that lower socioeconomic status is not enough to explain intentions to migrate for Mexican adolescents in Tijuana; other factors like perceived discrimination and pre-migration acculturation also play key roles. In the same vein, Wood et al. (2010); De Jong, Richter, Isarabhakdi (1996); and Drinkwater and Ingram (2009) highlight the importance of noneconomic factors as determinants of migration intentions.

In line with the above studies, we consider individual emigration as a two-step process. The first step relates to migration aspirations: “aspiring” migrants express an intention/desire to emigrate and search for migration opportunities, while others decide (or are forced) to stay put. The second step relates to the realization of these aspirations: some in the pool of aspiring migrants find opportunities to migrate (such as a job offer or visa), while others do not. Our model focuses only on the first step of the process. Theoretically speaking, we model migration aspirations of each individual i living in region r at year t as depending positively on the returns and negatively on the costs of migrating (as in Chort 2014; Docquier, Peri, and Ruyssen 2014; Ruyssen and Salomone 2018). Migration aspirations also depend on the availability of information about economic and social conditions abroad. Returns, costs, and information sets vary across individuals and are affected by a set of individual observed characteristics ( $X_{i r t}$ ) and a random component ( $∊_{i r t}$ ) that captures unobserved heterogeneity in preferences. Thus, the net benefit from migrating, $N B_{i r t}$ , can be written as:

{NB}_{irt} = f (X_{irt}) + ∊_{irt},

where f(.) is a general function that determines the net returns from migrating. The decision to become an aspiring migrant is expressed as

m_{i r t} = {\begin{cases} 1 i f {NB}_{irt} > 0 \\ 0 i f {NB}_{irt} \leq 0 \end{cases}

In the empirical section, we assume that ∊_irt follows a logistic distribution and that f(X_irt ) is a linear function of individual characteristics. Under these assumptions, a logit model governs the probability that individual i becomes an aspiring migrant:

Pr (m_{i r t} = 1) = Φ (X_{irt} β),

where Φ(.) is the logistic function and β is a set of parameters to be estimated.

The vector X_irt includes standard regressors (e.g., education, age, gender, presence of family members or friends abroad, etc.). The literature on the drivers of aggregate migration stocks and flows has identified a statistically significant effect of migration selection on education, which correlates with many unobserved factors and with cultural distance between countries (Krieger, Renner, and Ruhose 2018). At the micro level, we hypothesize that cultural traits influence the returns from and costs of migrating, since cultural distance can affect the net returns from migrating through labor-market discrimination, racism, violence, and other factors (Adsera and Pytlikova 2015; Falck et al. 2012; Falck, Lameli, and Ruhouse 2018).

2.2. Data Sources

To test whether aspiring migrants are culturally selected compared to intended stayers, we need micro-data on migration aspirations, cultural traits, and other individual characteristics. Such data can be obtained from the representative Gallup World Polls (GWP). Although the GWP database covers 148 countries, our sample is limited to 17 MENA countries where Gallup conducted at least one wave of its survey between 2007 and 2016.² On average, the sample includes about 1,000 randomly selected respondents per year and per country. For most countries in our sample, the data were collected through face-to-face interviews, except in Iran and Iraq, where interviews were conducted through phone calls, due to the high diffusion of telephone landlines. The sampling frame is such that GWP data are representative of the entire population aged 15 and over (including populations from rural areas). Our full sample includes 60,284 working-age native respondents (i.e., individuals aged 15 to 64).³

The GWP includes several questions capturing respondent beliefs and values, migration aspirations, preferred destination choices, and any active steps individuals were taking to emigrate. The fact that the GWP database covers many countries makes it exceptional. As the data are relatively new, the literature relying on these data to capture migration aspirations is limited, at both the micro (Manchin, Manchin, and Orazbayev 2014; Dustmann and Okatenko 2014; Bertoli and Ruyssen 2018; Ruyssen and Salomone 2018) and macro level (Docquier, Peri, and Ruyssen 2014; Docquier, Machado, and Sekkat 2015; Dao, Docquier, and Parsons 2018).

2.2.1. Measuring migration aspirations

The GWP questions used in this article to measure respondents’ migration aspirations include the following:

Ideally, if you had the opportunity, would you like to move permanently to another country, or would you prefer to continue living in this country?

To which country would you like to move?

Are you planning to move permanently to another country in the next 12 months, or not?

We define aspiring migrants as those who answered the first question affirmatively.⁴ Note that the last two questions were asked to aspiring migrants only. In line with Bertoli and Ruyssen (2018), we find that migration aspirations correlate with actual migration flows. Using the annual flow data from the International Migration Database of the OECD (the Organization for Economic Co-operation and Development), the correlation with aspirations is positive (0.435) and significant at the 1 percent threshold. Hence, patterns of migration aspirations are likely to be similar to patterns of actual migration, although not the same, due to the lack of realization of migration aspirations and information of flows through irregular channels.

The average share of aspiring migrants in our sample is around 24 percent, although large variations exist across countries. Countries exhibiting the greatest shares of aspiring migrants are Syria (35.6%) and Jordan (27.6%), while the smallest shares are present in Niger (16.1%) and Azerbaijan (18.3%). Cross-country variations in destination choices are even larger, due to historical and colonial reasons. On average, the share of aspiring migrants who would like to emigrate to an OECD destination country equals 52.3 percent.⁵ The latter share amounts to 90.9 percent in Morocco and 86.7 percent in Algeria but only 10 percent in Yemen and 12.8 percent in Niger.

2.2.2. Measuring cultural traits

Our objective in this article is to test whether cultural traits affect the aspiration to migrate. The GWP database includes several questions on cultural norms, beliefs, values, and attitudes. Aware that culture is a broad and vague concept (Shenkar 2012) and that measuring it is a complex task, we follow Guiso, Sapienza, and Zingales (2006), who define culture as those beliefs and values that are transmitted fairly unchanged through generations. Among those beliefs, we focus on religiosity (individual position toward religion) and gender attitudes (toward the role of women in society). Aware that confining culture to these two cultural traits is shortsighted, the literature shows that these traits exhibit peculiar distributions in MENA countries (Price 2015) and are highly correlated with economic development (e.g., Duflo 2012; Benabou, Ticchi, and Vindigni 2015). Specifically, we proxy those two cultural traits, using the following questions:

Q1. Is religion an important part of your daily life?

Q2. Have you attended a place of worship or religious service within the past seven days?

Q3. Do you agree that women and men should have equal legal rights?

Q4. Do you agree that women should be allowed to hold any job for which they are qualified outside the home?

Q5. Do you agree that women should have the right to initiate a divorce?

The first two questions were asked in all countries surveyed by GWP, while the last three were asked in specific geographical regions. We normalize responses between 0 and 1 such that sharing gender-egalitarian views and not being religious are coded as one. With those narrowed proxies, we capture individual beliefs, although concerning gender attitudes, we cannot grasp their actual influence in everyday life.

Several methods can be used to extract indices on cultural traits.⁶ We conduct a principal component analysis (PCA), which allows us to identify linear combinations of questions that explain the greatest share of variance in cultural traits. The PCA results are presented in Figure A1 and Table A3 in the supplemental appendix. The PCA’s first two components explain the highest amount of variance. The first reflects gender-egalitarian views whereas the second reflects religiosity. To compute our benchmark indices of Gender-egalitarian attitudes and Religiosity, we combine individual responses to the GWP questions, using the values of the eigenvectors as weights.⁷

Other methods can be used to aggregate cultural questions. In the supplemental appendix, we consider four alternative methods (factor analysis, multiple correspondence analysis, kernel principal component analysis, and simple average), all of which obtain very similar results (see Table A11 in the supplemental appendix). Each method generates new indices of gender-egalitarian attitudes and religiosity. Differences across indices are discussed in Table A5 in the supplemental appendix. Tables A8 and A9 in the supplemental appendix show that the correlation between the new indices and PCA ones is large and highly significant. Given the strong robustness of our findings, all results presented in the core of this article rely on the standard PCA aggregation method.

Table 1 reports the mean value of each indicator by country. Lebanon and Azerbaijan are the most progressive in terms of gender-egalitarian attitudes, probably due to the fact that Azerbaijan has a Soviet legacy and high adult literacy rate (UNDP 2000) and Lebanon is about 44 percent Christians and a pluri-confessional society (Karouby 2014). Iran, Tunisia, and Azerbaijan are the least religious countries. Finally, sub-Saharan African countries (Chad, Mauritania, Mali, and Niger) exhibit high levels of religiosity. The geographical distribution of these cultural traits is plotted in Figures 1 and 2.

Table 1.

Cultural Traits: Mean Levels by Country.

