AFRICAN HUMAN MOBILITY REVIEW Volume 3 Number 1 January– April 2017 Harnessing Economic Impacts of Migrant Remittances for Development in Sub-Saharan Africa: A Critical Review of the Literature Themba Nyasulu The Role of Trust and Migrant Investments in Diaspora- Homeland Development Relations Leander Kandilige Attitude, Risk Perception and Readiness of Ethiopian Potential Migrants and Returnees Towards Unsafe Migration Abebaw Minaye & Waganesh A. Zeleke The Impact of Migration on the Welfare of Households in Ghana: A Propensity Score Matching Approach Abdul-Malik Abdulai, Louis Boakye-Yiadom and Peter Quartey ii AHMR _________________________________________________________________________________ Chief Editor Dr Mulugeta F. Dinbabo University of the Western Cape, South Africa Board Members Dr Beneberu Assefa Wondimagegnhu Bahir Dar University, Ethiopia Dr Delali Margaret Badasu University of Ghana, Ghana Dr Edmond Agyeman University of Education, Winneba, Ghana Dr Ernest Angu Pineteh University of Pretoria, South Africa Dr Joseph Yaro University of Ghana, Ghana Prof Laurence Piper University of the Western Cape, South Africa Dr Linda Oucho African Migration and Development Policy Centre, Kenya Prof Loren Landau University of Witwatersrand, South Africa Dr Lothar Smith Radboud University, Netherlands Dr Meselu Alamnie Mulugeta Bahir Dar University, Ethiopia Prof Raul Delgado Wise Universidad Autónoma de Zacatecas, Mexico Dr Razack Karriem University of the Western Cape, South Africa Dr Sharon Penderis University of the Western Cape, South Africa Prof Simon Bekker University of Stellenbosch, South Africa Prof Shimelis Gulema Stony Brook University, New York Prof Thomas Faist Bielefeld University, Germany ____________________________________________________________________________________ AHMR is an interdisciplinary peer-reviewed on-line journal created to encourage and facilitate the study of all aspects (socio-economic, political, legislative and developmental) of Human Mobility in Africa. Through the publication of original research, policy discussions and evidence research papers, AHMR provides a comprehensive forum devoted exclusively to the analysis of contemporaneous trends, migration patterns and some of the most important migration-related issues. AHMR is published by the Scalabrini Institute for Human Mobility in Africa (SIHMA) a member of the Network of the Scalabrini Centers for Migration Studies, with institutions in New York, Paris, Rome, Buenos Aires, Sao Paulo and Manila. Articles and reviews in AHMR reflect the opinions of the contributors. All rights reserved. No part of this publication may be reproduced without the written permission from the publisher. ISSN 2410-7972 (online) ISSN 2411-6955 (print). Webpage: www.sihma.org.za Copyright © 2017 by the SCALABRINI INSTITUTE FOR HUMAN MOBILITY IN AFRICA 47, Commercial St, 8001 Cape Town – South Africa Tel. 0027 021 461 4741, Email: editor@sihma.org.za http://sihma.org.za/mulugeta-dinbabo http://sihma.org.za/beneberu-wondimagegnhu/ http://sihma.org.za/delali-badasu/ http://sihma.org.za/edmond-agyeman/ http://sihma.org.za/laurence-piper/ http://sihma.org.za/linda-oucho/ http://sihma.org.za/loren-landau/ http://sihma.org.za/lothar-smith/ http://sihma.org.za/meselu-alamnie-mulugeta/ http://sihma.org.za/raul-delgado-wise/ http://sihma.org.za/sharon-penderis/ http://sihma.org.za/shimelis-gulema/ http://sihma.org.za/thomas-faist/ mailto:editor@sihma.org.za iii African Human Mobility Review ___________________________________________________________________________ Volume 3 Number 1 January – April 2017 ARTICLES 645 Harnessing Economic Impacts of Migrant Remittances for Development in Sub-Saharan Africa: A Critical Review of the Literature Themba Nyasulu 671 The Role of Trust and Migrant Investments in Diaspora- Homeland Development Relations Leander Kandilige 702 Attitude, Risk Perception and Readiness of Ethiopian Potential Migrants and Returnees towards Unsafe migration Abebaw Minaye & Waganesh A. Zeleke 732 The Impact of Migration on the Welfare of Households in Ghana: A Propensity Score Matching Approach Abdul-Malik Abdulai, Louis Boakye-Yiadom and Peter Quartey AHMR, Vol.3 No1, January-April 2017 645 Harnessing Economic Impacts of Migrant Remittances for Development in Sub-Saharan Africa: A Critical Review of the Literature Themba Nyasulu* Abstract The recent rise in migrant remittances across Sub-Saharan Africa is one of the important issues currently dominating economic policy discourse in the region. Given the large volume of remittance flows, it is obvious that they have important positive and negative economic effects on the individual families and economies that receive them. Therefore, this paper critically examines channels through which remittance transfers affect microeconomic and macroeconomic activity, and suggests policy options available to Sub-Saharan African countries in terms of harnessing their development potential. The paper affirms that prospects for