Raising Awareness About the Risk of Irregular Migration: Quasi‐Experimental Evidence from Guinea

Tjaden, Jasper; Gninafon, Horace Article — Published Version Raising Awareness About the Risk of Irregular Migration: Quasi‐Experimental Evidence from Guinea Population and Development Review Provided in Cooperation with: John Wiley & Sons Suggested Citation: Tjaden, Jasper; Gninafon, Horace (2022) : Raising Awareness About the Risk of Irregular Migration: Quasi‐Experimental Evidence from Guinea, Population and Development Review, ISSN 1728-4457, Wiley, Hoboken, NJ, Vol. 48, Iss. 3, pp. 745-766, https://doi.org/10.1111/padr.12468 This Version is available at: https://hdl.handle.net/10419/266667 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Raising Awareness About the Risk of Irregular Migration: Quasi-Experimental Evidence from Guinea JASPER TJADEN AND HORACE GNINAFON In response to mounting evidence of harm inflicted on irregular migrants along their journeys from West Africa to Europe, international organizations, civil society organizations, and governments have scaled up campaigns as a tool for raising awareness about the risks of irregular migration. Campaigns aim to counter misinformation by smugglers and facilitate safe migration decisions. Despite the growing number of interventions, there is limited empirical evidence on the impact and effectiveness of such campaigns. Based on a difference-in-difference design, this study investigates the effect of a mobile cinema and community discussion intervention on the perceptions, knowledge, and intentions of potential irregular migrants in Northern Guinea in 2019. The results show that potential migrants who participated in events were significantly more likely to show awareness gains and less likely to report high intentions to migrate irregularly. While the relative importance of risk perceptions and their impact on migration flows remain unclear, the findings provide evidence supporting the assumption that risk awareness can be a relevant factor in the decisionmaking process of potential irregular migrants. While campaigns may be an effective tool in certain contexts, effect sizes highlight the need for policymakers to keep realistic expectations. Introduction This paper investigates the effect of awareness about the risks associated with irregular migration on intentions, perceptions, and knowledge among potential irregular migrants in rural areas of Northern Guinea.1While information campaigns are commonplace in other fields of development (Bongaarts and Casterline 2013; Battaglia and Pallarés 2020), rigorous evidence in the field of migration is still lacking, contributing to the overall “efficacy gap” in migration policy research (Czaika and de Haas 2013). Jasper Tjaden, Department of Economic and Social Sciences, University of Potsdam, Potsdam, Germany. Horace Gninafon, Department of Economics, Université Laval, Quebec, Canada. Email: jasper[email protected]. POPULATION AND DEVELOPMENT REVIEW 48(3): 745–766 (SEPTEMBER 2022) 745 © 2022 The Authors. Population and Development Review published by Wiley Periodicals LLC on behalf of Population Council. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. 746 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION In response to mounting evidence of harm inflicted on irregular migrants along their journeys from West Africa to Europe (Black 2020; Black et al. 2017; Galos et al. 2017; Mixed Migration Centre 2018; UNHCR 2018b), international organizations, civil society organizations, and governments have turned to campaigns as a tool for raising awareness about the risks of irregular migration in hopes of countering misinformation by smugglers, facilitating safe migration decisions, and ultimately reducing harm.2At the political level, such campaigns have attracted criticism for aiming to deter migration without explicitly stating this objective in campaign documentation. Migration from North, West, and East Africa to Europe has increased since 2013 alongside a surge in mixed migration flows from the Middle East. While the overwhelming majority of migrants arriving in the European Union (EU) since 2015 were nationals from the Syrian Arab Republic, Afghanistan, and Iraq, several African countries were also among the top 10 origins (Maher 2017; IOM 2018; UNHCR 2018a). From January to October 2019, Guinean nationals represented 4.5 percent of all arrivals, 10 percent of arrivals from Africa, and 28 percent from West Africa.3 Despite the growing number of campaigns on the risks of irregular migration, there is limited empirical evidence on their impact and effectiveness (Tjaden, Morgenstern, and Laczko 2018; Browne 2015). In addition to political concerns regarding their objective, scholars have also questioned these campaigns’ effectiveness. Accounts from qualitative research have argued that campaigns rely on wrong assumptions. Migrants may already be informed or ignore information that would run counter to their migration plans (Alpes and Sorensen 2015; Hernández-Carretero and Carling 2012; Mbaye 2014; van Bemmel 2020). A recent review of the policy on information campaigns highlights the many assumptions involved in the causal chain underlying the design of campaigns, yet the empirical evidence—particularly evidence using rigorous research designs with the ability to isolate program effects—remains severely underdeveloped (Schans and Optekamp 2016; Tjaden, Morgernstern, and Laczko 2018). This study attempts to address this gap by investigating the effects of an awareness-raising intervention in Guinea using a difference-in- difference (DiD) approach. The intervention was a mobile cinema caravan which organized awareness-raising events and movie screenings in 31 villages and small towns in Northern Guinea in 2019. In addition, 32 comparable villages were selected for the control group where no campaign took place. The events consisted of “infotainment” movies involving documentary-style testimonies by migrants highlighting