Sustainable Reintegration of Returning Migrants

2. Return migration: scope and drivers

Abstract

Return migration is difficult to measure, since much of it is spontaneous and goes unrecorded, although official data exist on forced return and voluntary assisted return in most OECD countries. The literature suggests that return decision-making is complex and influenced by a variety of factors including the conditions in the origin and destination country, individual and social factors, and, to a limited degree, policy interventions.

The chapter reviews the terminology used in discussing return migration. It then examines the scope and nature of return migration. It reviews the literature and evidence on the factors leading migrants to decide to return to their country of origin.

2.1. The scope and nature of return migration

Internationally comparable statistical information on return migration is limited. Attempts to measure the phenomenon in a comparable manner face two difficulties: the lack of a shared definition of return migration and the lack of data. Yet, knowing who returns and in what number matters for both the host and the home country, as the characteristics of returnees, in particular their education levels and the length and nature of their stay abroad, affect their probability of return, as well as the sustainability of their reintegration after their return.

Return migration occurs in different ways. The usual distinctions consider whether it is spontaneous, initiated by the migrant and without state involvement, or organised and enforced by state authorities (Box 2.1).

Box 2.1. Key Terminology: Forced, Voluntary and Assisted Voluntary Return

Forced return is “a migratory movement which, although the drivers can be diverse, involves force, compulsion, or coercion.” Voluntary return is “the assisted or independent return to the country of origin, transit or another country based on the voluntary decision of the returnee.” (IOM, 2019[1]).

Voluntary return can be either spontaneous or assisted: Spontaneous return is “the voluntary, independent return of a migrant or a group of migrants to their country of origin, usually without the support of States or other international or national assistance.” (IOM, 2019[1]). Assisted voluntary return (AVR) is the “administrative, logistical, financial and reintegration support to rejected asylum seekers, victims of trafficking in human beings, stranded migrants, qualified nationals and other migrants unable or unwilling to remain in the host country who volunteer to return to their countries of origin” (IOM, 2019[1]).

Assisted return programmes have come to include reintegration assistance in addition to return assistance. In addition to pre-departure counselling, return and travel assistance, Assisted Voluntary Return and Reintegration (AVRR) programmes offer cash and/or in-kind assistance to support re-insertion in their country of origin. Assistance may involve some or all of these: business start-up coaching and counselling, labour market counselling, vocational training – including on-the-job training – internships and job placement, housing, health care and children’s education.

In practice, return categories are not always distinct and involve varying degrees of voluntariness among the beneficiaries of both AVR and AVRR programmes (see Newland and Salant (2018[2])). For migrants in an irregular situation or asylum seekers with little chances of obtaining protection, AVRR may be a compelled choice, even in the absence of physical coercion. Some see return as voluntary only when individuals have alternative legal options and can make decisions based on a free and informed choice. As persons in these situations represent an expanding group of beneficiaries of AVRR, the line between forced and assisted voluntary return blurs.

Alternative descriptions have been used by scholars to describe situations where “voluntary return” occurs when options are severely constrained – e.g. “compelled return” (Cassarino, 2008[3]), “nominally voluntary return” (Gibney, 2008[4]) and “soft-deportation” (Leerkes, van Os and Boersema, 2017[5]). Several European countries, for example Austria, Norway and the United Kingdom, use the term “assisted return”, instead of ‘assisted voluntary return’, for their programmes.

Data on spontaneous outflows are difficult to capture, unless there is a reporting requirement in the destination and origin country, or a survey in the origin country capturing returns (see Table 2.1). The OECD collects data on reported outflows of the foreign population from selected OECD countries through its Continuous Reporting System on International Migration, which are published in the annual OECD International Migration Outlook (OECD, 2020[6]). Other approaches look at changes in the stock of foreigners and assume that residual outflows are directed to the origin country, rather than a third country. Regardless of the measure used for outflows, they appear to be significant: an OECD analysis (2008[7]) has found that, depending on the country of destination and the period of time considered, 20% to 50% of immigrants leave within five years after their arrival, either to return home or to move on to a third country (secondary emigration). Migrants who arrive for family or humanitarian motives have a lower return rate than economic migrants.

