Livelihoods and survival strategies pursued by international return migrants in Addis Ababa, Ethiopia

Abstract In Ethiopia as far as the livelihood and survival strategies pursued by international returnees is concerned, such scientific study is scanty or non-existent. The study was devoted to explore the livelihoods and survival strategies chased among the returnees in Addis Ababa to provide scientific evidences and insights for decision making and future intervention efforts. Data were collected from 402 randomly drawn sample returnees via a cross-sectional survey. The supplementary data were collected from purposively selected focus group discussion and semi-structured interview participants. Descriptive statistics and Ordinal Logistic Regression Model (OLRM) were employed to analyze the data. The study revealed that while unemployment was found to be the most categorized livelihood strategy by which returnees were hallmarked and rated above the scale mean on the three-point likert scale (mean = 2.18, Standard Deviation/SD = 0.91); wage-employment and self-employment were rated below the scale mean (mean = 1.41, SD = 0.61; and mean = 1.39, SD = 0.56) respectively, which reflects that majority of the study population weren’t involved in livelihood strategies common in urban settings. OLRM results confirmed that educational attainment and age of the returnees made the first and the second largest contributions in predicting wage-employment in the positive and negative directions with beta coefficients (β = 0.985; and β = |-0.611| respectively at (p ≤ 0.000) level of significance. A notable proportion of Ethiopian returnees deemed to be unemployed and not engaged in any urban livelihood activities and deprived of enjoying the potential benefits obtained from self-employment and wage-employment activities in the study area.


PUBLIC INTEREST STATEMENT
Over the last decades, return migration is considered as the concluding moment of the migratory process. Its dimensions, modalities, and other major aspects are poorly understood. This study has been conducted to broaden understanding of the link between livelihood strategies and return migration among international returnees in Ethiopian contexts. Returnees are confronted by many livelihood risks such as lack of access to resources that are needed for sustainable livelihoods. Return migrants had poor purchasing power, which may jeopardize their improved resilience to vulnerability to make a better living. The vast majority of returnees were not engaged in any occupation category at their homeland upon return. Return migration should not be viewed as the aftermath of the migration process that simply concludes migration. Rather it is an important stage in the migration cycle to the well-being of the return migrants and the welfare of the state.

Introduction
In our modern age, return migration remains a demanding agenda. Return migration has been receiving growing attention (Asiedu, 2005;Cassarino, 2004;Rodriguez & Egea, 2006). It has been a central point of discussions at the global level since migration was included in the 2030 Agenda for Sustainable development (Newland & Salant, 2018); and it has become one of the priority agendas of a growing number of emigrant origin countries (ILO, 2019;OECD, 2017). Nonetheless, over the last decades, it is the concept entangled with conceptual difficulties and it has even less consensus unanimity (Battistella, 2018;IOM, 2019Kuschminder, 2017. Return migration is the one which is least well understood among the different stages of the migration process (Bastia, 2011;IOM & ACP, 2013); and many aspects of the return migration remain in need of understanding (Battistella, 2018;Kuschminder, 2017;Wahba, 2014). Return migration is often badly understood, and its dimensions and modalities are often poorly understood (Battistella, 2018).
Evidences have heralded that return migration is a dynamic phenomenon. Undoubtedly, it is not always a process of simply "going-home" and uni-directional means (Arowolo, 2000;IOM & ACP, 2013;Wahba, 2014;Segal, 2016). Coming homeland as returnee can be harder than leaving and the process is grappled with severe obstacles. Return migration necessitates a complicated decision-making process demanding a variety of factors (IOM & ACP, 2013). Creating sustainable livelihoods for returnees can be difficult due to structural challenges and low economic opportunities in the country of origin (IOM, 2015;Kodom & Dako-Gyeke, 2017;Segal, 2016); and returnees previous social network has either been lost or damaged as they have spent a significant time in exile (ILO, 2013). Moreover, return migration is common, albeit our knowledge of its extent is hampered by lack of data, and whether those exiting return to their home country or move onto another destination is rarely known (GMG, 2017;Wahba, 2014); and return migration is difficult to measure as data on the numbers of return migrants are not well recorded (Jensen & Pedersen, 2007;OECD, 2008;Wahba, 2014).