remittances to facilitate economic development remain high provided that recipient countries put in place institutional frameworks capable of mitigating the malign effects and enhancing the benign effects of remittances. Keywords migrant remittances, microeconomic impacts, economic development, Sub-Saharan Africa. Introduction Remittances, referring to money sent by migrants to their countries of origin, are among the important topics currently dominating economic policy discussions across the globe. In line with the reality that more than 3.4 percent (i.e. more than 247 million people) of the world population lives outside their countries of birth, remittances have also been expanding rapidly in recent times. The World Bank (2016) observes that as of 2015, global remittance flows exceeded US$601 billion. A similar pattern has prevailed in Sub-Saharan Africa where remittances have meteorically risen so much so that they now, together with foreign direct investment and aid, constitute one of the three most important forms of capital inflows in the region (IMF, 2016). *Fellow in International Development, Institute for Development Research and Development Policy (IEE), Ruhr-University Bochum, Germany. Email: Themba.Nyasulu@ruhr-uni-bochum.de AHMR, Vol.3 No1, January-April 2017 646 From both an empirical and a theoretical point of view, migrant remittances are reputed to generate important economic impacts, which in turn profoundly influence development. In particular, these economic impacts are normally felt in terms of consumption, gross domestic product (GDP), foreign exchange reserves, exchange rates, exports and imports, and the country’s creditworthiness, among others. Moreover, remittance flows also have important impacts on human development through their influences on poverty reduction, education and health outcomes (Chami et al., 2008; Chami & Fullenkamp, 2013; Dinbabo & Carciotto, 2015). For Sub-Saharan Africa, which is the least developed region in the world and also home to over two- thirds of world’s poorest people, the above remittance-induced economic impacts, whether positive or negative, have a significant bearing on the development prospects of the region. This also suggests that there are multiple paths through which remittances affect both macroeconomic and microeconomic activity in recipient countries in this part of the world. As such, both the benign and malign economic impacts of migrant remittance inflows on development in Sub-Saharan African countries cannot simply be presumed but deserve critical empirical investigation. Therefore, the purpose of this paper is twofold: first, to analytically assess microeconomic and macroeconomic economic effects of remittances on development, and secondly, to suggest ways through which Sub-Saharan African countries can effectively utilise remittance receipts in order to maximise their development impact. In undertaking these tasks, the discussion critically reviews available theoretical and empirical literature on the subject from both Sub-Saharan Africa and other developing regions of the world. The paper is organised as follows: the economic motivations for remitting are analysed at the beginning of the discussion. After this, the paper examines the impacts of remittances on recipient households (i.e. micro-level) and the economies of Sub-Saharan Africa (i.e. macro-level). Furthermore, the discussion suggests some policy recommendations for leveraging the economic impacts of remittances for development. Finally, concluding remarks on the subject in question are presented in the epilogue. Economic Motivations for Migrant Remittances Among the important drivers of remittance flows are both the ability of migrants to generate and save income and their determination to remit savings back to their countries of origin (i.e. source/home countries). It must also be pointed out that the willingness to send remittances may also be driven AHMR, Vol.3 No1, January-April 2017 647 by the length of time migrants plan to stay abroad either on a permanent or temporary basis. For others, their family status (i.e. single or married) and social networks may play an important role in their decisions to remit. This indicates that there are several ways of analysing determinants of remittance flows, one of which is examining the motivations of migrants to remit. A careful survey of the literature reveals that there are four major economic motives for remitting, namely: altruism, self-interest, informal agreements with family members and portfolio management decisions. But this notwithstanding, there is no general theory that explains remittances (Rapoport & Docquier, 2005; OECD, 2006). Instead, what are available are empirical studies trying to describe the occurrence of remittances, even though their explanations are partial and often specific to certain geographical, socio-cultural and temporal contexts. The table below summarises the four major economic motives for remitting. Motivation for Remitting Rationale 1. Altruism  Migrants may remit money simply because they are concerned about the welfare of their relatives they have left behind in the origin country.  