risks involved in their own migration journeys followed by a community discussion about migration. The study finds effects on a range of different migration-related outcomes. Participation in the intervention further increased risk perceptions TJADEN AND GNINAFON 747 regarding the journey, reduced intentions to migrate irregularly, increased the perceived costs relating to irregular migration to Europe and marginally increased positive perceptions of economic opportunities at home. Further analyses suggest that effects do not differ by sex, do not wane within a two- to-five-month period after intervention, and increase for those potential migrants with no information at baseline. Overall, the study aims to contribute to the limited evidence based on information and awareness-raising treatments for potential migrants (Shrestha 2019, 2020; Tjaden and Dunsch 2021; Mesplé-Somps and Nilsson 2020). To the best of our knowledge, this study is one of the first (quasiexperimental) studies to evaluate the effectiveness of large-scale campaigns on irregular migration. Most of the available literature has either been limited to laboratory experiments (Bah and Batista 2018; Batista and McKenzie 2021) or observational studies (Mbaye 2014). In addition, this study offers a unique piece of evidence from rural areas where the living conditions of the population are more precarious and interest in irregular migration is high. Studies from rural areas are scarce, given the challenging environment for data collection. Risk awareness and irregular migration from West Africa Many irregular migrants arriving in the EU or stranded in Libya between 2015 and 2019 were from West Africa, including Guinea. According to a study by the Guinean National Institute of Statistics, the international emigration rate increased from 7 percent in 2007 to 30 percent in 2013 (Massandouno and Cissé 2017). According to recent UN estimates for 2018, half a million Guinean nationals lived abroad, approximately 24,000 as recognized refugees and 33,000 as asylum seekers. According to national Afrobarometer surveys for 2016–2018, one in five Guineans depend on money sent home from abroad, and almost one in three have a family member who lives abroad.4 According to data from the United Nations High Commissioner for Refugees (UNHCR) and the International Organization for Migration (IOM), Guinean nationals represented approximately four to five percent of all arrivals in the EU in 2018, 10 percent of arrivals from Africa, and 28 percent from West Africa.5While the overall percentage of Guinean migrant arrivals in Europe in more recent years appears low, Guinean nationals are consistently among the 10 largest migrant groups in terms of arrivals in Italy and among the three largest migration groups in terms of arrivals in Spain.6In 2018, the numbers of asylum claims from Guinea (3,000) launched in the European Union reached almost half of the number of asylum claims from Nigerian nationals despite the fact that Nigeria’s population is almost 16 times larger. Guinean migrants travel the land route via Mali and Niger to ports in Morocco or Libya before attempting to cross 748 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION the Mediterranean Sea. Alternatively, some migrants travel to Morocco through Senegal and Mauritania or, in few cases, take the sea route to the Canary Islands. A number of reports have indicated that some irregular migrants may embark on their journeys with limited awareness regarding the risks involved as well as the context in which migration to Europe takes place—both in transit countries and at destination. As a result, many may underestimate the risk (IOM 2017; European Commission 2018; Regional Mixed Migration Secretariat 2014; Gillespie et al. 2016; Foran and Iacucci 2017). According to a recent report by the United Nations Development Programme (UNDP 2020), only approximately half of recent African migrants in Europe had expected to encounter danger during their journey, yet 93 percent of them reported that they had actually experienced real danger. A study based on a sample of transit and returning migrants in Niger in 2016 showed that half (56 percent) declared that they did not collect information about migration before departing (IOM 2017). Over 80 percent of the individuals who provided feedback on their information sources said the information turned out to be false. A recent study showed that almost 43 percent of potential migrants who responded to a survey in Dakar reported that they did not feel well informed about how to migrate to Europe (Dunsch, Tjaden, and Quiviger 2019). Thirty-seven percent of young residents considering leaving Senegal said they were not well informed about the risks associated with migration, and data suggest that they were misinformed about the legal context of migration in terms of visa and international protection eligibility (Dunsch, Tjaden, and Quiviger 2019). Migration is a large business for people smugglers who often spread misinformation for monetary gain. The fees charged to smuggle migrants can differ substantially based on the point of origin, with figures ranging from USD 2,000 to USD 10,000 (Dunsch, Tjaden, and Quiviger 2019; Europol/Interpol 2016). A conservative estimate of the smuggling business along all sea smuggling routes for the year 2016 ranges between USD 320 million and USD 550 million (United Nations Office on Drugs and Crime 2018). There may be an incentive for smugglers to downplay the risks involved in the journey, the likelihood of reaching the destination, and the total costs involved, in order to recruit more customers and increase profits. Based on the assumption that a lack of awareness and misinformation contributed to increased risk of harm along the migration journey and at destination, many EU governments funded information and awareness-raising campaigns to counter narratives spread by smugglers. Many interventions between 2016 and 2019 focused on the risks involved in the