Return migration has been understudied for a long period of time. In migration studies, return migration is enormously a neglected area and has been less researched (Arowolo, 2000;Bastia, 2011;Kuschminder, 2013); and "return migration is the great unwritten chapter in the history of migration" (King, 2000, p. 7). However, since 2000, studies that explore return migration from multiple perspectives have been gradually increased. For instance, imagined and provisional return (Oxfeld & Long, 2004), second generation return (Tsuda, 2009), contributions to theory (Ammassari, 2009;Cassarino, 2004), the impacts of Assisted Voluntary Return (AVR) on returnees and their countries of return (Collyer, 2012;Majidi, 2012), and many other empirical studies. Despite the increase in research on different forms of return migration, a consensus on linking the economic life such as livelihood and survival strategies of the return migrants and the effects of return migration is still lacking.
Return migration has been viewed as a double-edged sword that affects both returnees and their country of origin. It has both positive and negative profound personal consequences for the returnees and a bundle of implications for their homeland (Newland & Salant, 2018;Wahba, 2014). On the one hand, the returnees can use their skills and knowledge gained in the country of destination to their advantage upon return, and the country of origin can benefit much from these inputs, turning "brain drain" into "brain gain" (Martin & Radu, 2012); and they can act as key agents for economic and social changes by bringing back their financial resources, and introducing innovations and technologies to their homeland (ADB, 2020). On the contrary, return migrants may comeback with skills that do not match the demands of the home economy and might not be able easily to reintegrate back in the labour market upon return (Hamdouch & Wahba, 2012;Lucas, 2008); and returnees' previous social network has either been lost or damaged, they are obliged to rebuild their social network and develop their means of livelihood in a new socio-economic and political setting (ILO, 2013;Battistella, 2018;Wahba, 2014).
It is a multitude process that depends on the environmental and structural capacities; development and economic opportunities available in the homeland amongst others (Dumont & Spielvogel, 2008;IOM, 2015). The livelihoods of returnees heavily rely on the prevailing socioeconomic and political environment; any of the dimensions of such environmental change may also change their livelihood status and survival strategies (ILO, 2013;Jacobsen, 2014); and return migration particularly the forced one and food insecurity and other vulnerability can feed into each other (IFPRI, 2017;Oxfam, 2018). In our globalized world, migration has become inevitable and almost touches every corner of the globe (Mulugeta & Mekonnen, 2017). However, in the overall migration landscape, the Horn of Africa has unique challenges as the nature of human movements in the region is irregular including return migration (Fransen & Kuschminder, 2009;Schroder, 2015;Siegel et al., 2016). Likewise, latest data demonstrate a rise in return migration in Ethiopia and the country is featured by frequent international return migration. Over the past decades, return migration has increased to Ethiopia (Kuschminder, 2013). Moreover, ILO (2019) elaborated the situation in a condensed form hereunder: . . . Ethiopia presents a case where an origin country has been grappling with increased returns of migrants, mainly as a result of mass deportations of nationals from abroad, especially from Saudi Arabia with 163,000 forcibly returned between November 2013 and March 2014. Ethiopian migrants are also returning voluntarily or forced to Ethiopia from other parts of the world. This poses serious challenges for reintegration as Ethiopian returnees are often in situations of vulnerability (ILO, 2019, p. 24).
Furthermore, very recently, from November 2018 to 2020 about 572,096 and 6,418 officially registered irregular cross-border migrants returned to Ethiopia in general and to Addis Ababa in particular, respectively (Ethiopian Ministry of Labour and Social Affairs/MoLSA, 2021), which in turn may imply that, Ethiopia is a hub for return migration. Having this in mind, an attempt was made to review some empirical studies that have been undertaken in Ethiopia on the issue. A study conducted in three countries (Ethiopia, Mali & Bangladesh) aimed at assessing the linkages between migration and sustainable livelihoods concluded that international migration is seen as just one of the livelihood strategies in Ethiopia (Fransen & Kuschminder, 2009); Ethiopian returned migrants were swamped by hopelessness and painful experiences as the result of physical abuses, restrictive mobility, and a variety of harassment by respective employers (Tadesse et al., 2019); and a survey study carried out on asset accessibility in Southern Ethiopia, Wolaita Sodo based on internal migration on urban migrants concurred that physical, human, social, financial, and natural capitals were ranked in decreasing order (Esayas et al., 2018).