There is no standardised definition that would allow for comparisons of the size and composition of spontaneous outflows across datasets. There are differences in the geographical references chosen (country of birth, nationality or country of previous residence); the time references (short-term or permanent migration and return, including circular movements; intended return or actual duration); and return motivations (excluding short-term visits or holidays). Further, existing data sources do not allow for random sampling, as there is a lack of appropriate sample frames (population registers, censuses, general surveys). The smaller specific surveys that are available, e.g. the MAFE survey in Senegal, are costly to undertake and are often non-representative as they are collected by directly seeking out returnees, or are selected by areas of high return incidence, which means they are not representative at country level. More generally, the fact that returnees are often a small and widely distributed group within the population affects the quality of analyses based on the data and makes evidence – on which policy decisions can be based – costly and less feasible to collect.

The European Union’s TEMPER project is a recent attempt at overcoming the gap in data and better understanding return dynamics, looking at both permanent return and temporary return/circular migration. It looks at return and circulation between certain destination (France, Spain, Germany, Italy and Poland) and origin countries (Senegal, Argentina, Romania and Ukraine). TEMPER aims to create a “Repository of Migration Surveys” (Table 2.1) and combine information in pre-existent immigrant surveys with new survey data collected in origin countries (González et al., 2018[8]), covering profiles, mobility patterns (return and circulation included), legal trajectories, contacts with origin countries, etc., for different types of migration.

Table 2.1. Examples of Specialised Return Surveys

Survey	Return Migration Definition	Sampling	Sample Size
ETF (Armenia, Georgia, Morocco 2011)	Left at 18 years or older, returned no more than 10 years prior	Nationally representative sample of non-migrants + snowball sample of returnees (in non-migrant households or in the neighbourhood)	1 311 return migrants, of which 398 from France, 272 from Spain, 203 from Italy
NOPOOR (Ecuador 2014)	Returned from Spain, lived there for one year or more, returned one year or more	Snow-ball sampling + recruitment workshop + website, leaflets	450 return migrants (out of 1 000 expected)
MAFE (Senegal 2008)	Left for Europe at 18 years or older; born in Senegal and has had Senegalese citizenship	Ares with different return incidence; and a listing operation was carried out in each of the selected survey sites	54 Return migrants; total 1 500 (including non-migrants)
ENAMIR (Colombia 2013)	Born in Colombia, returned with the intention to reside there again	Stratified sampling (geographical regions with high migration rates) in 2 stages	21.093 households (1 876 return migrants from all destinations)
MED-HIMS (Egypt 2013)	Last returned from abroad to the country of origin since 1 January 2000; who were 15 years or older on last return	Stratification to oversample where migration concentrated in some regions	90.012 households surveyed (5 085 return migrants from all destinations)

Source: EU Temper Project.

Return migration is self-selected. For spontaneous returns, the evidence suggests that the return rate changes over the life cycle of migrants, with higher rates for the young and for retirees. Returns by level of education also produce a U-curve; i.e. the return rate is higher at both ends of the education spectrum. Return migration tends to accentuate the selection that originally motivated migration. Return migration selection tends to be the reverse of the initial selection process for migration: if the host country attracts mainly skilled migrants, return migrants will likely be less skilled on average than the remaining immigrants in the host country, while if the host country attracts relatively unskilled workers, the better skilled among them are most likely to return (Borjas and Bratsberg, 1996[9]). Selection in return must be accounted for in analyses of return and reintegration outcomes, as it can lead to endogeneity problems and produce biased estimates.