journey because of horrifying evidence of harm inflicted on irregular migrants, especially among those stranded in Libya and those attempting to TJADEN AND GNINAFON 749 cross the Mediterranean (Black et al. 2017; Galos et al. 2017; Mixed Migration Centre 2018; UNHCR 2018b). Despite accounts of awareness gaps and information needs, the academic literature is predominantly skeptical toward efforts to provide information on risks. Several authors have argued—mostly based on qualitative studies—that migrants are fully aware of the potential perils but decide to migrate anyway (Alpes and Sorensen 2015; Fiedler 2020; Hernández- Carretero and Carling 2012; Mbaye 2014; van Bemmel 2020). Even when accurate information is provided, migrants may not trust it, only taking into account information from their immediate friend and family networks (Schans and Optekamp 2016). Others question whether migration decisions are rational cost-benefit calculations at all. Prothmann (2018), for example, claims that migration can be influenced by norms of masculinity, where migration can be seen as a means of recovering “lost values” of courage and determination. Ryo (2013) agrees that “the view of would-be migrants as atomistic, utility maximizing opportunists diverts our attention away from the complex and wide-ranging moral systems within which prospective migrants are embedded” (p. 593). In light of those arguments, the lack of robust empirical evidence is striking. The lack of evidence is even more glaring when considering that the role of risk and risk salience in migration decision-making is widely assumed implicitly and forms a key part of standard models of migration in the economic and sociological literature (Todaro and Maruszko 1987; Williams and Baláž 2012).7In the widely used Todaro and Maruszko (1987) model, migrants weigh the benefits of migration, the likelihood of those benefits to materialize, and the costs involved in migration. Greater risk awareness would reduce the perceived likelihood of the benefits of migration materializing (e.g., because irregular migrants will not reach the destination country or will not find a job there) and increase the costs (mental and financial) involved. A recent systematic review of available campaign evaluations revealed that the evidence base for programming and policymaking in the migration field is strikingly limited (Tjaden, Morgenstern, and Laczko 2018). Beyond evidence regarding particular campaigns, there is also a lack of general examination of the overall mechanism. Only a few experimental studies (outside the lab) are available that test the role of risk awareness and information for migration decision-making. Bryan, Chowdhury, and Mobarak (2014) and Beam, McKenzie, and Yang (2016) look at the impact of information incentivizing (internal) migration in Bangladesh and the Philippines. Both studies find that providing information on job opportunities elsewhere does not have any impact on migration behavior. Shrestha (2019) highlights the importance of access to information for potential migrants’ expectations and their subsequent (internal) 750 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION migration decisions using a randomized controlled trial (RCT) in Nepal. Findings suggest that providing information on mortality rates during the migration journey affects subsequent migration decisions. Providing information on mortality rates lowers the expected mortality rates which then increases the propensity to migrate. Providing information on wages at destination reduces expected wages and, in turn, reduces migration intentions. It is important to note that mortality rates as used by Shrestha (2019) are prone to measurement error as actual mortality is likely to be vastly underestimated based on available sources, particularly in the context of irregular migration from Africa to Europe (Black 2020). Recent work by Mesplé-Somps and Nilsson (2020) tests the effects of 20-minute documentary films on the migration intentions of a probability sample in one rural district in Mali using a RCT. The authors compare the effects of positive role models (i.e., migrants who were successful) and negative role models (i.e., migrants who failed and experienced considerable hardship). The study finds no evidence of any effect induced by the films, concluding that for most participants the returns of migration are too high for campaigns aiming to change perceptions of risks or opportunities at home to matter. Lastly, Tjaden and Dunsch (2021) recently conducted an RCT study on the effects of townhall events organized to raise awareness of the risks of irregular migration among potential migrants in Dakar, Senegal. Initial findings suggest that participation in such events has the potential to increase risk perceptions and reduce intentions to migrate irregularly. No effects on factual knowledge gains were found and the sample was limited to a large regional hub in urban settings. Urban settings may differ substantially from more rural settings in terms of livelihood opportunities, transnational family networks, availability of migration-relevant information (through organizations, media, and the internet), and prevalence of smuggling networks among other factors. Empirical methods Data The analysis is based on a (gross) sample of 2,825 potential migrants in the Guinean region of Boké, namely Boffa, Boké, Gaoual, and Koundara, collected in January and February 2019. All villages in the treatment group were selected during a scouting mission in December 2018. The provinces were chosen based on emigration levels and their border position with Senegal and Guinea-Bissau. Within these regions, the villages were selected according to accessibility of the village by the caravan (one truck and two minivans) and approval by local authorities (mayor or chief of the village), who were first approached during the scouting mission. For each selected TJADEN AND GNINAFON 751 FIGURE 1 Sampled control and treatment villages