It is evident from the preceding discussions that returnees are largely challenged by a tremendous number of setbacks upon their return. The main emphasis of previous studies was on understanding and analyzing decisions for departure, failing to acknowledge the link between return migration and livelihoods of the returnees in Ethiopia. The issue has no yet well documented in literatures, and only little is known in Ethiopian context. Consequently, development actors at various levels that are engaged in migration policy formulation and administration face critical knowledge gaps for which evidences are lacking to draw on sound livelihood interventions to curtail the resultant challenges of return migration. The main purpose of this study is therefore is to examine the livelihoods and survival strategies employed by Ethiopian return migrants in Addis Ababa to broaden understandings on the link between return migration and the livelihoods of the returnees. Moreover, the study is targeting at informing pragmatic and effective interventions and thereby enhances the livelihood of the target population. As a way to realize these objectives, we pose the following two basic research questions: 1) How do the returnees make a living, and in what types of livelihood strategies do they are being engaged?
2) What major background characteristics of the returnees are affecting their livelihood strategies?

The Sustainable Livelihood Approach as a conceptual framework
The present study is subjected to sustainable livelihood framework. Recently, the Sustainable Livelihoods Approach (SLA) developed by Department for International Development/DFID is the most cited framework and frequently adapted by many international organizations (CARE International, Oxfam, the World Bank and UNDP) and scholars to succinctly assess a household's methods of survival (Cordero, 2016;Rakodi & Lloyed-Jones, 2002). The approach has been endorsed by Migration Policy Institute (MPI) to examine and understand the impact of livelihood opportunities and interventions on migration patterns (Fratzke & Salant, 2018). The framework can help to identify the relationships that exist between assets and livelihood strategies (Carney, 1998;Ellis, 2000). It is a practical and analytical tool that outlines a holistic approach in addressing the analysis of livelihoods, as well as to the design and monitoring of the livelihood interventions (FAO, 2002;; and the approach sets out to be people-centered and holistic, and to provide an integrated view of how people make a living within evolving social, institutional, political, economic and environmental contexts (Bebbington, 1999;Carney, 1998).
Within the framework, three factors affect the scope of livelihood opportunities available to households or individuals (returnees in this case): (a) livelihood assets (such as human, financial capital, etc.,); (b) vulnerability contexts (describing concerns such as availability of employment, conflict, or environmental degradation); and (c) structures and processes (such as the legal and policy frameworks in different countries) (Fratzke & Salant, 2018); and it integrates three key concepts of capability, equity and sustainability (Chambers & Conway, 1992). In the approach not only the assets but also the individual are central to the sustainable livelihood approach, and it also illustrates the impact of policies, institutions as well as the vulnerability context on the individual (Cordero, 2016;Overseas Development Institute/ODI, 2002). Until recently, the livelihoods approach originated in rural contexts and has been worked on largely in the context of rural development, however it is now being increasingly discussed for urban settings and there is nothing inherent in urban settings to prevent their application there (DFID & World Bank, 2002;ODI, 2002). In other words, needless to say, the approach is just as relevant in the urban context as in the rural.
Broadly speaking, the application of the approach is two-fold: in developmental delays or problems; and in research spheres. The approach is widely applied in research (de Haan & Rogaly, 2002;Köberlein, 2003), and a means whereby development problems is analyzed, and policies and programmes are designed to realize the principal goal of poverty reduction (Rakodi & Lloyed-Jones, 2002). Accordingly, the SLA was conceptualized within urban contexts to build the current study on it. The rationale behind is that SLA has the capacity to better maximize understanding the livelihoods of the poor by taking the households including individual returnees as its main unit of analysis. Furthermore, the approach has due regard on the strategies that the poor people employ to make a living (Ellis, 2000;ODI, 2002;Tolossa, 2008).
To this effect, the present study is not an endeavor made to explore and interpret in great-depth all aspects of the SLA as applied to Ethiopian return migrants in the study area. The framework conceptualized above is only confined to exploring the livelihoods and survival strategies of the returnees in the area under consideration.

Research setting
The study was conducted in Addis Ababa, Ethiopia (see Figure A1 Appendix 2), which is located on a well-watered plateau surrounded by hills and mountains, in the geographic centre of Ethiopia (Addis Ababa City Administration/AACA, 2015; Addis Ababa Plan & Development Commission/ AAPDC, 2020). It is located at geographical coordinates: between 8°55"and 9°05" North Latitude and between 38°40"and 38°50" East Longitude. Its average elevation is 2,500 meters above sea level, and hence has a fairly favorable climate and moderate weather conditions. Addis Ababa is the capital and largest city of Ethiopia and it is the educational and administrative center of the country (AADPC, 2020;HABITAT, 2008). Moreover, it is the seat of the African Union (AU) and the United Nations Economic Commissions for Africa (UNECA), as well as various other continental and international organizations. It is often referred to as "the political capital of Africa" for its historical, diplomatic and political significance for the continent (HABITAT, 2008). The total land area of Addis Ababa is about 527 km 2 or 54, 000 hectors; and the city has a complex mix of highland climate zones, with temperature differences of up to 10°C, depending on elevation and prevailing wind patterns (World Meteorological Organization, 2019).