Countries	Gender	Religiosity	Countries	Gender	Religiosity
Afghanistan	0.502	0.156	Mauritania	0.684	0.146
Algeria	0.732	0.232	Morocco	0.675	0.234
Azerbaijan	0.846	0.633	Niger	0.552	0.104
Chad	0.767	0.098	Palestine	0.686	0.272
Egypt	0.738	0.188	Syria	0.668	0.245
Iran	0.8	0.342	Tunisia	0.739	0.355
Iraq	0.594	0.311	UAE	0.805	0.163
Jordan	n.a.	0.691	Yemen	0.676	0.237
Kuwait	0.761	0.137	All	0.699	0.256
Lebanon	0.853	0.313	Main Insurgents	0.713	0.267
Mali	0.644	0.164	Others	0.691	0.252

Source: Authors’ calculation on the Gallup World Poll.

All the values in the table are the mean values of each indicator. Main Insurgents: Algeria, Egypt, Syria, Tunisia and Yemen.

Figure 1.

Cross-country heterogeneity in gender-egalitarian views.

Figure 2.

Cross-country heterogeneity in religiosity.

GWP is not the only database documenting the distribution of cultural traits. For example, several questions of the World Values Survey (WVS) can also be used to document beliefs and values. However, the WVS includes a smaller set of countries and has no specific question on migration plans or aspirations.⁸ Still, some WVS questions closely relate to our two indices of cultural traits. For religiosity, the WVS’s sixth wave includes four questions: (1) How important in life is religion? (2) How often do you attend religious service? (3) How often do you pray? and (4) Are you a religious person? For gender-egalitarian attitudes, the WVS asked respondent opinions on two statements on gender-egalitarian views: (1) When jobs are scarce, men should have more right to a job than women; and (2) On the whole, men make better political leaders than women do. We normalize WVS responses between 0 and 1, using the same order as before.⁹ The upper panel of Table 2 reports the correlations between the country-specific mean levels of our indicators and of the WVS data. Our index of religiosity is highly correlated with WVS responses. Our gender-egalitarian index is poorly correlated with the WVS index of economic equality. It is, however, well correlated with the WVS index of equality in politics, indicating a convergence in measuring gender attitudes across data sources.

Table 2.

Correlation between Aggregate Cultural Gallup World Polls (GWP), World Value Survey (WVS), and Macro Indices.

Indicator	Gender		Religiosity
WVS indices
Important in life: religion	—	—	—	0.832***
Attending religious service	—	—	—	0.843***
Praying frequently	—	—	—	0.718**
Religious Person	—	—	—	0.947***
Gender equality: job	—	0.195	—	—
Gender equality: politics	—	0.646**	—	—
Macro indicators
Sunnis	−0.296	—	−0.456*	—
Shiites	0.366	—	0.651***	—
ln GDP	0.644***	—	0.693***	—
Rule of law	0.259	—	0.101	—
Control corr.	0.157	—	−0.049	—
Network	0.345	—	0.199	—

Source: Authors’ calculations based on the Gallup data, CIA World Factbook, World Bank indicators, Maddison Project, UN databases, and WVS.

For the WVS data, the list of countries available is the following: Algeria, Azerbaijan, Egypt, Jordan, Iraq, Lebanon, Morocco, Palestine, Tunisia and Yemen.

*p < 0.1. **p < 0.05. ***p < 0.01.

2.2.3. Correlates of cultural traits

We investigate whether our proxies for cultural traits correlate with six macro indicators capturing the branch of Islam (Sunni or Shia) that is prevalent in the origin country, the level of economic development, the quality of institutions, and past migration flows. Data on the shares of Sunnis and Shiites in the Muslim population are taken from the CIA World Factbook and the PEW Research Center; data on gross domestic product (GDP) per capita are obtained from the Maddison Project; data on the control of corruption and the rule of law are taken from the World Bank’s Worldwide Governance Indicators. As for past migration, we compute the percentage of respondents with a family member or friend abroad from the GWP data (a proxy for migration networks).¹⁰

Correlation coefficients are reported in the lower panel of Table 2. Two main findings emerge from this table. First, our index of religiosity is highly correlated with the Muslim population’s composition. Countries with a greater share of Shiites are less religious than countries with a greater share of Sunnis. Second, the level of development is highly correlated with religiosity and gender-egalitarian attitudes, in line with the empirical literature on culture and economic growth (Duflo 2012; Benabou, Ticchi, and Vindigni 2015; Chase 2014). The strong relationship between religiosity, gender-egalitarian attitudes, and economic development is the main reason we focus on these two cultural traits, as well as because of the poor correlation between our indices and the quality of institutions and previously existing emigrant stocks.

3. Empirical Strategy

Our goal is to analyze the determinants of migration aspirations and to test whether these aspirations are affected by cultural traits. Below, we describe the benchmark empirical specification and discuss some econometric issues.

3.1. Benchmark Specification

As explained in Subsection 2.1, our benchmark empirical model features the intention to migrate as the dependent variable. For respondent i originating from region r at year t, the variable $m_{i r t}$ takes a value of 1 if individual i expresses a desire to migrate abroad permanently and 0 otherwise. In line with Equation (3), we opt for a logit model:

Prob {m_{i r t} = 1 | C u l t u r e_{i r t}, X_{irt}} = Φ {α + β C u l t u r e_{i r t} + {ΓX}_{irt}},

where $C u l t u r e_{i r t}$ is our proxy of cultural traits ranging from 0 (when they do not share gender-egalitarian views or are religious) to 1; $X_{i r t}$ is the set of other determinants of migration aspirations that vary across individuals; α, β, and Γ are the parameters to be estimated. We cluster the error terms at the country level.

The set of control variables includes age, gender, marital status, the presence of children in the household, the level of income per household member and its square, education level (a dummy variable equal to one if the respondent has least 9 years of education), and the presence of a friend or relative abroad. These variables are denoted by x_irt , a subset of X_irt (as explained below). In line with the existing literature (Dustmann and Okatenko 2014; Manchin, Manchin, and Orazbayev 2014), these variables affect the size of migration costs, as well as the expected gains from migration.

As explained above, migration aspirations correlate with actual migration flows, in line with Bertoli and Ruyssen (2018). However, the number of aspiring migrants is much greater than the number of actual migrants. Hence, cultural traits and other determinants may affect realization rates and may have heterogeneous effects on the desire to emigrate and the capacity to realize that aspiration. Hence, as a robustness check, we also estimate Equation (4), using migration plans as a dependent variable. The question is: “Are you planning to move permanently to another country in the next 12 months, or not?” This question is asked if the answer related to migrate intention is affirmative. This dependent variable takes a value of 1 if the respondent i is actually making steps to move to another country within 12 months and 0 otherwise.

Equation (4) is estimated using the sample of working-age respondents living in the 17 MENA countries and the main cultural proxies identified in Subsection 2.2.2. In addition, our estimates can be affected by the presence of immigrants in the sample, since the latter are likely to exhibit different characteristics and cultural traits and to have different migration strategies (e.g., transit migrants). To keep the sample as homogenous as possible, we exclude the foreign born from the sample and only consider native residents.

3.2. Empirical Challenges

Our approach entails several methodological issues that might lead the logit model to generate inconsistent estimates. Below, we discuss how we deal with heterogeneous effects and endogeneity problems.

3.2.1. Heterogeneous effects

Migration patterns in general, and the role of cultural traits in particular, may vary according to the regional context, choice of destination, and individual characteristics (Ichino and Maggi 2000; Fernandez and Fogli 2009). To test for heterogeneity across origin countries, we first estimate Equation (4) at the country level. Since we find large variations across countries, we augment Equation (4) with some country-specific variables and their interaction with cultural proxies. In line with Table 2, the set of country-specific characteristics includes the country shares of Sunnis and Shiites among the Muslim population for the log of GDP per capita, two indicators of institutional quality, and the share of native citizens living in an OECD member state.