The altruistic hypothesis suggests that migrants derive utility from the welfare of their relatives. In the source country.  The amount of money remitted should vary with changes in the migrant’s level of income.  The volume of remittances should decrease not only as the household income of the migrant’s family declines but also when migrant’s social bonds with his/her family in the source- country weaken. AHMR, Vol.3 No1, January-April 2017 648  The level of remittances should gradually subside as the migrant settles in the destination country and brings over his/her relatives. 2. Self-interest  The self-interest model argues that migrants may send remittances to their parents because of the aspiration to inherit increased bequests from their parents.  Migrants owning assets in their countries of origin are more likely to remit to their relatives left behind in order to take care of the assets. 3. Implicit family agreements: (i) Co-insurance and loans  The decision to remit is taken as a family arrangement whereby remittances are considered as endogenous to the migration process. Households seek to find a Pareto-optimal allocation solution where certain family members are designated as migrants and remittances obtained are used as a redistributive tool for the acquired gains.  Firstly, migrants adopt the role of insurees while their dependents in the home country play the role of the insurers. But once migrants get employed, earning income and AHMR, Vol.3 No1, January-April 2017 649 (ii) The migrant’s saving target holding positive expectations about further income accumulation, they become insurers. Finally, in anticipation of a return home, migrants increase their remittances by investing their accumulated savings at home.  The desire to return home with a certain amount of savings is an important motivation for remitting. This ‘savings target’ motive is a direct result of a bargaining duel between a migrant and his/her family.  The migrant’s ability to remit is classified as the supply side while the family’s claim in the home country is the demand side of the remittance function.  The migrant is interested in not only minimising the leakage of his/her income (i.e. remittances and consumption expenditure) but also maximising the savings target. In contrast, the family’s main goal is assumed to be the accumulation of income and remittances above that of its neighbours. 4. Portfolio management decisions  Macroeconomic variables such as real rates of interest in both the source and destination countries AHMR, Vol.3 No1, January-April 2017 650 may significantly drive the decisions to remit and hence the volume of the total remittances.  Savings which are not required for household consumption can be remitted because of their profitability (i.e. high rates of return) in both the source and destination countries.  Portfolio management choice theory argues that savings may have an exogenous character which is intrinsically linked to the migration regime and are therefore dependent on macroeconomic factors such as exchange rates, rates of returns of real and financial assets, and real interest rates. Source: Adapted from Rapoport & Docquier (2005) and OECD (2006). Economic Impacts of Remittances The discussion classifies economic impacts of remittances on development into macroeconomic and microeconomic categories. Macroeconomic Impacts A careful review of the literature in developing countries clearly shows that the impact of remittances is felt on the macroeconomic level mainly through access to finance and economic growth in the recipient country. The following section outlines the remittance-effects on the macro economy through three important channels: economic growth and financial development, economic stability and international finance. AHMR, Vol.3 No1, January-April 2017 651 Remittances as a Macroeconomic Stabiliser Available empirical and theoretical evidence indicates that remittances behave in a counter-cyclical manner, meaning that their changes are inversely related to macroeconomic fluctuations. The main implication of this is that remittances insulate recipient households in home countries from the vagaries brought about by macroeconomic fluctuations (Chami et al., 2009). For instance, Ratha (2007) observes that money sent by migrants to their home countries rose significantly during the financial crisis of the 1990s in Mexico as well as during the 1997 Asian Financial Crisis, especially in countries such as Thailand and Indonesia. During this time, other private flows such as FDI and equity flows were at an all-time low. Apart from the surges experienced during financial crises, remittances have also shown a tendency to increase counter-cyclically in times of natural disasters and socio-political upheaval in the developing world (Clarke & Wallsten, 2004; Yang & Choi, 2007; Mohapatra et al., 2009; Sithole & Dinbabo, 2016). With this, therefore, Ratha (2003) and Frankel (2011) argue that remittances have different characteristics relative to most other private flows such as FDI, foreign aid or portfolio equity. This is so because remittances primarily consist of transactions made by households in destination countries. As such, they are less influenced by profit-maximising motives that are commonly associated with private resource flows. On the same note, Sayan (2006) observes that remittances show pro-cyclical patterns when they are sent in order to accomplish investment purposes in home countries. These patterns are more pronounced in middle-income developing countries than in low-income ones (Lueth & Ruiz-Arranz, 2008). Evidence from several Sub-Saharan African countries indicates that remittances show more stability in times of economic crises relative to other private capital flows such as private debt and equity flows and FDI (Dinbabo & Carciotto, 2015; Gupta et al., 2009). Remittances as a Source of External Finance and Guarantor for Creditworthiness It is now widely recognised in international finance that remittance inflows have a significant potential to boost global sovereign creditworthiness primarily by increasing the volume and stability of foreign exchange received by countries. Large remittance receipts have enabled many African countries to reduce their trade imbalances and hence lower their current account deficits (Avendaño et al., 2009). According to Chami et al. (2008), remittances have also enabled these countries to stabilise their current account AHMR, Vol.3 No1, January-April 2017 652 fluctuations by steadying the overall flow of capital. Similarly, Bugamelli and Patemo (2009) have argued that, as a rule of thumb, remittances significantly reduce current account deficits when they exceed 3 percent of the total national income. These examples give a clear indication of the importance of remittance flows to source countries as far as external financing is concerned. It is an open secret that effective accounting of remittances can go a long way in improving the evaluation of credit worthiness and debt sustainability of many African source countries. Ratha et al. (2011) argue that the foreign debt- to-exports ratio for many African source countries would be significantly lower if remittances were included in the analysis as a denominator. The importance of remittances in debt sustainability analysis is now recognised by multilateral agencies such as IMF and the World Bank. In recent times, remittances have formed important variables necessary for sovereign assessments and debt sustainability analyses in middle-income and low- income countries, respectively (IMF, 2016). However, despite the importance of the above, very few remittance-receiving African economies have managed to have their creditworthiness appraised by major sovereign rating agencies. Remittances as Collateral for Development Financing Remittances provide collateral for mobilising financing for economic development in international capital markets. The success in this endeavor of countries such as Brazil, El Salvador, Turkey and Kazakhstan suggests that Sub-Saharan African countries that often face severe capital constraints can also use future remittances as an important tool for mobilising development finance (Ratha, 2005; Ketkar & Ratha, 2009). Indeed, securitising future remittance receipts can help countries in the Sub-Saharan African region to mobilise not only cheaper but also long-term financial resources for development. Ratha et al. (2011) argue that since remittances in most cases are denominated in hard currencies, international banks can use them as collateral to obtain loan injections, provided their ability to settle domestic transactions in local currencies is not impeded. So, by securitising future remittance flows, migrant source countries are able to not only obtain higher sovereign credit ratings, but also attract a large number of potential foreign investors who can spur economic development. In a similar vein, a number of banks in the aforementioned countries have utilised remittance securitisation to obtain lower cost international financing and maturities. Undeniably, with the support of the African Export-Import Bank (Afreximbank), many countries throughout the region have used the AHMR, Vol.3 No1, January-April 2017 653 securitisation of future remittances to acquire significant medium-term development loans (Afreximbank, 2005; Rutten & Oramah, 2006; Ratha et al., 2011). Ghana, Nigeria and Ethiopia are some of the economies in the region that have used the remittance-backed syndicated facilities to obtain external financing for their development projects in sectors such as agriculture and infrastructure investment (Afreximbank, 2005). Ratha et al. (2011) argue that African countries can leverage a small fraction of their future remittance receipts to issue bonds. But, of