It is a chartered city having three layers of government: City government at the top, 10 sub-city administrations in the middle (of course, Lemi Kura, the 11 th sub-city isn't considered in the study as it is the newly emerging sub-city that isn't well established), and 121 woreda administrations at the bottom (Addis Ababa Plan & Development Commission/AAPDC, 2020). In this study, Addis Ababa is taken as a geographical unit of analysis as it is one of the leading hotspot areas where a significant number of international returnees are found in Ethiopia (MoLSA, 2021). Moreover, in today's vibrant economy, food market entered an era of instability and the competition to enter the labor market is very keen, and the rising of food prices and energy are hitting hard in urban areas and the livelihoods of urban people are to a large extent rely on labour markets on the other hand (de Haan et al., 2002). Furthermore, livelihoods tend to be at their most complexes in urban areas, with households drawing on a wide variety of activities to capture income and other resources (Beall, 1997;Rakodi, 1999;Walkker et al., 2001). In sum, the above factors can cause incalculable harm to the livelihood and survival strategies of the urban returnees which further exacerbated in capital cities like Addis Ababa than cities found elsewhere in Ethiopia.

Research approaches
In the present study, mixed methods research design was employed to associate both qualitative and quantitative approaches. Mixing both qualitative and quantitative data in a single study allows for the limitations of each approach to be neutralized while strengths are built upon thereby providing stronger and more accurate inferences (Bryman, 2006;Creswell & Clark, 2007;Creswell, 2009;Tashakkori & Creswell, 2007). Among the major types of mixed methods design, "Concurrent Embedded Design Approach" was used as the primary research design of the study. Concurrent embedded design is well-known for its use of a single data collection phase and a primary method guides the project and the secondary method offers a supporting role in the procedures (Creswell & Clark, 2010;Creswell et al., 2003;Johnson & Onwuegbuzie, 2004;Johnson et al., 2007).
The quantitative data were given more weight and the qualitative data were embedded within the former one to substantiate the numerical data obtained via the survey questionnaires. The target population in the study were irregular cross-border migrants who returned to Addis Ababa, Ethiopia, between the years 2016 and 2019 having at least one-year of stay in Ethiopia after return. Out of ten sub-cities clustered into inner-urban and peri-urban areas, Addis Ketema and Kirkos; and Akaki Kality and Kolfe Keraniyo were selected from the former and latter areas respectively using purposive sampling technique as the sample sub-cities of the study. The sample size of the study population was determined by employing Yamane (2001) sample size determination formula which assumes 50% (p = 0.5) variability and 95% confidence level with±5% precision error.
Where, n = sample size; N = population size; and e = level of precision.
As a whole, based on the formula, out of 5,228 returnees a sample of 416 returnees were drawn randomly for final sample of the study for structured survey questionnaires. Survey questionnaires, semi-structured interviews (SSIs), and focus group discussions (FGDs) were used as tools of data gathering. Using both types of data enable the researcher to expand an understanding from one method to another, to converge or confirm findings from different data sources, etc. (Bryman, 2006;Creswell, 2009;Tashakkori & Creswell, 2007). The survey questionnaire was pre-tested to check for its internal consistency, and a Cronbach's alpha (α) of 0.743 was obtained. The survey questionnaire was translated into the local Amharic language and tested for face validity. A principal component analysis (PCA) was also carried-out to reduce the factors into a smaller set of components and to summarize data so that relationships and patterns can be easily interpreted and understood (Abdi & Williams, 2010;Everitt, 2004;Field, 2009;Gray, 2017;O'Rourke & Hatcher, 2013). Kaiser-Meyer-Olkin (KMO) and Bartlett's test were undertaken to check the sample adequacy and the suitability of data for factor analysis respectively.