Second, to test for heterogeneity across periods, we distinguish between the pre– and post–Arab Spring periods. The Arab Spring started in December 2010 in Tunisia (with the attempted self-immolation of Mohamed Bouazizi) and triggered riots and political unrest in several MENA countries in subsequent months (Haas and Lesch 2016). Most MENA economies were adversely affected in the post-Arab Spring period with an increase of unemployment throughout the region and negative economic growth, especially in Tunisia and Yemen (Richards et al. 2014). However, the Arab Spring may have also impacted cultural norms and migration intentions. Gill (1999) notes that socio-economic crises can cause a loss in cultural identity, making individuals more susceptible to religion, as we observe in our case. In particular, the political instability and rise of authoritarianism that has characterized the post-Arab Spring period (sometimes referred to as the Arab Winter) may have affected aspiring migrants’ process of cultural selection.¹¹ We exploit this possible source of variation by focusing on the first GWP question on religiosity (see Q1 above), which is the only question that spans all years from 2007 to 2016. The correlation between responses to Q1 and our synthetic indicator of religiosity is large (0.688) and highly significant.¹² We then estimate the following equation:

Prob {m_{i r t} = 1 | R e l i g_{i r t}, X_{irt}, A S_{r t}} = Φ {α + β R e l i g_{i r t} + δ R e l i g_{i r t} * A S_{r t} + ζ A S_{r t} + η A S_{r t} * x_{i r t} + {Γ X}_{irt}},

where AS_rt is a dummy variable that captures the post–Arab Spring period (i.e., a dummy equal to 1 for the years 2011 to 2016), and Relig_irt is a dummy variable that takes the value of 1 if religion is an important part of daily life for individual i from region r at year t ∈ {2007,…, 2016}. Although Equation (5) is unable to capture additional shocks in the post–Arab Spring period, due to the Arab Spring’s historical relevance, it is not unreasonable to treat it as the main shock in the period of analysis. Moreover, the inclusion of time fixed effects as presented in Equation (6) will remove common time-trend across countries. Thus, the potential estimation bias due to shocks unrelated to the Arab Spring could be present, even though negligible.

The estimated coefficients β and δ capture the correlation between religiosity and intention to migrate and the Arab Spring’s effect on such correlation, respectively. If δ is positive and significant, it implies that intending migrants are less religious after the Arab Spring. We are fully aware that the Arab Spring was heterogeneous across countries. Governments were overthrown in some countries (e.g., Tunisia), while the Arab Spring had smaller effects in other countries (e.g., Mauritania) (Haas and Lesch 2016). To account for this heterogeneity, we estimate Equation (5) on the full set of countries, on a sample of countries that were highly impacted by the Arab Spring (i.e., Main Insurgents),¹³ and on the other MENA countries. We expect a stronger and more significant effect of Arab Spring among the Main Insurgents countries.

Third, at the individual level, we investigate whether cultural selection is affected by aspiring migrants’ intended destination. If migrants have cultural values that are more similar to those of the intended host country, we expect to find heterogeneous selection patterns across preferred destination types. To deal with this issue, we estimate our model with a modified dependent variable and distinguish between migration aspirations toward OECD member states and non-OECD destinations. Furthermore, we perform several robustness checks by sub-samples, distinguishing between age groups (15–30, 31–45, 45–65), education groups (respondents with less than 9 years of education or more), employment status (unemployed, employed, out of the labor force), gender and marital status (married and unmarried individuals, female and male), religious groups (Muslims, Christians, others), and place of residence (farm, town, city).

3.2.2. Omitted variables

Although we control for a traditional set of individual characteristics, migration aspirations can be influenced by additional characteristics that we do not observe. These omitted variables can be related to respondents’ regional environment (e.g., governance and security in the region, ethnic composition of the population, climatic conditions, distribution of cultural traits, percentage of native citizens abroad, etc.) or their own characteristics (e.g., cognitive skills and abilities, family ties, etc.). These unobserved characteristics may jointly affect the acquisition of cultural traits and migration aspirations (see Bisin and Verdier 1998). To deal with unobserved regional characteristics, we take advantage of the fact that GWP identifies the respondent’s detailed geographical location (within the country) and covers several years. For these reasons, we systematically augment the set of controls with spatial and year (or GWP wave) fixed-effects and estimate a fixed-effect logit model. Hence, the full set of control variables in Equations (4) and (5) is

{Γ X}_{irt} = θ_{r} + θ_{t} + γ x_{i r t},

where $θ_{r}$ and $θ_{t}$ are the intra-country region and year/wave fixed effects (justified here by the Hausman test), and γ is the vector of incidence parameters related to the observed controls.

As for individual characteristics, our specification could still lack unobserved factors that jointly influence migration aspirations and cultural traits. Our estimated coefficients would then capture spurious correlations. Moreover, disparities in the distribution of covariates between aspiring migrants and non-migrants may influence our estimates’ accuracy. As shown by Imbens and Rubin (2012), large covariates’ distributional gaps magnify the estimated coefficients’ sensitivity to any ostensibly minor change in the specification. To address this issue, we implement a design phase that precedes the empirical analysis and constructs a balanced sample in terms of observed covariates. In practice, we match aspiring migrants with non-migrants, using the Mahalanobis Metric Matching method to minimize the distance on observable covariates between them. Such matching technique generates samples with the same number of aspiring migrants and not-migrants in each country. This method allows us to compare intending and non-intending migrants that are really alike in the whole set of observables, apart from their cultural traits. If unobserved factors correlate with observed factors, we can also assume that a reduction of observable differences will reduce unobservable differences, minimizing this potential source of bias (see Altonji, Elder, and Taber 2005; Oster 2017). Moreover, if cultural traits are exogenous and persistent (see Guiso, Sapienza, and Zingales 2006), we can claim a certain degree of causality to our estimated coefficients. Otherwise, the Mahalanobis Metric Matching method simply identifies a cleaner correlation. Since unobserved factors can be also relevant in the estimation of Equation (5), we use the same method to build a balanced sample in terms of covariates, matching individuals pre– and post–Arab Spring. We then conduct our regressions on the balanced samples, making the estimates more credible and robust.

4. Results

This section begins by investigating the effect of cultural traits on migration aspirations, using fixed-effects logit regressions and distinguishing between OECD and non-OECD destinations. In addition, we check whether similar cultural selection patterns can be identified when considering short-run migration plans (instead of migration aspirations). In Section 4.2, we perform several robustness checks to explore how our results are influenced by individual characteristics (see Subsection 4.2.1) and whether they are robust to the matching technique (see Subsection 4.2.2). We account for heterogeneous effects across countries in Subsection 4.2.3. Finally, Subsection 4.2.4 investigates the Arab Spring’s effect on cultural self-selection.

4.1. Logit Regressions

Table 3 focuses on migration aspirations and describes the results of the fixed-effect logit regressions for the full sample of MENA countries and by destination type. Columns (1) and (2) report estimates for migration aspirations to all destinations, showing that aspirations correlate with religiosity. The coefficient of religiosity is positive (0.389) and significant at the 1 percent threshold. The logit model is non-linear. To illustrate the magnitude of this effect, we define the benchmark category of respondents as males with college education, aged 24 to 35, married with children, without friends or relatives abroad, and with a level of religiosity equal to the sample mean (0.259). The same benchmark category is used below to interpret the results of other regressions. For this category of respondents, increasing our indicator of religiosity by one standard deviation (+0.315) raises the desire to emigrate by 12.2 percentage points. Although positive, the effect of gender-egalitarian views is not significantly different from zero. Control variables are usually significant and have intuitive signs. In line with the literature (Dustmann and Okatenko 2014; Manchin, Manchin, and Orazbayev 2014), aspirations are higher for young, single men with higher education, lower levels of income per household member, and friends or relatives abroad. Interestingly, our quadratic specification identifies a hump-shaped effect of household income, in line with the mobility transition literature (see Zelinsky 1971).

Table 3.

Benchmark Regressions.

	(1)	(2)	(3)	(4)	(5)	(6)
Dep. Var. Culture Destination	Int Mig Gend All	Int Mig Rel All	Int Mig Gend OECD	Int Mig Rel OECD	Int Mig Gend Non-OECD	Int Mig Rel Non-OECD
Culture	0.051	0.389***	0.142	0.472***	–0.092	0.227
Culture	(0.101)	(0.102)	(0.103)	(0.101)	(0.150)	(0.168)
Female	–0.658***	–0.598***	–0.714***	–0.701***	–0.588***	–0.411***
Female	(0.107)	(0.097)	(0.100)	(0.090)	(0.154)	(0.149)
Network	0.677***	0.615***	0.708***	0.676***	0.689***	0.537***
Network	(0.067)	(0.060)	(0.083)	(0.069)	(0.072)	(0.076)
Age	–0.296***	–0.261***	–0.326***	–0.305***	–0.256***	–0.188***
Age	(0.039)	(0.040)	(0.044)	(0.040)	(0.040)	(0.044)
Education	0.179***	0.113	0.253**	0.221*	0.084	–0.019
Education	(0.058)	(0.074)	(0.103)	(0.116)	(0.074)	(0.078)
Marital	–0.257***	–0.401***	–0.354***	–0.481***	–0.167***	–0.291***
Marital	(0.055)	(0.088)	(0.058)	(0.076)	(0.061)	(0.108)
Children	–0.049	–0.021	–0.056	–0.039	–0.010	0.012
Children	(0.070)	(0.040)	(0.100)	(0.038)	(0.072)	(0.073)
Income	–0.088**	–0.075**	–0.046	–0.068*	–0.136**	–0.103***
Income	(0.036)	(0.032)	(0.042)	(0.041)	(0.053)	(0.030)
Income²	0.005**	0.002***	0.003	0.002**	0.008**	0.003***
Income²	(0.002)	(0.001)	(0.002)	(0.001)	(0.003)	(0.001)
Observations	32,805	34,017	28,937	30,038	28,065	28,240
Partial R ²	0.11	0.11	0.17	0.18	0.11	0.10
Region f.e.	Yes	Yes	Yes	Yes	Yes	Yes
Year-Wave f.e.	Yes	Yes	Yes	Yes	Yes	Yes

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. Std. errors are clustered at the country level. The dependent variable (Dep. Var.) is an intention to migrate (Int Mig) dummy. Gend = gender; Rel = religiosity; f.e. = fixed effect. OECD destinations: US, UK, France, Germany, Netherlands, Spain, Italy, Poland, Hungary, Sweden, Greece, Denmark, Israel, Canada, Australia, New Zealand, South Korea, Austria, Estonia, Finland, Japan, Mexico, Belgium, Turkey, Iceland, Ireland, Latvia, Norway, Portugal, Slovenia, Switzerland, Czech Rep.