Beyond spontaneous returns, destination countries arrange organised returns – either through assisted voluntary return programmes or through forced removals. Coverage of the target population is partial. To monitor return rates, Member States must provide Eurostat with statistics on the enforcement of migration legislation, which means that comparable data on return decisions and executions are available for EU countries. On the aggregate level (EU 28, including the UK until 01 February 2020), orders to leave peaked at more than 533 000 in 2015, a record high since 2011, and have remained at around a half million since, with annual variations. In line with this, returns peaked at 250 000 in 2016, the highest level since 2010, but have been falling steadily since then (Figure 2.1). In 2016, less than half of the migrants ordered to leave the European Union did so. In 2017, just above one third did. This figure is not the actual return rate, since it does not take into account time lags, appeals, spontaneous returns and other resolution of status obviating a return order. The 2017 renewed Action Plan of the European Union (European Commission, 2017[10]) aimed at increasing return rates.

Data trends on organised return vary greatly across EU Member States, due to different policies and enforcement capacities for forced returns, as well as to each country’s position in the irregular migration routes – and their evolution over time. Thus, in 2015, almost one in five orders to leave for the EU28 was issued by Greece, which drove the overall increase. Country trends have diverged from the aggregate picture. In Spain, for instance, orders to leave almost doubled in 2018 relative to the 2015 figure, reflecting shifts in irregular migration routes (from Eastern to Western Mediterranean route). In 2015, France issued a lower number of orders to leave than in 2013‑14, but the figures have grown steadily ever since.

In the EU, since 2014, Eurostat collects a broader set of data on return, including by type of return (voluntary, forced, other), assistance received by the returnee (assisted, non-assisted return), agreement procedure underlying return, citizenship, and destination country (i.e. country of return). Specifically, it includes data on “third country nationals – ordered to leave; returned following an order to leave; who have left the territory by type of return and citizenship; who have left the territory by type of assistance received and citizenship”. However, considering that those data are collected by countries on a voluntary basis, not all data for countries implementing AVR are reported in the Eurostat database. Further, spontaneous departures are usually not tracked.

Recent initiatives on the European level aim at overcoming various gaps in measuring return migration, with a focus on measuring return of migrants falling into categories which are subject to removal. The European Commission is implementing Integrated Return Management Application (IRMA), a secure web-platform for integrating all EU return activities. The restricted information exchange platform connects EU Member and Schengen States, the European Commission, the European Border and Coast Guard Agency (Frontex) and the relevant EU funded programmes at operational, practitioner level to collect data on return and readmission activities.

Other OECD countries also publish statistics on returns. In the United States, figures distinguish forced returns between removals and returns; the latter involves cooperation and does not carry the same re-entry ban. Removals averaged about 370 000 annually over the past decade, with a peak in 2012‑13, while returns have fallen sharply, from more than 800 000 in 2008 to 100 000 in 2017. In 2017, about two‑thirds of returns were to Mexico and one‑fourth to the Northern Triangle (El Salvador, Honduras and Guatemala). 41% of removals were of criminals. Returns were largely to the same destinations, although Asia comprised 20% of returns in 2017. Voluntary departure from the United States – when aliens facing deportation decide not to contest their removal and to leave on their own, rather than have a deportation on their record – have been increasing. Voluntary departure fell from 23 900 in FY2012 to 8 400 in FY2016, but rose again to 23 800 in 2018 (Thompson and Calderón, 2019[11]).

Figure 2.1. Number of third country nationals ordered to leave, persons returned, and the return rate (%) in the European Union, 2008-19

Source: Eurostat, 2019.

Canada distinguishes between forced removals and voluntary compliance with removal orders. Total removals in 2018‑19 stood at 9 600, down from 17 500 in 2012 but increasing since the 2016 figure of 7 400. About 60% of removals are forced, with the remainder voluntary, with travel costs borne by the returnee and not by the Canadian authorities. In Japan, annual removals averaged about 350 between FY 2013 and FY 2018; about 10 persons received AVRR annually. In addition, about 6 600 persons subject to removal left on their own without assistance annually. Israel publishes figures on voluntary departures, distinguishing between those returning to their origin country, a resettlement country or a third country. The number of voluntary returns has stood at about 3 000 annually in recent years, although the programme targets Eritreans and Sudanese and most do not return to their origin countries, but rather depart to other destinations. Australia removed about 7 700 refused asylum seekers between 2010 and 2015, comprising about one‑third of all asylum seekers with negative decisions over the same period (Cuthbert and Song, 2017[12]). In the past five years, Australia further removed about 6 000 persons.