The KMO measure of sampling adequacy calculated was 0.760, above the commonly recommended value of 0.60 (Field, 2009;Hair et al., 2010), and Bartlett's test of sphericity was significant (χ2 (66) = 5786.532, p = 0.000 < .05) and was considered adequate for performing a factor analysis (Hair et al., 2010;Pallant, 2010;Tabachnick & Fidell, 2007). Descriptive statistics and ordinal logistic regression model were employed to analyze quantitative data. The qualitative data were also transcribed, coded and interpreted thematically. The authors were committed to meet the ethical standards set forth by the APA from inception to completion of the study to protect and ensure the dignity and welfare of all participants.

Socio-cultural characteristics of the returnees
As can be observed from Table 1, male returnees constituted 16.2% of the final sample and female returnees 83.8%, which in turn shows that the number of women returnees assumes higher figure than men. In other words, majority of the returnee group is female-dominated in Ethiopian context in general and in the study area in particular. Thus, unlike the foregoing finding, the previous studies conducted by (Elbadawy, 2011;Songsore, 2003) found that males were more likely to migrate abroad compared to females that seems contradict with the above one. As to the marital status of the respondents, relatively as a whole unmarried respondents took a greater share than the rest respondents found in other marital status and account for 51.0%. Besides, the overall mean age of the sampled returnees of the study was found to be about 33 years; and the male and female mean age of the respondents were approximately 34 and 32 years respectively. A notable proportion of the respondents (50.6%) were in the age group ranged between 20-40 years, which in turn may indicate that majority of the study population were young adults who forced to move away irregularly. In several Sub-Saharan African countries, including Ghana, many youth are compelled to migrate abroad in search of better livelihood opportunities (Dako-Gyeke, 2015; Kodom & Dako-Gyeke, 2017), which may indicate the congruency between the two findings.
Moreover, as noted in Table 1, the role of the respondents in the family was found to be: Household Member/HHM 222(55.2%) and Head Household/HHH 180(44.8%) which may indicate that relatively majority of the participants were household members in their respective family. Regarding the educational status of the respondents, as a whole, about 47% of the returnees were only attending secondary school education. This finding seems to have some congruency with the finding of other authors. For instance, the highest level of education attained by the Ghanaian returnees was senior secondary school education, with majority of them completing junior high school (Kodom & Dako-Gyeke, 2017).
Table 1 further reveals that the percentages of returnees who had first Degree and Diploma were found to be only about 3.7% and 7.7% respectively. This may show that most returnees are deprived of getting further education, which in turn inevitably may negatively affect them to respond to a rapidly changing knowledge-based vibrant economy of the modern age to make a better living and viable livelihoods in an urban environment.

Economic characteristics of the returnees
As evident from Figure 1, the overall mean income of the respondents at abroad and homeland was found to be about 6233.46 ETB and 1038.06 ETB respectively. The overall mean income of the study population in abroad is about six times the average income at homeland that may signify that returnees having a significant number found in the study area had poor purchasing power, which may jeopardize returnees' access to services and improved resilience to vulnerability, such as food security or sustainable use of various resources to make a better living and viable livelihoods. The medians monthly income of the study population at abroad and homeland were found to be about 6,000 ETB and 800 ETB respectively.
Furthermore, as noted in Figure 1, the mean income of the respondents at abroad was about 6147.69 ETB and 6250.00 ETB for male and female respondents respectively. On average the  (2021) homeland income on monthly basis of the male respondents 1938.46 ETB and female respondents 864.39 ETB suggesting income differentials and gender pay gap among the two groups and female earnings is above half of male when disaggregating the results by gender. This finding is in harmony with the finding of Schuerkens (2010) which states that in most nations discrimination in employment is remained entrenched and women still earn less than men. By taking on average the current exchange rate of one USD for 50 ETB, male ($1.29) and female returnees ($0.58) earned a day which is below the threshold that has been defined by the World Bank for extreme poverty ($1.90 per person per day) (World Bank, 2015).

Occupations of the returnees at abroad and homeland
On the one hand, the results reported in Table 2 point out that domestic work is the most predominant occupation category in which returnees were engaged in their respective destinations, and the vast majority of the respondents were not engaged in any occupation category at their homeland upon return on the other hand. In numerical terms, 76.4% and 52.2% the respondents were engaged in domestic work when they were at abroad and without any kind of job at origin country upon their return respectively. Moreover, company employee as an occupation was the most suffered occupation and the percentages of returnees who were working as employee at abroad and homeland were below 1.0% and 3.7% in the order mentioned which directly related with their education levels as a whole in the study area.