*p < 0.1. **p < 0.05. ***p < 0.01.

The rest of Table 3 distinguishes between migration aspirations to OECD and non-OECD destinations. Columns (5) and (6) reveal that cultural traits have insignificant impact on migration aspirations to non-OECD countries. On the contrary, low levels of religiosity positively correlate with intentions to emigrate to OECD destinations at the 1 percent significance level, as shown in Column (4), and the effect is bigger compared to the results in Column (2). Increasing our indicator by one standard deviation raises the desire to emigrate to OECD destinations by 14.9 percentage points (for the benchmark category of respondents). The correlation between gender-egalitarian views is also bigger; however, it still not precisely estimated. Aspirations to migrate to OECD destinations are even more influenced by education attainment and the presence of network members abroad. Hence, Table 3 evidences that aspiring migrants from MENA countries self-select along cultural traits but only when they intend to migrate to OECD destinations and exhibit lower levels of religiosity than those who do not intend to migrate.

In Table 4, we check whether similar selection patterns apply to individual with concrete migration plans (i.e., those taking concrete steps to leave their country within the next 12 months). Columns (1) and (3) report the results from Table 3 for migration aspirations. Relying on the same specification, Columns (2) and (4) provide the results for migration plans. Column (2) shows that the effect of gender-egalitarian views is insignificant. By contrast, the effect of religiosity is highly significant and greater than for migration aspirations. We obtain a coefficient of 0.771, which means that increasing our indicator by one standard deviation (+0.315) raises the probability to have concrete migration plans by 24.3 percentage points. Note that the mean proportion of individuals taking steps to move within 12 months equals 1.9 percent and that its standard deviation equals 13.9 percentage points. We thus find evidence of religiosity’s effect on migration plans, implying that emigration to OECD countries affects the distribution of cultural traits in the population left behind.

Table 4.

Results for Intentions and Plans to Migrate.

	(1)	(2)	(3)	(4)
Dep. Var. Culture Destination	Int Mig Gend OECD	Plan Mig Gend OECD	Int Mig Rel OECD	Plan Mig Rel OECD
Culture	0.142	–0.077	0.472***	0.771***
Culture	(0.103)	(0.173)	(0.101)	(0.243)
Observations	28,937	10,834	30,038	12,199
Partial R ²	0.17	0.16	0.18	0.23
Region f.e.	Yes	Yes	Yes	Yes
Year-Wave f.e.	Yes	Yes	Yes	Yes

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. Std. errors are clustered at the country level. OECD destinations. Dependent variable (Dep. Var.): intention to migrate (Int Mig); plan to migrate in the next 12 months (Plan Mig). Gend = gender; Rel = religiosity; f.e. = fixed effect.

*p < 0.1. **p < 0.05. ***p < 0.01.

4.2. Robustness Analysis

4.2.1. Robustness by subsample

This section investigates whether the identified self-selection patterns vary by intended destination country or group of individuals. We begin by splitting the set of OECD destinations into three subsets of countries frequently reported as preferred destinations in the data: the European Union, North America (i.e., Canada and the United States), and Turkey. Separate fixed-effect logit regressions are used to explain migration aspirations to these three sets of countries. Results are provided in Table 5. Columns (1), (3), and (5) confirm that the effect of gender-egalitarian views remains insignificant for all sets of destination. Columns (2) and (4) show that religiosity’s effect is highly significant when considering OECD, high-income destinations. We also notice that the intensity of the self-selection process is greater for individuals intending to migrate to North America (0.679) than for those intending to migrate to Europe (0.408). We find no evidence of cultural selection toward Turkey.

Table 5.

Robustness Analysis by Preferred Destination Countries.

Culture Destination	(1)	(2)	(3)	(4)	(5)	(6)
Culture Destination	Gend EU	Rel EU	Gend US/Can	Rel US/Can	Gend Turk	Rel Turk
Culture	0.014	0.408***	0.170	0.679***	–0.021	0.095
Culture	(0.115)	(0.108)	(0.160)	(0.135)	(0.283)	(0.306)
Observations	25,617	27,210	24,645	24,840	11,973	12,031
Partial R ²	0.20	0.21	0.16	0.13	0.27	0.13
Region f.e.	Yes	Yes	Yes	Yes	Yes	Yes
Year-Wave f.e.	Yes	Yes	Yes	Yes	Yes	Yes

Source: Author’s calculations on Gallup Data. Std. errors in parentheses. Std. errors are clustered at the country level. The dependent variable is an intention to migrate dummy. Col. (1) and (2): European destinations that are member states of the OECD; Col. (2) and (3): US and Canada; Col. (5) and (6) Turkey. Gend = gender; Rel = religiosity; f.e. = fixed effect.

*p < 0.1. **p < 0.05. ***p < 0.01.

We now focus on migration aspirations to OECD destination countries and test whether the intensity of cultural selection varies by group of individuals. Kaestner and Malamud (2014) note that individual characteristics determine migrants and non-migrants in a Mexican context. We run logit regressions, using different subsamples of individuals. We distinguish between gender and education groups (men, men high educated, men low educated), age groups (between 15–30, 31–45, 46–65), religious groups (Muslims, Christians, other confessions), gender and marital status (single men, single women, married men, married women), place of residence (rural area, town/small village, city/suburb), and employment status (unemployed, employed, out of the labor force). Table 6 shows the results for gender-egalitarian views. We identify a positive correlation between gender-egalitarian views and intentions to emigrate for individuals aged 15 to 30 (i.e., individuals exhibiting the greatest probability to realize their aspirations), for single women, and for people living in rural areas. The coefficient is usually significant at the 5 percent threshold. On the contrary, the effect is insignificant for other groups.

Table 6.

Gender Views and Migration Aspirations towards OECD Countries — Subsample Analysis.

Subsamples	Benchmark	Gender & Education			Age
Subsamples	(1)	(2) Men	(3) Men HS	(4) Men LS	(5) 15–30	(6) 31–45	(7) 46–65
Gender Att.	0.142	0.205	0.226	0.167	0.258**	–0.213	0.351
Gender Att.	(0.103)	(0.149)	(0.201)	(0.190)	(0.113)	(0.188)	(0.268)
Observations	28,937	14,425	5,312	8,678	13,586	8,657	4,832
Partial R ²	0.17	0.16	0.18	0.15	0.18	0.16	0.14
	Religion			Gender & Marital Status (S, M)
	(8) Muslim	(9) Christian	(10) Other	(11) Men (S)	(12) Men (M)	(13) Women (S)	(14) Women (M)
Gender Att.	0.198	–0.519	0.104	0.269	0.098	0.447***	–0.148
Gender Att.	(0.142)	(0.350)	(0.104)	(0.194)	(0.156)	(0.164)	(0.218)
Observations	21,638	2,008	5,204	5,691	8,179	4,900	7,487
Partial R ²	0.19	0.11	0.16	0.15	0.13	0.15	0.17
	Place of Residence			Employment Status
	(15)	(16)	(17)	(18)		(19)	(20)
	Rural	Town	City	Unemp		Employ	Out LF
Gender Att.	0.492**	0.212	–0.055	0.668		0.186	0.139
	(0.211)	(0.166)	(0.127)	(0.408)		(0.156)	(0.174)
Observations	6,946	7,304	12,926	1,347		12,181	9,589
Partial R ²	0.23	0.19	0.15	0.22		0.16	0.20

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. Std. errors are clustered at the country level. In each regression we control for a set of individual level characteristics, region and year fixed effects. The dependent variable is an intention to migrate dummy towards OECD destination countries. The benchmark results are available in column (1). We then perform a subsample analysis on the following characteristics: Gender and Education (Men (2), Men with more than 9 years of education (3), Men with less than 9 years of education (4)); Age groups (15–30 (5), 31–45 (6), and 46–65 (7)); Religion group (Muslims (8), Christians (9), and Others (10)); Gender and Marital status group (Single Men (11), Married Men (12), Single Women (13), and Married Women (14)); Place of Residence (Rural area (15), Town or Village (16), and City or suburbs (17)); and Employment status (unemployed (18), employed (19), and out of the labor force (20)).