In addition to governments, implementers and service providers also acquire statistics. IOM, the largest global provider of both AVR and AVRR programmes, collects voluntary return data under its various programmes on a regular basis. IOM data include the number of migrants assisted, migrants’ host and origin countries, as well as sex, age and reintegration support. IOM data also include information on assisted migrants by specific vulnerability (namely, unaccompanied and separated migrant children, migrants with health-related needs and victims of trafficking).

Since 2010, IOM publishes these data on its AVRR page and in its publication on “Return and Reintegration: Key Highlights” (IOM, 2019[13]). In 2018, IOM assisted a total of 63 300 migrants through its AVRR programmes worldwide, representing a 12% decrease compared to 2017. Of the migrants assisted, almost half – 30 900 individuals, or 49% of all AVRR beneficiaries – returned from two countries, Germany and Niger, each with about 15 000 return migrants. The other countries sending large numbers of AVRR beneficiaries home included Greece (5 000 migrants), Austria (3 500), Djibouti (3 400), Belgium (2 800), Netherlands (2 100), Morocco (1 500), Turkey (1 500) and Italy (1 000). This reflects that a large part of the return assistance provided through IOM is from non-OECD countries. Over the past years, the figures for assisted return of stranded migrants within Africa has increased. Since January 2018, IOM Libya has assisted more than 17 500 stranded migrants to return to 32 countries across Africa and Asia.

Spontaneous returns are more difficult to measure and the return of naturalised immigrants or even legal permanent residents is generally unrecorded or partial. Precise data are thus largely limited to persons whose return involves state authorities in the destination countries, especially forced returns. For many of these returns, the data available do not go beyond nationality, age and gender. Rarely do these include the duration of stay, the grounds for removal or more detailed individual characteristics. Kosovo, for example, registers repatriated Kosovars in its own Case Management System, with detailed information, but does not capture information on other returns. There remain substantial gaps in the knowledge about who is returning to which countries and in what numbers. Since returnees’ characteristics – and their capabilities and vulnerabilities – determine their outcomes upon return, as well as their specific needs for reintegration support, AVRR programmes require better understanding of target group profiles along different return corridors.

2.2. Drivers of return decisions

While there is a well-established literature on how migrants make the decision whether and where to move, and some research on refugee return, there has been less research on the migrant return decision-making process. Most research has been conducted with migrants who have already returned, giving a retrospective picture of what affected a decision to return, but less information on what drives migrants not to return. The literature suggests that return decision-making is complex and influenced by a variety of factors including the conditions in the origin and destination country, individual and social factors, and, to a limited degree policy interventions. Within this literature, much of the evidence is from cases of spontaneous return – i.e. by migrants who were under no obligation to leave the host country or were under no constraint to return.

The main structural reasons assumed to explain migrants’ decision to return relate to the situation in the host country, but also to the opportunities open to them in their origin country. Existing studies based on interviews with potential returnees from/to different countries find contradicting results as to whether destination or origin country factors weigh more heavily (Black et al., 2004[14]; Koser and Kuschminder, 2015[15]; Strand et al., 2011[16])):

• conditions in the destination country: prospects for a future in the country, failure to integrate into the host country, living conditions, asylum policies, chances of forced return and others.

• conditions in the origin country: safety, human rights, political, economic and social, personal relationships and personal resources in the home country (family, housing, job opportunities, networks, ties, debt), etc.

Most existing studies on return migration seek to explain the return of labour migrants or those who migrated irregularly for economic reasons, with different migration theories generating competing hypotheses with regard to the determinants of return (Figure 2.3).