NOTE: Data were collected in a total of 63 villages between January and May 2019. Figure 1 shows the geographical position of both treatment villages (the yellow dots) and the control group (the blue dots). village along the route of treatment villages, we carefully selected a controlgroup village. Control villages were selected based on similarity (language, ethnic composition, infrastructure) and distance to treatment villages, that is, six km apart to avoid spillover. One more control village was added to adjust for oversampling in treatment villages (in anticipation of noncompliance in treatment villages) and to preempt attrition issues. Figure 1 shows a map of the sampled sites. Baseline data were collected one to two days prior to the date of the treatment event in both control and treatment villages. It is important to note that, throughout the campaign, the impact evaluation activities did not overlap or interfere with campaign implementation. Data were collected by a trained team of local Guinean nationals who conducted interviews with likely event attendees one or two days before the event team arrived. Enumerators were allocated various key sites in the village (such as mosque, market, school, and main street) and conducted random walks in the vicinity. After identifying a specific location for the enumerators to start, households were selected through the household selection procedure used in the Afrobarometer surveys.8At the household level, from a list provided by the head of household or his/her representative, the interviewer randomly selected a respondent who agreed to participate in the survey. Eligible target group members were required to meet three criteria: (1) the respondent must be between the ages of 15 and 39 years of age9; (2) the 752 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION TABLE 1 Sample overview and adjustment Stage Control Treatment Running total Baseline 1,086 1,739 2,825 Reduced sample due to… Noncompliance 204 Spillover 3 Attrition 43 41 Trimming based on propensity score 331 826 Final sample for analysis 709 668 1,377 NOTE: Final sample in the model may vary depending on item nonresponse. respondent had generally considered migration in broader terms; and (3) the respondent had an interest in attending a movie event on migration. These criteria were applied in both control and treatment villages to attenuate the potential risk of bias due to differences in selection into treatment. The endline was conducted in April and May 2019, between three to five months after the intervention. The analysis accounts for minor variations in the time between intervention and endline. Based on a complete list of participants from the baseline survey, enumerators set up appointments via mobile phones to schedule face-to-face interviews. Respondents were offered a phone credit of 30,000 Guinean francs (approximately USD 4) for successful participation in the endline survey. It is important to note that no incentives were used in the baseline survey to reduce selection into the study. Considering low and balanced attrition overall (three percent), incentives are unlikely to introduce substantial bias. In cases where a face- to-face interview was not achieved, the survey was conducted by telephone. When the interviewer was unable to identify or locate the baseline respondents, several strategies were used to locate them. Upon the arrival of the survey team in the village, enumerators approached the head of the village, passers-by, religious leaders, or market workers for a referral to find the missing baseline respondents. The contact details of baseline respondents who were still not reinterviewed were sent to a subcontracted call center in Conakry that followed up with respondents several times over the phone. As a result of these intense efforts to reduce attrition, only 84 baseline respondents were not re-interviewed in the endline (three percent). The endline survey also verified compliance and contamination. Out of 1,739 re-interviewed individuals in the treatment villages, 1,535, 88 percent (i.e., compliers) attended the intervention (main film and discussion). Only three respondents in the control group were exposed to treatment (i.e., contamination). Table 1 shows the sample at different stages of the process including information on attrition, spillover, and trimming (see the next section). TJADEN AND GNINAFON 759 in the campaign induced additional information-seeking behavior or additional informal exchanges at the local level that reinforced awareness rather than reducing it. The timing of the endline was associated with the location of treatment and control villages along the route of the mobile caravan. Therefore, this effect remains tentative and requires further research. Robustness checks The results are robust against different specifications (see Table A4 in the Supporting Information). As an alternative specification to our main DiD estimation of choice (DiD based on a trimmed sample and individual fixed effects, see section Treatment and specification IV in Table A4), we report results for (I) DiD estimation without covariates based on the gross sample, (II) DiD estimation based on the gross sample with individual fixed effects, (III) DiD estimation with covariates (see caption below Table A4), and (V) DiD estimation based on the trimmed sample and covariates (without individual-fixed effects). The results are qualitatively similar across specifications. Threats to validity In line with the theory of change of the intervention, we interpret the effects as evidence that potential migrants adjust their perceptions and intentions in response to learning from the experiences of their peers through the films and discussion. However, response biases such as social desirability bias and reciprocity bias may threaten the validity of this interpretation. These response biases are common to most experimental and quasi-experimental studies and are generally difficult to avoid. Social desirability