*p < 0.1. **p < 0.05. ***p < 0.01.

Results for religiosity are provided in Table 7. They confirm that less religious people are more prone to migrate toward OECD destination countries and that men self-selected more than the total population (0.655 against 0.472). The intensity of self-selection is homogenous across education groups but increases with age. Columns (8) to (10) show that cultural selection is explained by the Muslim population’s behavior. Columns (16) to (19) reveal that self-selection is stronger for employed than for unemployed individuals. Moreover, as for gender attitudes, individuals living in rural areas are more selected also on religiosity. This finding suggests that emigration from rural regions is more likely to make the population left behind less progressive.¹⁴

Table 7.

Religiosity and Migration Aspirations Towards OECD Countries - Subsample Analysis.

Subsamples	Benchmark	Gender & Education			Age
Subsamples	(1)	(2) Men	(3) Men HS	(4) Men LS	(5) 15-30	(6) 31-45	(7) 46-65
Religiosity	0.472***	0.665***	0.688***	0.696***	0.420***	0.522***	0.512***
Religiosity	(0.101)	(0.105)	(0.221)	(0.123)	(0.134)	(0.111)	(0.177)
Observations	30,038	15,031	5,192	9,515	13,771	9,260	5,374
Partial R ²	0.18	0.17	0.22	0.15	0.16	0.16	0.14
	Religion			Gender & Marital Status (S, M)
	(8) Muslim	(9) Christian	(10) Other	(11) Men (S)	(12) Men (M)	(13) Women (S)	(14) Women (M)
Religiosity	0.518***	0.180	0.511	0.604***	0.757***	0.122	0.380*
Religiosity	(0.110)	(0.164)	(0.327)	(0.123)	(0.147)	(0.215)	(0.209)
Observations	22,421	1,952	5,487	6,171	8,515	5,441	7,652
Partial R ²	0.18	0.10	0.13	0.12	0.14	0.16	0.15
	Place of Residence			Employment Status
	(15) Rural	(16) Town	(17) City	(18) Unemp		(19) Employ	(20) Out LF
Religiosity	0.555**	0.272*	0.397***	0.306		0.433***	0.305
Religiosity	(0.227)	(0.162)	(0.111)	(0.411)		(0.126)	(0.199)
Observations	6,729	5,619	12,733	874		8,389	6,862
Partial R ²	0.24	0.19	0.15	0.22		0.17	0.24

Source: Author’s calculations on Gallup Data.

*p < 0.1. **p < 0.05. ***p < 0.01.

4.2.2. Robustness by using matched samples

We now investigate whether our results are driven by differences in the composition of the samples of aspiring migrants and non-migrants and whether the threat of biased estimation due to unobservables can be mitigated. In line with comparisons between treated and control groups, we use the Mahalanobis Metric Matching technique to identify samples of aspiring migrants and non-migrants that are balanced in terms of covariates. We run logit regressions, using a sample of individuals that are similar in terms of observed covariates (apart from their cultural traits) and potentially similar in terms of unobservables if the latter correlate with observed characteristics. The matching procedure minimizes the Mahalanobis metric. For each covariate x, we compute the normalized difference:

Δ_{x} = ({\bar{x}}_{M i g} - {\bar{x}}_{N o n}) {(\frac{s_{x, M i g}^{2} + s_{x, N o n}^{2}}{2})}^{- 1 / 2},

where the difference between the mean value of the covariate for aspiring migrants and non-migrants, ${\bar{x}}_{M i g} - {\bar{x}}_{N o n}$ , is divided by the mean of the standard deviations of the covariate over the whole sample ( $s_{x, M i g}$ and $s_{x, N o n}$ ).

Results of the matching technique are described in Table A13 in the supplemental appendix. For each sample, we report the difference in terms of covariates before and after the matching procedure. Before matching, the distribution of covariates is unbalanced for both samples; differences in characteristics are always statistically different from zero. By contrast, the matching technique allows us to generate a matched sample exhibiting a balanced distribution of covariates. After matching, the only variable along which aspiring migrants and non-migrants exhibit statistically different outcomes is the level of income per household member.

Table 8 provides the results of the fixed-effect logit regressions, using the matched samples; they can be easily compared with those of Table 3 for the non-matched samples. All conclusions of the benchmark regressions hold when using the matched samples. Columns (1) and (2) confirm that aspiring migrants self-select in terms of religiosity but not along gender-egalitarian views. Column (4) and (6) confirm that the results are driven by migration aspirations to OECD destination countries only. The coefficient of religiosity equals 0.423 and is significant at the 1 percent threshold; it is almost identical to that of Table 3. Increasing the indicator of religiosity by one standard deviation raises the desire to emigrate by 13.4 percentage points.¹⁵ The slight decrease in the estimated coefficient suggests that the matching technique allows us to reduce an upward bias due to omitted variables.

Table 8.

Analysis on the Matched Sample.

Culture Destination	(1)	(2)	(3)	(4)	(5)	(6)
Culture Destination	Gend All	Rel All	Gend OECD	Rel OECD	Gend Non-OECD	Rel Non-OECD
Culture	0.078	0.333***	0.125	0.423***	–0.014	0.188
Culture	(0.129)	(0.107)	(0.135)	(0.115)	(0.160)	(0.185)
Observations	11,184	12,100	8,594	9,442	8,145	8,320
Partial R ²	0.04	0.04	0.09	0.08	0.10	0.10
Region f.e.	Yes	Yes	Yes	Yes	Yes	Yes
Year-Wave f.e.	Yes	Yes	Yes	Yes	Yes	Yes

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. Std. errors are clustered at the country level. The dependent variable is an intention to migrate dummy. Gend = gender; Rel = religiosity; f.e. = fixed effect. OECD destinations: same as Table 3.

*p < 0.1. **p < 0.05. ***p < 0.01.

4.2.3. Heterogeneity across countries

We now explore whether cultural selection varies across countries and with country-specific characteristics. The set of country characteristics includes the shares of Sunnis and Shiites among the Muslim population, the log of GDP per capita, two indicators of institutional quality, and the share of native citizens from the same origin country living in an OECD member state, as a proxy of network size abroad. Country-specific coefficients of religiosity and gender-egalitarian attitudes against the above-mentioned country characteristics are plotted in Figures A3 and A4 in the supplemental appendix. On average, only the shares of Sunni/Shiites and the control of corruption are significantly correlated.

To generalize this descriptive analysis of correlations, we ran regressions accounting for interactions between country characteristics and cultural traits. Results are provided in Table 9. Overall, the interaction between progressive views on religiosity and country characteristics is never significant. In addition, the effect of religiosity remains significant in all specifications. Hence, our conclusion that aspiring migrants from virtually all MENA countries self-select along religiosity levels is highly robust. For gender-egalitarian attitudes, the results in Column (1) suggest that aspiring migrants from countries with a Sunni majority have less gender-egalitarian views. This finding is confirmed in Column (2), which reports a positive coefficient for the interaction between cultural traits and share of Shiites. Column (3) shows that selection along gender-egalitarian views also becomes significant when controlling for a measure of control of corruption in the origin country, suggesting that improving institutions reduces self-selection on gender-egalitarian attitudes.

Table 9.

Adding Interactions with Macroeconomic Variables.

Destination Interaction	(1)	(2)	(3)	(4)	(5)	(6)
Destination Interaction	OECD Sunnis	OECD Shiites	OECD Cont. Corr.	OECD Rule Law	OECD ln GDP	OECD Net.
Religiosity	0.589**	0.356**	0.490**	0.773**	0.373**	0.461***
Religiosity	(0.242)	(0.138)	(0.219)	(0.333)	(0.178)	(0.151)
Interaction	–0.002	0.006	–0.150	–0.797	0.000	0.006
Interaction	(0.003)	(0.004)	(0.819)	(0.687)	(0.000)	(0.017)
Observations	30,038	30,038	27,141	27,141	30,038	30,038
Partial R ²	0.18	0.18	0.18	0.18	0.18	0.18
Gender Att.	0.764***	0.024	0.633**	0.639	0.140	0.233*
Gender Att.	(0.183)	(0.104)	(0.313)	(0.392)	(0.155)	(0.134)
Interaction	–0.009***	0.009***	–1.918*	–1.199	0.000	–0.029
Interaction	(0.002)	(0.003)	(1.029)	(0.827)	(0.000)	(0.020)
Observations	28,937	28,937	27,469	27,469	28,937	28,937
Partial R ²	0.17	0.17	0.18	0.18	0.17	0.17

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. The dependent variable is an intention to migrate dummy. We control for gender, network, age cohort, level of education, marital status, presence of children, income, income squared, year-wave and region fixed effects (f.e.). Std. errors are clustered at the country level. OECD destinations: same as Table 3.