Studies on migrant return have focused on persons who have returned, rather than those who did not choose to return. For most, return is motivated by economic considerations. On the one hand, return of labour migrants is explained as a logical step after migrants have earned sufficient assets and knowledge to invest in their origin country. For example, looking at Turkish migrants who returned from Germany in the 1980s, (Dustmann and Kirchkamp, 2002[17]) propose that the reason why migrants return is that they expect higher returns from self-employment opportunities in their country of origin in the long run. Dustmann and Weiss (2007[18]) make a similar argument looking at migrants in the United Kingdom: migrants return if the human capital acquired in the host country has a higher return at home. Preferences for consumption at home and higher purchasing power in the country of origin are other factors motivating return.

A contrasting strand in the literature associates return migration to the failure to integrate in the destination country. The supposed intention behind migration projects is seen as maximizing utility by migrating to places that allow individuals to be more productive. Thus, if the expected increase in productivity cannot be realised, the individuals are likely to return. For Germany (Constant and Massey (2002[19]) and Canada (Lam, 1994[20]), two studies find that exposure to unemployment in the host country labour market increases the probability of return. For immigrants who find it difficult to join the labour market, however, access to a social security system in the host country can reduce their propensity to return. Reagan and Olsen (2000[21]), Jensen and Pedersen (2007[22]) and Nekby (2006[23]) obtain similar results for the United States, Denmark and Sweden, respectively.

The empirical evidence on return decisions suggests that ultimately, a complex interplay between home- and host-country related reasons motivates return. A large household survey in 11 countries of origin (OECD, 2017[24]), captured more than 3 000 return migrants. They reported their main reason for return (Figure 2.2). The main reason for return was an overall preference for the home country, followed by the failure to obtain legal status in the host country and difficulties in integration.

Figure 2.2. Reported reasons returnees came back to their home country

Source: IPPMD survey data. See (OECD, 2017[24]).

De Haas, Fokkema and Fihri (2015[25]) examine return intentions based on survey data of Moroccan migrants across different European countries. The survey indicates that structural integration through labour market participation, education and the maintenance of economic and social ties with receiving countries do not significantly affect return intentions. At the same time, investments in and social ties to Morocco are positively related to return, and socio-cultural integration in destination countries is negatively related to return migration intentions.

Social relations, both in the destination and origin country, are key in migrants’ return decision (Flahaux, 2017[26]). In their survey of (potential) return migrants in both destination and origin countries, Koser and Kuschminder (2015[15]) find that changes infamily relations were the most frequently mentioned reason for return. Constant and Massey (2002[19]) find that having a partner in the origin country increases propensity for return for all migrants, whereas having a partner in Europe reduces return probability. The presence of children reduces return probability. Aradhya (2018[27]), for example, finds that immigrant fathers of daughters in Sweden are less likely to return, with the explanation that they expect better opportunities for them in a more gender equal Swedish system. Ultimately, while social relations are a key factor influencing the decision to return, they are largely beyond the scope of policy interventions. A 2014 survey of returnees in Tunisia by the National Institute of Statistics found that the main reasons for return were to accompany family, to return home, but also – for 12% ‑ to get married (OECD, 2018[28]) (Hammouda, 2020[29]).

Figure 2.3. Factors influencing individual return decisions by level

Source: Based on Black et al. (2004).

Several studies show that reasons for initial migration are linked to return (Aydemir and Robinson (2008[30]) in Canada, Klinthäll (2006[31]) in Sweden, Shortland (2006[32]) in New Zealand, Bijwaard (2010[33]) in the Netherlands). Students are the most likely to return while refugees are the least likely, and students and labour migrants return earlier. As expected, work-related factors are more important in predicting return of migrants who left for economic reasons. In comparison with labour and family migrants, return decisions of (rejected) asylum seekers depend more strongly on structural factors in the origin country. For asylum seekers, economic factors are less important than political factors (Black et al., 2004[14]; King, 2000[34]; van Wijk, 2008[35]), although they also have socio-economic reintegration concerns (Klinthall, 2007[36]; Zimmermann, 2012[37]).