bias emerges when respondents adjust their response to what they perceive to be the “correct” or socially desirable response. Respondents in treatment villages were exposed to messages surrounding the dangers of migrating; therefore, respondents may infer that the “correct” response at endline is to state high risk awareness following the intervention. Respondents in the control group were not exposed to any intervention. As a result, respondents in the control group may not form a conception of what the desirable response would be. We introduced several measures at the data collection stage to reduce the risk of this bias. All enumerators were locals and had no association with the campaign itself. It was made clear to the respondents that data collection was conducted as part of an independent exercise, that their responses would have no impact on follow-up interventions, and that data would not be shared with third parties including governments. Moreover, the endline survey took place three to five months after the intervention. Respondents had time to deliberate their preferences, seek additional information, discuss with friends, 760 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION etc. Social desirability bias should present a greater threat shortly after the intervention took place. Reciprocity bias is a concern when respondents adjust their responses in anticipation of a (financial) reward provided by the data collection team. We believe this bias is not a main concern for our study. First, all respondents, in both the treatment and control groups, received the same incentive (phone credit). If reciprocity applied, it applied equally to both groups and as such does not affect the treatment estimates. Second, the amount was very limited (approximately USD 4). Third, it was made clear to respondents at the beginning of the interview that the incentive was conditional on completing the survey and was not linked to the type of responses provided. Fourth, it was made clear that the data collection team had no links with the campaign organization and that survey responses had no impact on entitlements or participation in other projects. In sum, we argue that response bias is of minor concern; however, as with most DiD studies, we cannot exclude the possibility that treatment effects are—at least partially—influenced by social desirability bias. Discussion In response to mounting evidence of the harm inflicted on irregular migrants along their journeys from West Africa to Europe, international organizations, civil society organizations, and governments have scaled up awareness-raising campaigns about the risks of irregular migration. The official objective of many campaigns is reportedly to increase awareness, counter misinformation by smugglers, and facilitate safe migration decisions. Despite the growing number of interventions, there is limited empirical evidence on the impact and effectiveness of such campaigns. In contrast, there is widespread doubt expressed in some parts of the literature showing that campaigns are based on wrong assumptions about how migrants make decisions. More generally, empirical evidence measuring the role of risk perceptions in the decision-making process of potential irregular migrants remains limited while it is generally assumed in theoretical models (Todaro and Maruszko 1987). This study aimed to contribute an empirical account based on a case study in North-Western Guinea—a region with a high prevalence of outmigration. Based on a quasi-experimental DiD design, we estimated the effect of participating in awareness-raising events (screening edutainment films about the risks of irregular migration based on personal testimonials of migrants, followed by moderated community discussion among participants from the local community) on potential migrants’ perceptions, knowledge, and intentions regarding irregular migration. The results show that potential migrants who participated in movie and discussion events were significantly less likely to report high intentions TJADEN AND GNINAFON 761 to migrate irregularly and significantly more likely to report higher awareness regarding the costs involved in irregular migration; they also had heightened perceptions of the level of certain (not all) risks related to irregular migration and a more positive perception of future economic opportunities at home. The effects were larger for those potential migrants with little to no information about migration to Europe at baseline. While general and abstract risk perceptions are already high among potential migrants in the sample at baseline, the intervention appeared to further increase perceptions of specific risks associated with the irregular migration journey. In addition, treatment effects appear to increase marginally with time, suggesting that the intervention may have induced additional information-seeking behavior among potential migrants following the campaign. These results suggests that general dismissals of information and awareness-raising interventions targeting potential irregular migrants require nuance. Consistent, statistically significant effects on various migration-relevant outcomes underscore the potential of campaigns to inform the decision-making process of certain groups of potential irregular migrants and, in some cases, potentially reduce harm for those facing perils associated with irregular journeys from West Africa to Europe by land and sea. The general finding that context information matters for migration choices is consistent with emerging experimental evidence from different settings (Tjaden and Dunsch 2021; Shrestha 2019, 2020; Bah and Batista 2018). The results are also consistent with general rational-choice models of migration decisions which assume that risk perceptions may condition the likelihood of expected benefits of migration (e.g., income) to materialize (Todaro and Maruszko 1987; Massey et al. 1993). However, the results are inconsistent with recent