*p < 0.1. **p < 0.05. ***p < 0.01.

4.2.4. Effect of the Arab Spring

We finally explore whether the link between cultural traits and migration was affected by the Arab Spring. Since most questions on religiosity and gender-egalitarian views were asked from 2007 to 2011, we use the only proxy for religiosity available in all GWP waves (2007–2016): Is religion an important part of your daily life? Responses to this question are highly correlated with the index of religiosity resulting from our PCA, as is shown in Table A6 in the supplemental appendix.

We run fixed-effect logit regressions accounting for the interactions between a post–Arab Spring dummy (equal to 1 for the years 2011 to 2016) and religiosity. Results for the full sample are provided in Column (1) of Table 10.

Table 10.

Cultural Selection before and after the Arab Spring.

	(1)	(2)	(3)	(4)	(5)	(6)
Destination Sample Countries	OECD Full All	OECD Full Main Ins.	OECD Full Others	OECD Matched All	OECD Matched Main Ins.	OECD Matched Others
Religiosity	0.473***	0.560***	0.409***	0.446***	0.559***	0.396***
	(0.080)	(0.088)	(0.116)	(0.062)	(0.108)	(0.082)
Interaction	–0.110	–0.475**	0.081	–0.149	–0.425*	0.010
Interaction	(0.177)	(0.215)	(0.203)	(0.173)	(0.244)	(0.228)
Observations	100,544	34,914	65,630	41,801	14,328	27,473
Region f.e.	Yes	Yes	Yes	Yes	Yes	Yes
Year-Wave f.e.	Yes	Yes	Yes	Yes	Yes	Yes
Partial R ²	0.139	0.170	0.127	0.142	0.173	0.131

Source: Author’s calculations on Gallup Data.

Std. errors in parentheses. The dependent variable is an intention to migrate dummy. We control for gender, network, age cohort, level of education, marital status, presence of income squared, the interaction terms with individual controls and Arab Spring, children, income, and region fixed effects (f.e.). Std. errors are clustered at the country level. Sample legend: Full Sample (FS); Matched on After Arab Spring (Matched). OECD destinations: same as Table 3.

*p < 0.1. **p < 0.05. ***p < 0.01.

The Arab Spring affected MENA countries differently. In countries like Tunisia, Egypt, and Yemen, the government was overthrown after turmoil and riots. Algeria experienced two years of major protests, while the Arab Spring led Syria into a civil war, with more than 400,000 casualties and high numbers of refugees in neighboring countries (Haas and Lesch 2016). The uprisings in these countries were triggered by the high level of unemployment (particularly for the young), the persistence of economic inequality, and political authoritarianism. We refer to these five countries (Algeria, Egypt, Syria, Tunisia, and Yemen) as the group of Main Insurgents. We then run a subsample analysis of the Arab Spring’s effect on the Main Insurgents and on the other countries. Intuitively, we expect a greater effect for the Main Insurgents. Columns (1) to (3) use the full sample; Columns (4) to (6) use matched samples, with matching based on the post-Arab Spring period. Results of the matching are available in Table A14 in the supplemental appendix and minimize possible biases due to the unbalanced distribution of covariates and unobservables.

In all specifications, selection by religiosity is always positive and significant. Although the Arab Spring did not affect the intensity of cultural selection in the less-affected countries, it reduced it in the Main Insurgent countries. The latter findings are highly robust to the use of the matching technique. In other words, the Arab Spring increased aspiring migrants’ relative religiosity. The data do not allow us to investigate whether the degree of cultural selection changed after the recent conflicts and political unrest in the Middle East. Given similarities between the Arab Spring period and the 2015 wave of asylum-seekers, however, it is plausible that the latter is less culturally selected than previous migration waves.

5. Concluding Remarks

This article uses a unique database on migration aspirations, opinions, and beliefs in MENA countries to test whether migrants positively self-select along cultural traits. We conduct fixed-effect logit regressions, using full or matched samples of aspiring migrants and non-migrants. We find that aspiring migrants in general and those who have concrete migration plans to migrate in particular are culturally selected and that this selection on cultural traits depends on the type of preferred destination. Intended migrants to OECD, high-income countries exhibit lower level of religiosity. This result is robust across gender, age groups, and education levels. As far as attitude toward women’s rights is concerned, aspiring migrants have more gender-egalitarian views when they are from 15 to 30 years old, when they are single women, or when they originate from countries with a Sunni minority. Finally, we find a robust effect of the Arab Spring on the intensity of cultural selection only in countries highly impacted by the Arab Spring. In these countries, the Arab Spring decreased the degree of cultural selection. These results have implications from the viewpoint of both origin and destination countries.

From the viewpoint of destination countries, selection on religiosity and gender-egalitarian attitudes implies that the cultural distance between migrants and host-country citizens is smaller than that between the countries’ overall populations. On the one hand, existing empirical studies suggest that smaller cultural distance facilitates immigrants’ economic and socio-demographic outcomes at destination (Pendakur and Pendakur 2005; Fernandez and Fogli 2009; Alesina and Giuliano 2010; Battu and Zenou 2010; Casey and Dustmann 2010; Bisin et al. 2011; Islam and Raschky 2013; Lundborg 2013). As far as public opinions are concerned, selective migration from MENA to high-income OECD countries should be less of a concern from the viewpoint of OECD member states. Informing public opinion in this regard might influence attitudes toward immigration and discrimination practices. On the other hand, a key finding of our analysis is that the aggregate effect of cultural selection should not be overestimated.

Figure 3(a) compares the average level of religiosity of OECD natives (vertical line at 0.623) with the average level of intended stayers in MENA countries, of intended migrants toward OECD destinations, and of current migrants from MENA countries in OECD countries (extracted from the GWP database). When considering all respondents, the religiosity index of intended migrants (0.291) is 13 percent higher than that of intended stayers (0.257). Remember, zero corresponds to the maximal religiosity level. Hence, self-selection along religiosity levels reduces the gap between MENA and OECD countries by 9 percent only.¹⁶ Similar findings appear when comparing young intended migrants and non-migrants. When considering highly educated respondents, the average religiosity index is slightly greater (0.301), but cultural differences between intended migrants and non-migrants become negligible. The GWP data also enable us to compare actual migrants from MENA countries (those who have already migrated) with OECD native citizens. Figure 3(b) shows that actual migrants’ religiosity index is much closer to that of natives, especially for older migrants and the highly educated. The gap between actual and intended migrants can be due to several reasons: (1) cultural selection in the realization of migration aspirations (in line with the results of Table 4), (2) a gradual decline in cultural selection over time (in line with our findings about the Arab Spring, see Table 10), (3) a sign of cultural assimilation abroad, or (4) estimation biases due to the survey’s under-representation of the foreign-born population (the GWP database only includes a small sample of 282 immigrants from the MENA in the OECD countries).

Figure 3.

Average cultural indices by group of respondents. (a) Religiosity index (migrants vs. non-migrants). (b) Gender-Egal. index (migrants vs. non-migrants). (c) cultural traits in the MENA.

Another important finding is that the Arab Spring decreased the strength of cultural selection. Although the external validity of our results cannot be assessed, this result suggests that conflicts and political turmoil in developing countries are likely to increase not only migration pressures to OECD countries but also the average cultural distance between immigrants and host-country citizens. Migration peaks and cultural distance are two sources of tension and negative attitudes toward immigrants in OECD member states (see the effect on voting, Mendez and Cutillas 2014; Halla, Wagner, and Zweimüller 2017). Hence, we see cultural selection as another argument justifying coordinated interventions to bring peace, security, and stability in countries at risk.

Finally, from the viewpoint of the home country, the selection patterns documented in this analysis can lead to a situation in which the distribution of cultural traits in the population left behind skews toward more religiosity and less gender-egalitarian attitudes, with potential implications for modernization, growth, and democracy. On this basis, it could be argued that emigration to OECD countries in particular should be combated if home-country governments hope to achieve higher levels of economic development. Our results, however, do not support this view. Figure 3(c) compares the observed average indices of religiosity and gender-egalitarian view of the current MENA population (in blue) with those obtained if all intended migrants left their country (in red) or all young intended migrants left (in green). Given the proportion of intended migrants toward OECD countries (12.3% on average) and the small cultural differences between groups, the average cultural traits of the population left behind hardly changes under these two counterfactuals. In sum, despite cultural selection, emigration from MENA countries is unlikely to induce negative effects on modernization, growth, and democracy. On the contrary, emigration toward OECD countries could even reverse the selection effect on average cultural traits if migrants abroad transfer more progressive norms and beliefs to their home country, as argued in Rapoport, Sardoschau, and Silve (2018) and Bouoiyour and Miftah (2017). This latter effect, still understudied in the literature, could even reduce cultural distance between origin and destination countries. Our article shows the cultural selection on religious and gender egalitarian attitudes of aspiring migrants from MENA countries; we hope that this study will stimulate further research on the link between migration and culture.