Livelihood and survival strategies of the returnees
As depicted in Table 3, returnee respondents reported that: no job (not engaged in any livelihood activities), daily labor, and vendor and associated petty trades were attached the greatest values above an average on the three Point Likert Scale and ranked as the most three major livelihood strategies pursued by returnees in the study area with the mean values of 2.18, 2.17, and 1.84 respectively. This shows that the vast majority of a group of returnees were without any kind of job in Addis Ababa, Ethiopia upon their return, which in turn may negatively affect the livelihood of the returnees and their respective family members. This is because the mean of returnee respondents on "No job" was slightly greater than the scale mean when compared with the mean values of the rest items, taking the scale mean 2.00 as a point of reference.
However, respondents found out that amongst all designated 12 potential livelihood and survival strategies that need to be practiced in the study area, the following three sources of livelihoods: metal works, wood works, and wholesale with the mean values of 1.31, 1.30, and 1.24 respectively as the three least livelihood strategies that were materialized in the study area, which is below an average on the indicated scale. That is, majority of returnee respondents of the study rated their livelihood strategies between the ranges of "not at all engage" and "engage rarely". Moreover, an attempt was made to condense the aforementioned livelihood strategies into the following three major categories (see Table A1): self-employment, unemployment, and wageemployment based on the PCA and the responses obtained are analyzed as follows.

Summary of the categorized occupation of the returnees
As indicated in Figure 2, returnee respondents singled out that unemployment is the most dominant categorized livelihood strategy by which the study population were hallmarked with the values of (mean = 2.18, SD = 0.91) rated above the scale mean. However, wageemployment and self-employment were rated below the scale mean by the returnees with the values of (mean = 1.41, SD = 0.61; and mean = 1.39, SD = 0.56) respectively, which reflects that majority of returnee respondents were unemployed and deprived of enjoying the potential benefits obtained from the two remunerative livelihood strategies: wage-employment and self-employment which complements the data obtained via qualitative research approach.
In other words, majority of the study population is featured by lack of required assets to enter two more remunerative livelihood strategies indicated and actively seeking employment. The above findings are not congruent with the following result of study: a large number of poor urban households are opportunistic and diversifying their sources of income through a variety of income-generating activities (formal waged employment, informal trading and service activities concurrently (DFID, 2001;Walkker et al., 2001).
Besides quantitative research approach, based on data obtained via semi-structured interviews (SSIs) and focus group discussions (FGDs) qualitative data analysis was undertaken to corroborate the quantitative findings. Correspondingly, the data are analyzed hereunder: Asked about how to make a living upon her return, a 35 years old female returnee explicated the situation as follows: . . . Nowadays, getting viable job is very difficult upon return. There are no regular sources of income to make a living and to support our families. Consequently, economic problems are bottlenecks to discharge our responsibilities of caring and supporting our family members. Post-return situation is characterized by decrease in various opportunities for livelihood strategies and many returnees are entering into small business activities mainly on daily basis in private sectors to secure their family income in Addis Ababa. The reintegration process as a whole is poor. Generally, sustainable livelihood upon return is hindered by the absence regular income, though the issue may vary from returnee to returnee depending on the economic background of the respective family at homeland. So, government should work at macro, middle as well as micro levels to improve the economic problems such as unemployment and underemployment (29/7/2021).
Still one of the male returnee interviewee, narrated the issue in his words as follows: . . . I migrated in irregular manner to change my livelihood pattern and way of living. Unfortunately, I returned back due to a number of reasons. This time urban labour market is highly diversified and competitive in Addis Ababa. Some jobs require better levels of education and skills. For unskilled and illiterate dweller/returnee like me job opportunities are unreliable, irregular and subject to high seasonal variation. Sometimes, let alone formal jobs, it is very difficult to obtain the non-formal ones on daily basis in sustainable manner due to high unemployment rate. Though, non-formal economic activity is the main means of living for majority of the dwellers in Addis Ababa, its capacity to accommodate the ever growing urban population has been diminished. Due attention should be given by Addis Ababa city administration so as to create ample job opportunities for the dwellers unless re-migration will continue and unabated irrespective of its resultant challenges (30/7/2021).