findings in a similar study in Mali (Mesplé-Somps and Nilsson 2020) which found no effects of films featuring positive and negative role models on migration intentions. Unlike the intervention we have assessed, the Malian study relied only on a 20-minute documentary screening without follow-up discussion and without any in-person contact with migrants who shared their stories. In addition, the Malian sample was not tailored to potential migrants with reported interest in migration but to the general population. Bryan et al. (2014) and Beam, McKenzie, and Yang (2016) also found no effect on migration behavior of information on job opportunities elsewhere. In taking stock, the mixed results underscore the need for further research on campaign—and more broadly, information— effects on migration. In the meantime, it appears that campaign effects depend highly on the context, the target group, and specifics of the intervention. Awareness-raising interventions are commonplace in other fields of development (i.e., health, family planning, and education; see e.g., Battaglia and Pallarés 2020; Bongaarts and Casterline 2013). The evidence presented 762 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION in this study suggests that—similar to other fields—there is reason to believe that campaigns can be an effective tool in certain circumstances. The conversation should go beyond dismissing the general approach and move toward a discussion on which types of campaigns may be useful at a particular time and location. For example, the results suggest that interventions targeting individuals who are in early stages of their decision-making process may be particularly effective. Some limitations of this study should be mentioned which also put potential policy lessons into context. First, this study was unable to measure actual behavior which is notoriously difficult given the mobile and irregular setting of migration from Guinea to Europe. Intentions are an important predictor of actual behavior (Tjaden, Auer, and Laczko 2019); however, the reasons that lead to leaving may differ from those that shape prior preferences and intentions. This question is inherently linked to policy decisions about the intended outcomes of campaigns. Knowledge and awareness gains may be suitable and ethical indicators of campaign effectiveness when campaigns aim to promote safe migration and facilitate informed decisions (regardless of eventual migration behavior). While actual behavior is of key interest, it may not be an adequate indicator of success in cases when the way in which migration takes place (i.e., safely or unsafely) cannot be measured. Effect sizes in this study appear to be large for social scientists (approximately 8–20 percentage points relative to the control group), yet modest from a policy perspective. Campaigns by themselves are no silver bullet. On average, one to two potential migrants out of 10 change their mind based on new information acquired through interventions. Structural factors driving the initial consideration of migration (such as livelihoods, income opportunities, access to health care, etc.)—which have been held constant in this study—are likely to outweigh the role of information. More insights are needed on the relative cost-effectiveness of campaigns vis-à-vis alternative approaches to promote safe migration and prevent harm, including supporting local livelihood interventions or alterative legal pathways within the region or beyond. Second, sustainability of effects may be of concern. The study assessed outcomes three to five months after the intervention. More research is needed to assess long-term impacts. Tentative evidence from heterogenous treatment effects in this study suggests that effects could actually increase over time but more research is needed. Lastly, as is common in survey-based research, social desirability bias may be an issue. We attempted to mitigate the risk of this bias, yet it is not possible to test for it directly. In sum, awareness-raising interventions appear to be a viable option in the toolbox of migration policymakers, yet they are well advised to keep TJADEN AND GNINAFON 763 realistic expectations, carefully tailoring campaigns and clearly defining campaign objectives. Data availability Statement Data and materials to replicate the analysis are available upon request. Please send an email to jasper[email protected]. Acknowledgment Open access funding enabled and organized by Projekt DEAL. Notes 1 The authors would like to thank Marie-Luce Zafinikamia (IOM) for her pivotal support on data collection, field coordination, and data management, as well as the team at the International Organization for Migration’s (IOM) Global Migration Data Analysis Centre (GMDAC). We would also like to thank the team at the Guinea country mission of the IOM for their support, including Lucas Chandellier, Mohamed Doumbouya, and Mohammad Coulibaly; and Giulia Falzoi, Shiraz Jerbi and Flavio Di Giacomo at the IOM Coordination Office for the Mediterranean-Liaison Mission for Italy, Malta, and the Holy See. In addition to the anonymous reviewers, earlier versions of the paper benefited greatly from feedback from David McKenzie (World Bank); Bernd Beber (WZB Berlin Social Science Center); Catia Batista (NOVA School of Business and Economics); Yvonne Giesing and Nadzeya Laurentsyeva (ifo Institute, Munich); Tobias Stöhr (IfW Kiel); Gilles Spielvogel (Organisation for Economic Co-operation and Development); Nahomi Ichino (Emory University); Sandra Morgenstern (University of Konstanz); and Benjamin Chibuye (University of Kiel). The work was supported by UKAid by the Government of the United Kingdom as part of the project titled Safety, Support and Solutions in the Central Mediterranean Route; and by the German Federal Ministry for Economic Development and Cooperation through staff contributions. 