course, the quantity of bonds issued would very much depend on the level of collateral possible for banks to secure. Available estimates project that the potential for securitisation of future remittance flows in Africa stood in excess of US$35 billion as of 2009 (Ratha et al., 2009). However, it is worth noting that at present, securitisation of remittance flows faces significant challenges in many low-income African countries. Among the major bottlenecks identified in the literature include: underdeveloped domestic financial systems, limited integration of domestic banks into the global financial system, and exorbitant fixed costs incurred when acquiring legal, credit rating and investment banking services (Ketkar & Ratha, 2009). It must also be pointed out in passing that remittance-backed bonds may, however, pose a risk to issuers, especially if there is a currency mismatch that often accompanies foreign currency debt. Ratha et al. (2011) contend that developing countries need to put in place prudent risk management mechanisms before obtaining additional debt. Additionally, the existence of political instability may expose African economies to volatility in remittance flows as well as disruption of social networks between migrants and their countries of origin. What is more, increased foreign currency inflows resulting from the issuing of remittance-backed bonds may trigger currency appreciation in African countries with fragile macroeconomic structures. Remittances as a Driver of Economic Growth, Financial Development and Competitiveness Remittance flows are also known to positively affect economic growth, competiveness and financial development in many migrant source countries. The positive effect of remittances on growth and development is mainly transmitted through the following channels: consumption and investment increases, productivity enhancing social spending, and both macroeconomic and microeconomic stability of consumption and production (Chami et al., 2009; Dinbabo & Nyasulu, 2015; Mohapatra et al., 2009; Sithole & Dinbabo, AHMR, Vol.3 No1, January-April 2017 654 2016). The above benefits in turn boost investment supply by spurring the demand for financial intermediation. This is what ultimately conveys the positive impacts of remittance flows on economic growth in developing countries. Literature on the subject is awash with empirical examples from Sub-Saharan African countries confirming the above scenario (Rajan & Zingales, 1998; Gupta et al., 2009; Akinlo & Egbetunde, 2010). Apart from the positive effects, remittances can also negatively affect growth. Indeed, several theoretical reasons have been advanced as to why remittances may reduce the level of economic growth. To begin with, researchers argue that a large inflow of remittances may trigger an appreciation in the value of the domestic currency, which in the end can curtail growth through reduction of tradable production and other scale economies and externalities. This negative exchange rate effect on growth is what is commonly known as the ‘Dutch disease’ (World Bank, 2006; Acosta et al., 2010 Gupta et al., 2009). However, Rajan and Subramaniam (2005) find no empirical evidence suggesting that remittance-induced exchange rate surges have negatively affected growth in developing countries across the globe. A second channel through which remittances may depress growth is the reduction of labour supply. In theory, since migration takes away some of the most productive individuals, it therefore deprives the home country of an important growth driver, labour supply (Lucas, 1987; Azam & Gubert, 2006; Bussolo & Medvedev, 2007; Chami et al., 2008). However, there is very little empirical evidence supporting the above argument. In fact, in many developing countries that are characterised by high unemployment and underemployment levels, the loss of one laborer can easily (and cheaply) be replaced by another. As such, the remittance-induced labour supply- reductions will likely have a very limited negative effect on economic growth. On another note, some economic commentators suggest that additional income emanating from remittance receipts eases pressure on migrant destination countries to improve the quality of social service delivery, since with the help of remittances the recipient households can afford to access alternative private services (Abdih et al., 2008; Sithole & Dinbabo, 2016). The above researchers also find evidence suggesting that remittances greatly boost home countries’ foreign exchange position and hence ease balance of payment bottlenecks and fiscal deficits that characterise many developing countries. In contrast, Catrinescu et al. (2006 observe that the impact of remittances on growth is only positive in countries that already possess high- AHMR, Vol.3 No1, January-April 2017 655 quality political and economic institutions. For these economists, many developing countries in Sub-Saharan Africa are unlikely to harvest the growth dividends from remittance