The above results disclosed that majority of the interviewees reported that currently both formal and non-formal jobs as the two major economic activities are non-abundant and even non-existent to absorb the urban dwellers in general and the returnees in particular due to high unemployment rate in the country including Addis Ababa. Both detrimental economic situation and poor reintegration process upon return orient the return migrant to go abroad again. Furthermore, as indicated by most of the interviewees measures should be taken for the actual creation of job opportunities for citizens as a whole and for returnees in particular to reintegrate them into the community. They also forwarded that government bodies must work towards economic activities operating within the official legal framework that can attract individual returnees for employment and ultimately curb down the outward looking for re-migration. These results corroborate the findings of the quantitative data analysis secured through survey questionnaires.
Further analysis of data obtained via focus group discussion (FGDs) was conducted to substantiate these results. The section to follow is dealt with the situation as follows.
During the FGDs, the participants shared the same ideas and reported the following: The livelihood schemes are ill-treated not only in our country including Addis Ababa, but in almost all developing regions of the world as informal employment represents the lion's share of the job opportunities. Developing regions are highly entangled with immense problems to carry out livelihood schemes for their respective returnees. The main problems were and/are: poor linkage between intentions and actions, between policy and actions, loose linkage between the grassroots level organs and higher officials, absence of mandates for the concerned bodies to implement the schemes properly, and scarcity of resources for designing effective and targeted livelihood strategies for an overwhelming number of returnees who came back to home on frequent basis. Though majority of the returnees are without any kind of formal jobs, the informal economy offers some sort of opportunities for improving livelihoods of the returnees. In other words, some returnees are being absorbed in the informal economy that constitutes: informal sector (informal self-owned enterprises or household enterprises, and the like); and informal employment (informal domestic workers, broker, laborer, and so on) (31/7/ 2021 and 1/8/2021).
This may imply that as far as livelihood strategies of the returnees are concerned, majority of the respondents reported the sources of livelihoods as serious problem and impediments to make a living, which need more immediate attention and an attempt is to be made to apply sound job opportunity schemes in Addis Ababa and in Ethiopia at large. This finding is in harmony with the study conducted: informal employment represents more than half of the opportunities in developing regions of the world; and in the years between 2010 and 2014, the informal economy constituted 49% of non-agricultural employment in Northern Africa; 57% in Latin America and the Caribbean; 66% in Sothern and South Eastern Asia and up to 74% in Sub-Saharan Africa and informal employment in the informal sector rises approximately equal to 80% in Sub-Saharan Africa (OECD, 2017).
Moreover, the factorability of the livelihood and survival strategy items and criteria for the factorability of a correlation were examined. Firstly, the Kaiser-Meyer-Olkin measure of sampling adequacy calculated was 0.760, above the commonly recommended value of 0.60, and Bartlett's test of sphericity was significant (χ2 (66) = 5786.532, P = 0.000 < .05). Secondly, factors whose average communality is greater than 0.60 and principal components with eigenvalues above one were considered based on Kaiser's criteria (when the sample size exceeds 250 and the average communality is greater than 0.60, then retain all factors with eigenvalues above one). Thirdly, confirming that each item shared some common variance with other items. Given these overall indicators, factor analysis was deemed to be suitable with variables associated with livelihoods and survival strategies in the study area.
To this effect, principal component analysis was conducted to single out and compute composite scores underlying the livelihoods and survival strategies of returnees in the study area. The factor loadings after rotation using a significant factor criterion of 0.50 and above (see Table A1 Appendix 1). Accordingly, out of 12 initial variables 10 variables were retained and wholesale, and vendor and associated petty trades were removed from the final analysis as they were not significant in the model. In addition, an attempt was made to identify common themes and a three factor structure for variables was evident. Correspondingly, the first six variables that loaded highly on (factor 1) seem to all relate to the state of working for oneself instead of an employer to earn an income by business directly. Therefore, they were labelled as self-employment livelihood activities. The two variables (no job, and daily labor) loaded highly on (factor 2) all look to be associated with the form of contingent work with no promise that more work will be available in the future; hence, denominated as unemployment.
Finally, the last two variables loaded highly on (factor 3) all contain bundles of tasks performed by employees under administrative job titles; thus, they were labelled as wage-employment. Overall, the factor analysis of the ten variables with varimax rotation yielded the following three distinct factors: self-employment; unemployment; and wage-employment as the three categorized livelihood strategies (see Table A1 Appendix 1).

The Effects of predictor variables on latent variables
In the analysis, an attempt was made to examine the relative contribution of the predictor variables-general characteristics of the returnees to the criterion variables-the three underlying livelihood strategies. The details and concluding remarks, therefore, are presented here.