2 According to an internal review by the European Commission, the European Union and European Union governments have funded, implemented, or endorsed more than 100 such information and awarenessraising campaigns targeted at migrants since 2014. The provision of information and awareness-raising activities form part of major policy framework between the EU and Africa on the issue of migration such as the EU Emergency Trust Fund for Africa (EUTF) and the Migration Partnership Framework. 3 See https://data2.unhcr.org/en/situatio ns/mediterranean and Europe | Flow monitoring (iom.int). 4 See afrobarometer.org/sites/default/fil es/publications/Dispatches/ab_r7 _dispatchno288_looking_for_opportunity _africans_views_on_emigration1.pdf. 5 See https://data2.unhcr.org/en/situati ons/mediterranean. 6 See https://data2.unhcr.org/en/situati ons/mediterranean. 7 One notable exception is the case study of migration between Mexico and the United States. Several studies have examined the potential role of border enforcement and inferred risks on the decision-making process of potential irregular migrants (Espenshade 1994; Ryo 2013). Given its unique context (large diaspora, shared border, bilateral trade), comparability with West African migration to the EU is limited. 8 “Team members must walk in opposite directions to each other. If A walks towards the sun, B must walk away from the sun; C and D must walk at right angles to A and B. Use a 5/10 interval pattern to select a household. That is, walking in your 764 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION designated direction away from the starting point, select the fifth household for the first interview, counting houses on both the right and the left (and starting with those on the right if they are opposite each other). Once you leave your first interview, continue on in the same direction, this time selecting the tenth household, again counting houses on both the right and the left. If the settlement comes to an end and there are no more houses, turn at right angles to the right and keep walking, continuing to count until finding the tenth dwelling.” See https://afrobarometer. org/sites/default/files/survey_manuals/ab_r7 _survey_manual_FR.pdf. 9 Massandouno and Cissé (2017) used data from the 2014 General Population and Housing Census to show that the large share of international migrants are aged from 15 to 39 years old. 10 Hemisphere Africa, “Migrants, return from hell,” June 9, 2017. Available at https://www.youtube.com/watch? v=icfztPXNiDk&t=1401s. References Abadie, Alberto and Matias D. Cattaneo. 2018. “Econometric Methods for Program Evaluation.” Annual Review of Economics 10(2018): 465–503. Alpes, M.J., and N. Sorensen. 2015. “Migration Risk Campaigns Are Based on Wrong Assumptions.” DIIS Policy Briefs. Copenhagen: Danish Institute for International Studies. Bah, T., and Catia Batista. 2018. “Understanding Willingness to Migrate Illegally: Evidence from a Lab in the Field Experiment. ” NOVAFRICA Working Paper No. 1803. Lisbon: Universidade Nova de Lisboa, Nova School of Business and Economics, NOVAFRICA. Banerjee, Abhijit, Eliana La Ferrara, and Victor Orozco-Olvera. 2019. “The Entertaining Way to Behavioral Change: Fighting HIV with MTV.” Cambridge, MA: National Bureau of Economic Research. Batista, Catia, and David McKenzie. 2021. “Testing Classic Theories of Migration in the Lab.” CEPR Discussion Paper No. DP16469. London: Centre for Economic Policy Research. Battaglia, M., and N. Pallarés. 2020. “Family Planning and Child Health Care: Effect of the Peruvian Programa de salud reproductiva y planificación familiar, 1996–2000.” Population and Development Review 46(1): 33–64. Beam, Emily A., David McKenzie, and Dean Yang. 2016. “Unilateral Facilitation Does Not Raise International Labor Migration from the Philippines.” Economic Development and Cultural Change 64(2): 323–368. https://doi.org/10.1086/683999 Bernard, T., S. Dercon, K. Orkin, and A. Seyoum Taffesse. 2015. “Will Video Kill the Radio Star? Assessing the Potential of Targeted Exposure to Role Models through Video.” The World Bank Economic Review 29: S226–S237. Black, J. 2020. “‘No One Talks about What It’s Really Like’– Risks Faced by Migrants in the Sahara Desert.” Migration in West and North Africa and Across the Mediterranean: Trends, Risks, Development and Governance. Geneva: International Organization for Migration. Black, J., K. Dearden, A. Singleton, and Frank Laczko. 2017. “Global Overview of the Available Data on Migrant Deaths and Disappearances.” Fatal Journeys, Volume 3, Part 1. Geneva: International Organization for Migration: 1–25. Bongaarts, J., and J. Casterline. 2013. “Fertility Transition: Is Sub-Saharan Africa Different?” Population and Development Review 38(Suppl 1): 153–168. https://www.ncbi.nlm.nih.gov/pmc/ articles/pmc4011385/ Browne, E. 2015. “Impact of Communication Campaigns to Deter Irregular Migration.” GSRDC Helpdesk Research Report 1248. Birmingham, UK: GSDRC, University of Birmingham. Bryan, Gharad, Shyamal Chowdhury, and Ahmed Mushfiq Mobarak. 2014. “Underinvestment in a Profitable Technology: The Case of Seasonal Migration in Bangladesh.” Econometrica 82(5): 1671–1748. https://doi.org/10.3982/ECTA10489 TJADEN AND GNINAFON 765 Crump, Richard K., V. Joseph Hotz, Guido W. Imbens, and Oscar A. Mitnik. 2009. “Dealing with Limited Overlap in Estimation of Average Treatment Effects.” Biometrika 9(1): 187–199. https://doi.org/10.1093/biomet/asn055 Czaika, M., and H. de Haas. 2013. “The Effectiveness of Immigration Policies.” Population and Development Review 39(9): 487–508. Dunsch, Felipe, Jasper Tjaden, and W. Quiviger. 2019. “Migrants as Messengers: The Impact of Peer- to-Peer Communication on Potential Migrants in Senegal.” Impact Evaluation Report. Geneva: International Organization for Migration. Espenshade, T.J. 1994. “Does the Threat of Border Apprehension Deter Undocumented US Immigration?” Population and Development Review 20(4): 871–892. https://www.jstor.org/stable/ 2137667 European Commission. 2018. “How West African Migrants Engage with Migration Information En-Route to Europe.” Luxembourg: European Union. Europol/Interpol. 