These findings have implications for return policy. Unsurprisingly, Brekke (2015[38]) finds that AVR uptake by rejected asylum seekers in Norway is highly dependent on political factors in the origin country, especially in (post‑)conflict contexts. In a somewhat similar study, Leerkes et al. (2014[39]), who examine AVR uptake amongst rejected asylum-seekers in the Netherlands, find that voluntary return is less common towards countries with low levels of freedom and/or safety and/or GDP. A recent study of Syrian refugees in Lebanon found that return intentions were contingent on the situation in Syria rather than conditions in Lebanon, including the socio-economic well-being of refugees and their access to services (Alrababa’h, Casalis and Masterson, 2020[40]).

Regarding AVRR more specifically, one question which has been investigated is whether financial incentives – cash payments, in particular, but also reintegration services and support – affect the decision to return voluntarily. The literature suggests that financial incentives alone do not significantly influence potential returnees’ decisions. Studies in the French context suggest that most beneficiaries of AVRR had already planned to return, considering the funding as a bonus rather than a decisive factor (Daum, 2002[41]). In Norway, return support appears to be one of the factors contributing to the decision, but not the principal factor (Sønsterudbråten, 2018[42]). A transnational study in the Netherlands and Nigeria underlines that among those who decided to return, the most recurrent reasons are the stress of being undocumented, the family and related ‘triggers’ that occurred in Nigeria, and the threat of deportation – rather than AVRR (Benhayyoun, 2018[43]). Black et al. (2004[14]) find that most of their respondents believed assistance would in any case not be sufficient to overcome more fundamental obstacles to return – most significantly insecurity and longer-term unemployment. Brekke (2015[38]) finds that “while return and reintegration programs may facilitate and increase the quality of assisted returns, [there is] no clear-cut link to the quantity of such returns”. During the global financial crisis of the late 2000s, several OECD countries – Spain, Japan and the Czech Republic – introduced cash payments to unemployed foreign workers as an incentive to return home, even if they held residence permits allowing them to stay. These programmes appeared not to affect the return decision – and indeed, many returned home without taking the incentives when the benefit was tied to a re-entry ban (OECD, 2009[44]).

The role of provision of information and return counselling has received little attention in the literature so far. Nonetheless, there is some evidence that uptake of AVR can be influenced by information from certain credible figures. Leerkes, van Os and Boersema (2017[45]) find that uptake of AVR is highest when “native counsellors” from the migrants’ origin community are employed. The use of “native counsellors” was most effective with irregular migrants living outside of state reception facilities who are not engaged in regular return counselling that occurs in the centres. This finding is likely related to higher trust placed in counsellors and the ability to communicate in the origin language. Apart from published studies, different agencies administering AVRR programming also conduct analyses within their programmes, surveying participants about their return decisions. Questions about factors leading to the decision to take assistance are usually included in programme evaluations, both internal and through external evaluators or researchers. For example, an evaluation following up 1 300 participants of a German programme asked about the influence of counselling and the availability of financial assistance on the return decision. For about one-third of respondents, counselling strongly affected their decision, while only one in seven reported that financial aid strongly influenced their decision Invalid source specified.. More comparative research on differences in AVR uptake based on different models of return counselling applied in EU destination countries is necessary (Kuschminder, 2017[46]).

Regardless of the information provided, return decisions are the “outcome of a complex mix of migrants’ choices and constraints on staying or moving” (van Houte, 2016[47]). The literature on migrants’ decision to return voluntarily depicts a complex process largely determined by factors beyond the scope of punctual policy intervention in the destination country, e.g. structural conditions in the country of origin or family relations. The existing evidence suggests that incentives set by AVRR, while they might accelerate return, cannot initiate a decision to return. Many of the factors which are found to affect the return decision have little to do with the incentives, support and conditions offered under the heading of “return and reintegration” policy. These include, for example, likelihood of access to international protection, the prospect of alternatives to protection such as leave to remain for other grounds, the conditions for migrants in an irregular situation, and the likelihood of ending up in removal procedures.

References

[40] Alrababa’h, A., M. Casalis and D. Masterson (2020), Returning Home? Conditions in Syria, Not Lebanon, Drive the Return Intentions of Syrian Refugees, Innovations for Poverty Action , Washington D.C.