As evident from Table 4, the relationship between the categorical variables and/or ordinal variables was examined to look for associations. Accordingly, an ordinal logistic regression was performed between the criterion variable-self-employment and the predictor variables (personal, economic characteristics, and modality of returns of the returnees). Amongst the predictor variables indicated in Table 4, only two variables: role in the family; and income at homeland had statistically significant association with self-employment with the odds ratio of (Exp (B) = 10.83, p < 0.05; and Exp (B) = 0.04, p ≤ 0.001) respectively. This indicates that, there is a strong evidence to reject the null hypothesis that self-employment as a livelihood strategy and role in the family and income of the returnees in the homeland are independent and there is, therefore, some relationship between the predicator and criterion variables.
While the association between self-employment and role of returnees in the family is positive (β = 2.382); and on the contrary the correlation between self-employment and income in the homeland is negative (β = −3.360) as indicated via their respective coefficients. On examining the contributions made by the independent variables in the model, it was found that the scores received from income at homeland made the largest contribution with (β = |-3.360|; which is followed by the score received by role in the family among returnees with beta coefficient (β = 2.382). As illustrated in Table 5, age and educational status of the returnees had a negative and positive statistically significant association with wage-employment respectively at (p ≤ 0.000) level of significance. The results of analyses further disclose that the rest explanatory variables tested in the model weren't indicating any significant association with the two aforementioned categorized livelihood strategies. The final observation from Table 5 is about the relative importance of independent variables in predicting the dependent variables. Accordingly, the results of analysis indicated that educational attainment and age of the returnees made the first and the second largest contributions in predicting wage-employment with values of beta coefficients (β = 0.985; and β = |-0.611| respectively.

Conclusions and recommendations
The main intent of the study was to investigate the livelihoods and survival strategies of Ethiopian returnees in the study area. The study was built on the discourses of Sustainable Livelihood Approach (SLA) in urban context as the main framework to better understand the current situation of the issue under way. The study could contribute for further understandings of the nexus between livelihood strategies and return migration in Ethiopian contexts, with the aim of providing insights for policymakers and development actors to develop pragmatic strategies and ultimately to reverse and combat the livelihood and survival challenges of the returnees in the study area. Moreover, the study has paramount importance to gauge the salience of "the economic life of international returnees upon return" in the academic literature. The results of the study also offer lines of enquiry worthy of further investigations in the study area as well as in the global south at large.
The study has pinpointed out that on average the homeland income on monthly basis of the male and female returnees suggesting income differentials and gender pay gap among the two groups and female earnings is above half of male when disaggregating the results by gender. The results of analysis vividly revealed that the sources of livelihoods deemed to be major impediments to make a living among the vast majority of the study population, which require more immediate concerted actions and an attempt is to be made to apply sound job opportunity schemes in the study area. Company employee as an occupation was the most suffered occupation and the number of returnees who were working as employee upon return constituting a minimum percentage. Moreover, the results of ordinal logistic regression analysis depicted that the association between self-employment and role of returnees in the family is positive; while the correlation between self-employment and income in the homeland is negative.
In a nutshell, conclusions drawn from the present study is three-fold: there is no definite and well-defined major livelihoods and survival strategies for the study population and they were seriously hampered by the absence of regular sources of livelihoods; an overwhelming number of a group of returnees were unemployed and not engaged in any urban livelihood activities and deprived of enjoying the potential benefits obtained from the two remunerative livelihood strategies (self-employment and wage-employment); and currently both formal and non-formal jobs as the two major economic activities are non-abundant to absorb the urban return migrants in the study area. Therefore, targeted interventions are required to entitle and warrant the urban poor returnees with ample livelihoods opportunities via economic inclusion and job creation schemes and thereby building sustainable livelihoods.
It is believed that the present study is not free from some drawbacks. First, some pertinent data were collected by relying on the opinions and perceptions of the returnees in which the methods employed may limit the robustness of the results of the study. Second, in this study, data were gathered at snapshot and did not consider the dynamic nature of the livelihood situations of the returnees through time that can be better addressed via longitudinal studies, which obviously may have an impact on the study. Therefore, the aforementioned limitations should be kept in mind when evaluating the conclusion of the present study. Furthermore, researchers are encouraged to investigate factors affecting the livelihood and survival strategies of international returnees using various variables and larger samples at national level.