inflows because they lack the necessary policy and institutional frameworks. Remittances have also been touted as an important mechanism for alleviating liquidity and credit constraints in many developing economies. Therefore, they play an active role in generating capital for microenterprises and, thus, act as an important alternative for financial development. Giuliano and Ruiz- Arranz (2009) examine the contribution of remittances to economic growth and financial development in developing countries. These economists find robust econometric evidence suggesting that remittances have the strongest positive impact on economic growth only when the level of financial development in the country is low. The above findings firmly resonate with the experience of many Sub-Saharan African economies that contain undeveloped financial systems and face severe liquidity and credit constraints. Microeconomic Impacts of Remittances Apart from having developmental impacts on the macro-level, remittances can also influence microeconomic development fundamentals such as poverty reduction, household consumption and investment decisions, as well as access to social services such as health and education. It is the aim of the following section to review available empirical literature on the microeconomic development impacts of migrant remittances on recipient regions of the developing world, including Sub-Saharan Africa. Remittances as a Driver of Household Capital Investment and Entrepreneurship To begin with, the fungibility of money and the unreliability of migration data in developing countries make it difficult to calculate the proportion of remittances allocated to physical capital investment and entrepreneurship in source countries. Nevertheless, ample empirical evidence still exists suggesting that recipient households devote a significant chunk of remittance receipts to investment in housing and the acquisition of land, especially in regions where there are no credible alternative stores of value for money. On the same note, empirical studies by Taylor and Wyatt (1996) reveal that in many developing home nations, the shadow value of remittances necessary for offsetting risk and liquidity bottlenecks is vital for low and middle-income recipient households that usually face challenges to obtain credit from formal AHMR, Vol.3 No1, January-April 2017 656 financial institutions. Concurring with this assertion, Adams and Cuecuecha (2010) find that remittance-receiving families in Guatemala allocate very minimal amounts of the remittances to housing expenditure. However, the positive micro-effects of remittances are not limited to urban areas of developing regions. For instance, Adams (1998) finds that migrant remittances significantly boost investment in agricultural land in rural areas of Pakistan. Despite the abundance of empirical evidence in Latin America and Asia, studies assessing the utilisation of remittances for investment and entrepreneurial purposes in Africa appear to be scarce. Among the few available studies are those done by McCormick and Wahba (2001), which find that remittances from return migrants greatly enhance the prospects of entrepreneurship and investment in Egypt. In addition, Osili (2004) found that not only do remittances from United States-based migrants boost investment expenditure on housing in Nigeria, but also that the investment is very responsive to macroeconomic fluctuations in exchange rate, inflation and political climate. Ratha et al. (2011) observe that the majority of studies on Africa conducted under the auspices of the Africa Migration Project (AMP) reveal that a significant proportion of the remittances are spent on housing infrastructure, land acquisition, establishment of small-scale businesses and farming activities. There is strong empirical evidence on the above tendencies gathered from household surveys conducted in Burkina Faso, Kenya, Uganda, Nigeria and Senegal. Remittances as an Enabler for Recipient Households to Attain High Quality Education Migrant remittances boost expenditure on education by contributing to funding of school activities in many migrants’ home countries. The increased education financing reduces the need for child labour in these countries. There is ample empirical evidence in African countries suggesting that migrant remittances can contribute to increasing school enrolment and attendance levels. But despite these positive education spinoffs, migration may sometimes negatively affect children’s performance in schools since the absence of parents/guardians at home as a result of migration increases the burden for kids to perform household chores (Ratha et al., 2011). AHMR, Vol.3 No1, January-April 2017 657 Even though there is scarcity of empirical household surveys analysing the remittance-education nexus in Sub-Saharan Africa, in some countries such as Egypt, remittances are known to increase university enrollment rates and reduce domestic chores for school-going children (Elbadawi & Roushy, 2009). On the same note, Adams & Page (2005) found that in Ghana, households that received remittances from abroad invested more in education than families with no remittance receipts. This suggests that remittances may be a potent tool for financing education in Africa. Remittances as a Financier of Household Health Expenditure Money sent by migrants can also enable home countries to achieve their health outcomes by, among others, empowering families in rural areas to acquire more food and healthcare services and improving their access to information on modern health practices. In support of the above hypothesis, Drabo and Ebeke (2010) found in a survey of 56 developing countries that remittances not only increased access to private treatment for communicable diseases such as malaria and diarrhea, but also augmented foreign aid in financing health outcomes. In a similar vein, an empirical study covering 84 countries by Chauvet et al. (2009) found that an increase in the level of remittances significantly reduced infant mortality, even though the reduction was higher for richer recipient households over those that were poorer. Going further into the above analysis, Guzmán et al. (2007) found that female- headed households in Ghana spend more of their remittances on healthcare than those headed by men, irrespective of whether the remittances came from outside or within Africa. Earlier studies done in Mali by Birdsall and Chuhan (1986) also found that remittances triggered not only a rise in healthcare demand but also modern health facilities. In South Africa, Nagarajan (2009) found that between 1993 and 2004, increased migrant remittances enabled poor households in KwaZulu-Natal Province to spend more on healthcare and food items and improved access to better medical facilities. Remittances as a Form Insurance against Adverse Socio-Economic and Climatic Shocks There is now growing recognition among economists that migration can contribute positively to diversifying household income sources and thereby minimise their risk exposure to adversities such as famine, drought and other vagaries of nature. For example, Calero et al. (2009) observe that remittances were vital in keeping Ecuadorian children in school at a time when their AHMR, Vol.3 No1, January-April 2017 658 households faced adverse shocks. Moreover, Weiss-Fagan (2006) and Ratha (2010) empirically demonstrate that remittances from relatives and well- wishers in the United States enabled recipient-households to ably cope with several cyclone and earthquake crises that affected Haiti in 2004 and 2010, respectively. Remittances are also reputed to have accelerated the recovery of Indonesian households affected by the 2004 tsunami disaster. On the same note, Wu (2006) observes that remittance-receiving families showed evidence of a quicker recovery than those with no access to money from migrants. A similar episode was also reported in Pakistan where remittances enabled the recovery and reconstruction of household livelihoods after a ravaging earthquake (Suleri & Savage, 2006). It is a well-known fact that in many disaster-prone regions of Africa, migration and remittances play a vital role in enabling recipient households to cope with the resultant shocks and thereby smooth their consumption patterns (Block & Webb, 2001; Dinbabo & Carciotto, 2015). Lucas and Stark (1985) buttress the above point by empirically showing that families facing drought in Botswana were able to avert income shocks by utilising their increased remittance receipts to purchase livestock and food for sustenance. In a similar study, Mohapatra et al. (2009) unearth a somewhat startling result that during droughts, remittance-receiving families in Ethiopia are less likely to pawn their product assets such cattle for food as a way of consumption-smoothing. However, a study by Quartey and Blankson (2004) reveals that migrant remittances are the major source of consumption-smoothing for Ghanaian small-holder farmers in rural areas. Furthermore, there is empirical evidence emanating from rural Mali suggesting that remittances are an effective tool for dealing with various types of household income shocks (Gubert, 2002). Apart from cushioning households from income shocks, remittances were also found to be an important risk-diversification tool for Malian and Senegalese farming households in times of adverse climatic conditions (Azam & Gubert, 2005 and 2006). Besides, money from migrants abroad can enable their relatives in home countries to construct durable and more resilient housing. In concurrence with this view, Mohapatra et al. (2009) found that in Burkina Faso and Ghana, remittances enabled recipient households to construct more concrete and iron-roofed houses than non-remittance receiving families. AHMR, Vol.3 No1, January-April 2017 659 Remittances as a Potent Tool for Poverty and Inequality Reduction In the literature, it is a well-established fact that remittances have the ability to reduce the