[27] Aradhya, S. (2018), Diversity of Legacy. The Experience of Immigrants and their Descendants in Sweden, Lund University.

[30] Aydemir, A. and C. Robinson (2008), “Global labour markets, return, and onward migration”, Canadian Journal of Economics, Vol. 41/4, pp. 1285-1311.

[43] Benhayyoun, S. (2018), Here no king. There no king. Perceptions of Return among Undocumented Nigerians in the Netherlands and Returnees in Nigeria, Leiden University, Faculty of Humanities.

[33] Bijwaard, G. (2010), “Immigrant migration dynamics model for the Netherlands”, Journal of Population Economics, Vol. 23/4, pp. 1213-1247.

[14] Black, R. et al. (2004), Understanding Voluntary Return, Home Office Report 50/04.

[9] Borjas, G. and B. Bratsberg (1996), “Who Leaves? The Outmigration of the Foreign-Born”, Review of Economics and Statistics, Vol. 1/78, pp. 165-176.

[38] Brekke, J. (2015), Why go back? Assisted return from Norway, Institute for Social Research, https://www.cmi.no/publications/file/5800-why-go-back-assisted-return-from-norway.pdf.

[3] Cassarino, J. (2008), Return Migrants to the Maghreb Countries: Reintegration and Development Challenges.

[19] Constant, A. and D. Massey (2002), “Return Migration by German Guestworkers: Neoclassical versus New Economic Theories”, International Migration, Vol. 40/4, pp. 5-38.

[12] Cuthbert, N. and J. Song (2017), Removal of failed asylum seekers in Australia: A comparative perspective, Lowy Institute for International Policy, Sydney, https://www.lowyinstitute.org/sites/default/files/documents/Removal%20of%20failed%20asylum%20seekers%20in%20Australia.pdf (accessed on 7 August 2020).

[41] Daum, C. (2002), “Aides au “retour volontaire” et réinsertion au Mali : un bilan critique”, Hommes et Migrations, Vol. 1239/1, pp. 40-48.

[25] de Haas, H., T. Fokkema and M. Fihri (2015), “Return Migration as Failure or Success?: The Determinants of Return Migration Intentions Among Moroccan Migrants in Europe”, Journal of International Migration and Integration, Vol. 16/2, pp. 415-429.

[17] Dustmann, C. and O. Kirchkamp (2002), “The optimal migration duration and activity choice after re-migration”, Journal of Development Economics, Vol. 67, pp. 352-372.

[18] Dustmann, C. and Y. Weiss (2007), “Return Migration: Theory and Empirical Evidence from the UK”, British Journal of Industrial Relations, Vol. 45/2, pp. 236-256.

[10] European Commission (2017), On a More Effective Return Policy in the European Union - A Renewed Action Plan, European Commission, Brussels.

[26] Flahaux, M. (2017), “The Role of Migration Policy Changes in Europe for Return Migration to Senegal”, International Migration Review, Vol. 51/4.

[4] Gibney, M. (2008), “Asylum and the Expansion of Deportation in the United Kingdom”, Government and Opposition, Vol. 43/2, pp. 146-167.

[8] González, A. et al. (2018), “Preparation of survey tools for merged dataset”, Working Paper Series Temporary versus Permanent Migration, No. 14, Temper EU Project.

[29] Hammouda, H. (2020), Retour et Réintegration Durable en Tunisie, OECD, Paris.

[13] IOM (2019), 2018 Return and Reintegration. Key Highlights.

[1] IOM (ed.) (2019), Glossary on Migration.

[22] Jensen, P. and P. Pedersen (2007), “To Stay or Not to Stay? Out-Migration of Immigrants from Denmark”, International Migration, Vol. 45/5, pp. 87-113.

[34] King, R. (2000), “Generalizations from the History of Return Migration”, in Ghosh, B. (ed.), Return Migration: Journey of Hope or Despair?, International Organization for Migration.