Supplemental Material

Supplemental Material, MRX849011_Supplemental_Material - Do Emigrants Self-Select Along Cultural Traits? Evidence from the MENA Countries

Supplemental Material, MRX849011_Supplemental_Material for Do Emigrants Self-Select Along Cultural Traits? Evidence from the MENA Countries by Frédéric Docquier, Aysit Tansel and Riccardo Turati in International Migration Review

Footnotes

Authors’ Note

The contents of this document are the sole responsibility of the authors and can under no circumstances be regarded as reflecting the position of the European Union.

Acknowledgments

The authors are grateful to Michel Beine, Bastien Chabé-Ferret, William Parienté, Christopher Parsons, Pedro Vicente, the editor, and anonymous referees for helpful comments. This article has also benefited from discussions at the FEMISE Annual Conference (Casablanca, May 2017), at the Workshop on “Migration and Conflicts” (Louvain-la-Neuve, June 2017), at the CEMIR Junior Economist Workshop on Migration Research (Munich, June 2017), and at the EDP Jamboree (Bonn, September 2017).

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 the following financial support for the research, authorship, and/or publication of this article: This article has been produced with the financial assistance of the European Union within the context of the European Commission-FEMISE project on “Support to Economic Research, Studies and Dialogue of the Euro-Mediterranean Partnership” (Agreement No. FEM42-03).

Supplemental Material

The supplemental appendix is available in the online version of this article at .

Notes

References

Adsera

Pytlikova

. 2015. “The Role of Language in Shaping International Migration.” Economic Journal 125(586):F49–F81.

Alesina

Giuliano

. 2010. “The Power of the Family.” Journal of Economic Growth 15(2):93–125.

Alesina

Harnoss

Rapoport

. 2016. “Birthplace Diversity and Economic Prosperity.” Journal of Economic Growth 21(101):1–38.

Alexander

Welzel

. 2011. “Islam and Patriarchy: How Robust Is Muslim Support for Patriarchal Values?” International Review of Sociology 21(2):249–76.

Alexander

J. C.

2008. “Clifford Geertz and the Strong Program: The Human Sciences and Cultural Sociology.” Cultural Sociology 2:157.

Algan

Cahuc

. 2010. “On the Origin of Gender Roles: Women and Plogh.” Quarterly Journal of Economics 128(2):469–530.

Altonji

J. G.

Elder

T. E.

Taber

C. R.

. 2005. “Selection on Observed and Unobserved Variables: Assessing the Effectiveness of Catholic Schools.” Journal of Political Economy 113(1):151–84.

Battu

Zenou

. 2010. “Oppositional Identities and Employment for Ethnic Minorities: Evidence from England.” Economic Journal 120(542):F52–F71.

Baxter

Kane

E. W.

. 1995. “Dependence and Independence: A Cross-National Analysis of Gender Inequality and Gender Attitudes.” Gender & Society 9(2):193–215.

10.

Becerra

2012. “The Impact of Anti-Immigration Policies and Perceived Discrimination in the United States on Migration Intentions among Mexican Adolescents.” International Migration 50(4):20–32.

11.

Becerra

Gurrola

Ayon

Androff

Krysik

Gerdes

Moya-Salas

Segal

. 2010. “Poverty and Other Factors Affecting Migration Intentions among Adolescents in Mexico.” Journal of Poverty 14:1–16.

12.

Belot

Ederveen

. 2012. “Cultural Barriers in Migration between OECD Countries.” Journal of Population Economics 25(3):1077–1105.

13.

Belot

Hatton

T. J.

. 2012. “Immigrant Selection in the OECD.” Scandinavian Journal of Economics 114(4):1105–28.

14.

Benabou

Ticchi

Vindigni

. 2015. “Religion and Innovation.” American Economic Review 105(5):346–51.

15.

Bergh

2007. “Gender Attitudes and Modernization Processes.” International Journal of Public Opinion Research 19(1):5–23.

16.

Berlinschi

Harutyunyan

. 2016. “Do Migrants Think Differently? Evidence from Eastern European and Post-Soviet States.” Licos Discussion Paper.

17.

Bertoli

Ruyssen

. 2018. “Networks and Migrants Intended Destination.” Journal of Economic Geography 18(4):705–28.

18.

Bisin

Pataccini

Verdier

Zenou

. 2011. “Ethnic Identity and Labour Market Outcomes of Immigrants in Europe.” Economic Policy 25(65):57–92.

19.

Bisin

Verdier

1998. “The Economics of Cultural Transmission and the Dynamics of Preferences.” Journal of Economic Theory 97:298–319.

20.

Bouoiyour

Miftah

. 2017. “Do Migrants Transfer Political and Cultural Norms to Their Origin Country? Some Evidence from Some Arab Countries.” Economic Research Forum (ERF) Working Paper No. 1098.

21.

Card

Dustmann

Preston

. 2006. “Understanding Attitudes to Immigration: The Migration and Minority Module of the First European Social Survey.” In Measuring Attitudes Cross-Nationally – Lessons from the European Social Survey, edited by Jowell

Fitzgerald

Eva

. Thousand Oaks, CA: SAGE Publications.

22.

Card

Dustmann

Preston

. 2012. “Immigration, Wages, and Compositional Amenities.” Journal of the European Economic Association 10(1):78–119.

23.

Casey

Dustmann

. 2010. “Immigrants’ Identity, Economic Outcomes and the Transmission of Identity across Generations.” Economic Journal 120(2):F31–F51.

24.

Chase

2014. “Religiosity, Corruption and Economic Growth.” Quarterly Review of Business Disciplines 1(3):263–72.

25.

Chort

2014. “Mexican Migrants to the US: What do Unrealized Migration Intentions Tell Us about Gender Inequalities?” World Development 59:535–52.

26.

Collier

2017. “Culture, Politics, and Economic Development.” Annual Review of Political Science 20:111–25.

27.

Dao

T. H.

Docquier

Parsons

Peri

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

28.

De Jong

G. F.

Richter

Isarabhakdi

. 1996. “Gender, Values and Intention to Move in Rural Thailand.” International Migration Review 30(3):748–70.

29.

Desmet

Le Breton

Ortuño-Ortín

Weber

. 2011. “The Stability and Breakup of Nations: A Quantitative Analysis.” Journal of Economic Growth 16(3):183–213.

30.

Dibeh

Fakih

Marrouch

. 2018. “Labor Market and Institutional Drivers of Youth Irregular Migration in the Middle East and North Africa Region.” Journal of Industrial Relations, 0022185618788085.

31.

Di Tella

MacCulloch

. 2014. “Culture, Beliefs and Economic Performance.” Motu Working Paper No. 14-06.

32.

Diwan

Tzannatos

Akin

. 2018. “Debunking the Myth: Economic Values in the Arab World through the Prism of Opinion Polls.” Middle East Development Journal 10(1):31–63.

33.

Docquier

Lohest

Marfouk

. 2007. “Brain Drain in Developing Countries.” World Bank Economic Review 21:193–218.

34.

Docquier

Machado

Sekkat

. 2015. “Efficiency Gains from Liberalizing Labor Mobility.” Scandinavian Journal of Economics 117(2):303–46.

35.

Docquier

Peri

Ruyssen

. 2014. “The Cross-Country Determinants of Potential and Actual Migration.” International Migration Review 48:37–99.

36.

Drinkwater

Ingram

. 2009. “How Different Are the British in Their Willingness to Move? Evidence from International Social Survey Data.” Regional Studies 43(2):287–303.

37.

Duflo

2012. “Women’s Empowerment and Economic Development.” Journal of Economic Literature 50(4):1051–79.

38.

Dustmann

Okatenko

. 2014. “Out-Migration, Wealth Constraints, and the Quality of Local Amenities.” Journal of Development Economics 110:52–63.

39.

El Mikawy

Mohieddin

El Ashmaouy

. 2017. “Egypt: The Protracted Transition from Authoritarianism to Democracy and Social Justice.” In Democratic Transitions in the Arab World, edited by Elbadawi

Ibrahim

Makdisi

Samir

. Cambridge: Cambridge University Press.

40.

Falck

Heblich

Lameli

Suedekum

. 2012. “Dialects, Cultural Identity and Economic Exchange.” Journal of Urban Economics 72:225–39.

41.