2016. “Migrant Smuggling Networks: Joint Europol-INTERPOL Report. Executive Summary.” The Hague: Europol/Interpol. Fiedler, Anke. 2020. “From Being Aware to Going There: On the Awareness and Decision-Making of (Prospective) Migrants.” Mass Communication and Society 23(3): 356–377. Foran, R., and A. Iacucci. 2017. “Lost in Translation: The Misinformed Journey of Migrants Across Italy.” London: Internews. https://internews.org/wp-content/uploads/legacy/ Internews_Lost_In_Translation_Publication_2017-05-23.pdf Galos, E., L. Bartolini, H. Cook, and N. Grant. 2017. “Migrant Vulnerability to Human Trafficking and Exploitation: Evidence from the Central and Eastern Mediterranean Migration Routes.” Geneva: International Organization for Migration. https://publications.iom.int/books/ migrant-vulnerability-human-trafficking-and-exploitation-evidence-central-and-eastern Gibson, John, and David McKenzie. 2014. “The Development Impact of a Best Practice Seasonal Worker Policy.” The Review of Economics and Statistics 96(2): 229–243. https://doi.org/10.1162/ REST_a_00383 Gillespie, M., L. Ampofo, M. Cheesman, B. Faith, E. Iliadou, A. Issa, S. Osseiran, and D. Skleparis. 2016. “Mapping Refugee Media Journeys: Smartphones and Social Media Networks.” The Open University /France Médias Monde. Hernández-Carretero, María and Jørgen Carling. 2012. “Beyond ‘kamikaze migrants’: Risk taking in West African boat migration to Europe,” Human Organization 71(4): 407–416. https://doi. org/10.17730/humo.71.4.n52709742v2637t1 International Organization for Migration (IOM). 2017. “IOM Niger—2016 Migrant Profiling Report.” Geneva: IOM. ———. 2018. “Displacement Tracking Matrix—Flow Monitoring Europe.” Geneva: IOM. https://migration.iom.int/europe?type=arrivals Maher, S. 2017. “Historicising ‘Irregular’ Migration from Senegal to Europe.” Anti-Trafficking Review 9: 77–91. Massandouno, L., and M. Cissé. 2017. “Rapport d’analyse des données du recensement général de la population et de l’habitation 3: Migration et urbanisation [Analysis of the General Population and Housing Census 3: Migration and Urbanization].” Conakry: Ministre du Plan et la Cooperation Internationale. Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor. 1993. “Theories of International Migration: A Review and Appraisal.” Population and Development Review 19(3): 431. https://doi.org/10.2307/2938462 Mbaye, L.M. 2014. “Barcelona or Die: Understanding Illegal Migration from Senegal.” IZA Journal of Migration 3(1): 1–21. Mesplé-Somps, S. and B. Nilsson. 2020. “Role Models and Migration Intentions.” Research Paper No. 200. Paris: Agence française de développement. Mixed Migration Centre (MMC). 2018. “Fraught with Risk: Protection Concerns of People on the Move Across West Africa and Libya.” Geneva: MMC. https://mixedmigration.org/resource/ fraught-with-risk/ 766 RAISING AWARENESS ABOUT THE RISK OF IRREGULAR MIGRATION Prothmann, Sebastian. 2018. “Migration, Masculinity and Social Class: Insights from Pikine, Senegal,” International Migration 56(4): 96–108. https://doi.org/10.1111/imig.12385 Regional Mixed Migration Secretariat. 2014. “Blinded by Hope: Knowledge, Attitudes and Practices of Ethiopian Migrants.” Mixed Migration Research Series. Nairobi: Regional Mixed Migration Secretariat. Ryo, Emily. 2013. “Deciding to Cross,” American Sociological Review 78(4): 574–603. https://doi.org/ 10.1177/0003122413487904 Schans, D., and C. Optekamp. 2016. “Raising Awareness, Changing Behavior? Combatting Irregular Migration Through Information Campaigns.” The Hague: Ministry of Security and Justice, WOCD. Shrestha, Maheshwor. 2019. “Death Scares: How Potential Work-Migrants Infer Mortality Rates from Migrant Deaths.” Journal of Development Economics 141: 102368. https://doi.org/10.1016/ j.jdeveco.2019.07.001 ———. 2020. “Get Rich or Die Tryin’: Perceived Earnings, Perceived Mortality Rates, and Migration Decisions of Potential Work Migrants from Nepal.” The World Bank Economic Review 34(1): 1– 27. https://doi.org/10.1093/wber/lhz023 Tjaden, Jasper, D. Auer, and Frank Laczko. 2019. “Linking Migration Intentions with Flows: Evidence and Potential Use.” International Migration 57(1): 36–57. https://doi.org/10.1111/imig. 12502 Tjaden, Jasper, and Felipe Dunsch. 2021. “The Effect of Peer-to-Peer Risk Information on Potential Migrants—Evidence from a Randomized Controlled Trial in Senegal.” World Development 145: 105488. https://doi.org/10.1016/j.worlddev.2021.105488 Tjaden, Jasper, Sandra Morgernstern, and Frank Laczko. 2018. “Evaluating the Impact of Information Campaigns in the Field of Migration.” Central Mediterranean Route Thematic Report Series. Geneva: International Organization for Migration. https://eapmigrationpanel8.idudka.com/ sites/default/files/evaluating_the_impact_of_information_campaigns_in_field_of_migra tion_iom_gmdac.pdf Todaro, Michael P. , and Lydia Maruszko. 1987. “Illegal Migration and US Immigration Reform: A Conceptual Framework.” Population and Development Review 13(1): 101. https://doi.org/10. 2307/1972122 UNDP. 2020. “Scaling Fences: Voices of Irregular African Migrants to Europe.” New York: United Nations. www.undp.org/content/dam/rba/docs/Reports/UNDP-Scaling-Fences-EN- 2019.pdf UNHCR. 2018a. “Central Mediterranean Situation: Supplementary Appeal, January–December 2018.” Geneva: United Nations. ———. 2018b. “Desperate Journeys: Refugees and Migrants Arriving in Europe and at Europe’s Borders, January–December 2018.” Geneva: United Nations. United Nations Office on Drugs and Crime. 2018. “Global Study on Smuggling of Migrants 2018.” (United Nations Publication, Sales No. E.18.IV.9). New York: United Nations. van Bemmel, Sjoerd. 2020. “The Perception of Risk among Unauthorized Migrants in Ghana.” Journal of Risk Research 23(1): 47–61. https://doi.org/10.1080/13669877.2018.1517376 Williams, Allan M., and Vladimir Baláž. 2012. “Migration, Risk, and Uncertainty: Theoretical Perspectives.” Population Space and Place 18(2): 167–180. https://doi.org/10.1002/psp.663