[36] Klinthall, M. (2007), “Refugee Return Migration: Return Migration from Sweden to Chile, Iran and Poland 1973 1996”, Journal of Refugee Studies, Vol. 20/4, pp. 579-598.

[31] Klinthall, M. (2006), “Retirement Return Migration from Sweden”, International Migration, Vol. 44/2, pp. 153-180.

[15] Koser, K. and K. Kuschminder (2015), Comparative Research on the Assisted Voluntary Return and Reintegration of Migrants, International Organization for Migration, Geneva.

[46] Kuschminder, K. (2017), “Taking stock of Assisted Voluntary Return from Europe : decision making, reintegration and sustainable return : time for a paradigm shift”, EUI RSCAS Global Governance Programme, No. 31, European University Institute.

[20] Lam, K. (1994), “Outmigration of Foreign-Born Members in Canada”, The Canadian Journal of Economics, Vol. 27/2, p. 352.

[39] Leerkes et al. (2014), Rejected and departed from the Netherlands? A study into the backgrounds of the variation in assisted voluntary return among rejected asylum seekers, Wetenschappelijk Onderzoek- en Documentatiecentrum Cahiers 2014–03.

[5] Leerkes, A., R. van Os and E. Boersema (2017), “What drives ‘soft deportation’? Understanding the rise in Assisted Voluntary Return among rejected asylum seekers in the Netherlands”, Population, Space and Place, Vol. 23/8, pp. 1-11.

[45] Leerkes, A., R. van Os and E. Boersema (2017), “What drives ‘soft deportation’? Understanding the rise in Assisted Voluntary Return among rejected asylum seekers in the Netherlands”, Population, Space and Place, Vol. 23/8.

[23] Nekby, L. (2006), “The emigration of immigrants, return vs onward migration: Evidence from Sweden”, Journal of Population Economics, Vol. 19/2, pp. 197-226.

[2] Newland, K. and B. Salant (2018), “Balancing Acts: Policy Frameworks for Migrant Return and Reintegration”, Migration Policy Institute.

[6] OECD (2020), International Migration Outlook 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/ec98f531-en.

[28] OECD (2018), Talents à l’étranger: Une revue des émigrés tunisiens, OECD Publishing, https://dx.doi.org/10.1787/9789264308855-fr.

[24] OECD (2017), Interrelations between Public Policies, Migration and Development, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264265615-en.

[44] OECD (2009), International Migration Outlook 2009, OECD Publishing, Paris, https://dx.doi.org/10.1787/migr_outlook-2009-en.

[7] OECD (2008), “Return Migration: A New Perspective”, in International Migration Outlook 2008, OECD Publishing, Paris, https://dx.doi.org/10.1787/migr_outlook-2008-7-en.

[21] Reagan, P. and R. Olsen (2000), “You can go home again: Evidence from longitudinal data”, Demography, Vol. 37/3, pp. 339-350.

[32] Shortland, P. (2006), People on the Move: A Study of Migrant Movement Patterns to and from New Zealand.

[42] Sønsterudbråten, S. (2018), Fafo-rapport 2018:16 Assistert retur En kunnskapsstatus, Fafo, Oslo.

[16] Strand, A. et al. (2011), Between two societies Review of the Information, Return and Reintegration of Iraqi Nationals to Iraq (IRRINI) programme, CMI Report.

[11] Thompson, C. and A. Calderón (2019), More Immigrants Are Giving Up and Leaving the U.S. | The Marshall Project, The Marshall Project, https://www.themarshallproject.org/2019/05/08/more-detained-immigrants-are-giving-up-court-fights-and-leaving-the-u-s (accessed on 7 August 2020).

[47] van Houte, M. (2016), Return Migration to Afghanistan, Springer International Publishing, Cham.

[35] van Wijk, J. (2008), Reaching out to the unknown. Native counselling and the decision making process of irregular migrants and rejected asylum seekers on voluntary return, International Organization for Migration.

[37] Zimmermann, S. (2012), “Understanding repatriation: refugee perspectives on the importance of safety, reintegration, and hope”, Population, Space and Place, Vol. 18/1, pp. 45-57.