Falck

Lameli

Ruhouse

. 2018. “Cultural Biases in Migration: Estimating Nonmonetary Migration Costs.” Papers in Regional Science 97(2):411–38.

42.

Falck

Lameli

Rotondi

. 2016. “The Less Extreme, the More You Leave: Radical Islam and Willingness to Migrate.” World Development 88:122–33.

43.

Felbermayr

Toubal

. 2010. “Cultural Proximity and Trade.” European Economic Review 54(2):279–91.

44.

Fernandez

Fogli

. 2009. “Culture: An Empirical Investigation of Beliefs, Work and Fertility.” American Economic Journal: Macroeconomics 1(1):146–77.

45.

Gallup. 2010. “Measuring the State of Muslim-West Relations – Assessing the ‘New Beginning.’” Abu Dhabi Gallup Center.

46.

Gill

1999. “Government Regulation, Social Anomie and Protestant Growth in Latin America: A Cross-National Analysis.” Rationality and Society 11:287–316.

47.

Gorodnichenko

Roland

. 2017. “Culture, Institutions and the Wealth of Nations.” Review of Economics and Statistics 99(3):402–16.

48.

Grogger

Hanson

G. H.

. 2011. “Income Maximization and the Selection and Sorting of International Migrants.” Journal of Development Economics 95(1):42–57.

49.

Guiso

Sapienza

Zingales

. 2006. “Does Culture Affect Economic Outcomes?” Journal of Economic Perspectives 20(2):23–48.

50.

Guiso

Sapienza

Zingales

. 2009. “Cultural Biases in Economic Exchange?” Quarterly Journal of Economics 124(3):1095–1131.

51.

Haas

M. L.

Lesch

, eds. 2016. The Arab Spring: The Hope and Reality of the Uprisings. Boulder, CO: Westview Press.

52.

Halla

Wagner

A. F.

Zweimüller

. 2017. “Immigration and Voting for the Far Right.” Journal of the European Economic Association 15(6):1341–85.

53.

Hammel

E. A.

1990. “A Theory of Culture for Demography.” Population and Development Review 16(3):455–85.

54.

Hoffman

Marsiglia

Ayers

. 2015. “Religiosity and Migration Aspirations among Mexican Youth.” Journal of International Migration and Integration 16:173–86.

55.

Ichino

Maggi

. 2000. “Work Environment and Individual Background: Explaining Regional Shirking Differentials in a Large Italian Firm.” Quarterly Journal of Economics 115(3):1057–90.

56.

Imbens

Rubin

. 2012. Casual Inference for Statistic, Social and Biomedical Sciences. Cambridge: Cambridge University Press.

57.

Islam

Raschky

P. A.

. 2013. “Cultural Distance, Immigrants’ Identity, and Labour Market Outcomes.” Discussion Paper 21/13. Monash University.

58.

Just

Monaghan

. 2000. Social and Cultural Anthropology: A Very Short Introduction. Oxford: Oxford University Press.

59.

Kaestner

Malamud

. 2014. “Self-Selection and International Migration: New Evidence from Mexico.” Review of Economics and Statistics 96(1):78–91.

60.

Karouby

2014. “The Role of the Maronite Diaspora.” In Non-State Actors in the Middle East: Factors for Peace and Democracy, edited by Golan

Salem

. London: Routledge, Taylor and Francis Group.

61.

Knack

Keefer

. 1997. “Does Social Capital Have an Economic Payoff? A Cross-Country Investigation.” Quarterly Journal of Economics 112(4):1251–88.

62.

Kostenko

Ponarin

Shteiwi

Strebkova

. 2017. “Historical Legacies and Gender Attitudes in the Middle East.” Economic Research Forum (ERF) Working Paper No. 1105.

63.

Krieger

Renner

Ruhose

. 2018. “Long-Term Relatedness between Countries and International Migrant Selection.” Journal of International Economics 113:35–54.

64.

Kuran

2012. The Long Divergence: How Islamic Law Held Back the Middle East. Princeton, NJ: Princeton University Press.

65.

Lundborg

2013. “Refugees’ Employment Integration in Sweden: Cultural Distance and Labor Market Performance.” Review of International Economics 21(2):219–32.

66.

Manchin

Orazbayev

. 2014. “Desire to Migrate Internationally and Locally and the Importance of Satisfaction with Amenities.” Paper presented at the FIW-wiiw Seminars in International Economics, April 10, Vienna.

67.

Mayda

A. M.

2006. “Who Is Against Migration? A Cross-Country Investigation of Individual Attitudes toward Immigrants.” Review of Economics and Statistics 88(3):510–30.

68.

Mendez

Cutillas

I. M.

. 2014. “Has Immigration Affected Spanish Presidential Elections Results?” Journal of Population Economics 27(1):135–71.

69.

Murtin

Wacziarg

. 2014. “The Democratic Transition.” Journal of Economic Growth 19(2):141–81.

70.

Myers

2000. “The Impact of Religious Involvement on Migration.” Social Forces 79:755–83.

71.

Norris

Inglehart

. 2011. Sacred and Secular. Religion and Politics Worldwide. Cambridge: Cambridge University Press.

72.

Oster

2017. “Unobservable Selection and Coefficient Stability: Theory and Evidence.” Journal of Business & Economic Statistics 40(1):1–18.

73.

Ottaviano

Peri

. 2006. “The Economic Value of Cultural Diversity: Evidence from US Cities.” Journal of Economic Geography 6(1):9–44.

74.

Pendakur

. 2005. “Ethnic Identity and the Labour Market.” Working Paper, Department of Economics, Simon Fraser University.

75.

Piketty

1995. “Social Mobility and Redistributive Politics.” Quarterly Journal of Economics 110(3):551–84.

76.

Price

2015. “How National Structures Shape Attitudes toward Women’s Right to Employment in the Middle East.” International Journal of Comparative Sociology 56(6):408–32.

77.

Rapoport

Sardoschau

Silve

. 2018. “International Migration and Cultural Convergence.” Manuscript.

78.

Richards

Waterbury

Cammett

Diwan

. 2014. A Political Economy of the Middle East. 3rd ed. Boulder, CO: Westview Press.

79.

Ruyssen

Salomone

. 2018. “Female Migration: A Way out of Discrimination?” Journal of Development Economics 130:224–41.

80.

Shenkar

2012. “Cultural Distance Revisited: Towards a More Rigorous Conceptualization and Measurement.” Journal of International Business Studies 43(1):1–11.

81.

Shweder

R. A.

Levine

R. A.

. 1984. Culture Theory: Essays on Mind, Self and Emotion. New York, NY: Cambridge University Press.

82.

Spolaore

Wacziarg

. 2012. “Long-Term Barriers to the International Diffusion of Innovations.” In NBER International Seminar on Macroeconomics 2011, edited by Frankel

Pissarides

, chapter 1, 11–46. Cambridge, MA: National Bureau of Economic Research.

83.

Spolaore

Wacziarg

. 2013. “How Deep Are the Roots of Economic Development?” Journal of Economic Literature 51(2):325–69.

84.

Spolaore

Wacziarg

. 2016. “Ancestry, Language and Culture.” In The Palgrave Handbook of Economics and Language, edited by Ginsburgh

Victor

Weber

Shlomo

, chapter 6, pp. 174–211. London: Palgrave Macmillan.

85.

Stetter

2008. World Society and the Middle East: Reconstructions in Regional Politics. New York, NY: Palgrave Macmillan.

86.

Tabellini

2010. “Culture and Institutions: Economic Development in the Regions of Europe.” Journal of European Economic Association 8(4):677–716.

87.

Tzannatos

1999. “Women and Labor Market Changes in the Global Economy: Growth Helps, Inequalities Hurt and Public Policy Matters.” World Development 27(3):551–69.

88.

UN. 2005. Household Sample Surveys in Developing and Transition Countries. New York, NY: United Nations.

89.

UN. 2015. Human Development Report 2015: Work for Human Development. New York, NY: United Nations.

90.

UNDP (United Nations Development Programme). 2000. Azerbaijan Human Development Report, 2000. Baku: Azerbaijan Republic.

91.

Weber

1946. “The Social Psychology of World Religions.” In From Max to Weber, edited by Gerth

H. H.

Wright Mills

, 267–301. New York, NY: Oxford University Press.

92.

Wood

C. H.

Gibson

Ribeiro

Hamsho-Diaz

. 2010. “Crime Victimization in Latin America and Intentions to Migrate to the United States.” International Migration Review 44(1):3–24.

93.

Young

H. P.

2008. “Social Norms.” In The New Palgrave Dictionary of Economics, 2nd ed. London: Palgrave.

94.

Zelinsky

1971. “The Hypothesis of the Mobility Transition.” Geographical Review 91(2):219–49.

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.

0.00 MB

0.58 MB