level of household poverty. This is so because, among others, remittances supplement the incomes of poor recipient families and boost aggregate demand levels and, in turn, create employment opportunities and wage earnings of the poor (Ratha et al., 2011). Seminal studies by Adams and Page (2003 and 2005) strongly support the above theoretical assertions by illustrating that increased inflows of remittances in migrant source countries greatly diminish the share of poor people in the country. There is a plethora of econometric studies suggesting that remittances play an important role in poverty reduction in Africa. Ajayi et al. (2009) and Anyanwu and Erhijakpor (2010) find evidence that suggests that official international remittances as a percentage of the GDP have a significant positive effect on poverty reduction in 33 African countries. Likewise, Gupta et al. (2009) find that the remittance-induced poverty reduction in Africa is much higher when compared to other developing regions of the world. Similarly, numerous case study analyses in Burkina Faso, Ghana, Lesotho, Morocco and Nigeria, among others, firmly endorse the poverty-reducing effect of remittances in rural regions (Lachaud, 1999; Quartey & Blankson, 2004; Adams & Page, 2005; Sorensen, 2004; Odozia et al., 2010). These studies further suggest that remittances not only reduce the overall number of people living in poverty but also reduce the severity and depth of the deprivation. Despite the availability of numerous studies analysing the relationship between remittances and poverty, the same cannot be said about the remittance-inequality nexus. The World Bank (2006) and Ratha and Mohapatra (2007) argue that the ambiguity of the impact of remittances on income inequality is largely a result of the unavailability of counterfactuals, i.e. how inequality levels in the source countries would be in the absence of remittance inflows. Nevertheless, the influence of income inequality on the decision to emigrate and actual flows of remittances is quite clear. As a matter of fact, households receiving remittances from outside Africa may need a certain income threshold for them to ably sponsor a family member to emigrate in the first place. Remittance-receiving families may also have higher incomes than those households with no access to remittances. AHMR, Vol.3 No1, January-April 2017 660 Some Policy Options for Leveraging Remittance Impacts for Economic Development in Sub-Saharan Africa The positive role that remittances play in poverty reduction, growth enhancement and social transformation has been explored at length in this paper. Both theoretical and empirical studies broadly agree that remittances contribute to socio-economic development in source countries by meeting the basic needs of recipients such as housing, education and health. This is in addition to the facilitation role that remittances play in transferring skills and knowledge from returning migrants and the diaspora. The above merits suggest a strong case for Sub-Sahara African countries to integrate remittance inflows into their overarching migration and national development strategies. The discussion has indicated a reasonable degree of consensus that exists among migration economists on the potential that remittances have to positively contribute to economic development in recipient countries. However, for these positive development spinoffs to be maximised there is need for proactive and targeted policies to be implemented. For instance, remittance recipient countries should improve the levels of financial literacy and financial inclusion for both their emigrants and the beneficiaries of remittances. This would ensure that the remittance funds are channelled into productive sectors. Apart from this, both migrant source and destination countries can cooperate in providing technical training on money transfers and financial services to both the remitters and the recipients. Training programs can also take the form of migrant entrepreneurship and SME (small and medium scale enterprise) development sessions. Moreover, the securitisation of remittances through initiatives such as diaspora bonds can also greatly improve the remittance-impact on socio-economic development. Even though Sub-Saharan African countries must keep in mind that remittances are private flows, their economic effects can still be harnessed for development. One of the important ways of maximising the development impact of remittances is by formalising their flows. Furthermore, remittances can have a maximum impact on development if Sub-Saharan African governments make deliberate efforts to put in place enabling and competitive environments that allow these funds to flow into productive sectors. Another important solution is for the recipient countries to upgrade and increase coverage of their underdeveloped financial systems. Improved access to financial services facilitates the linkages between remittances and other AHMR, Vol.3 No1, January-April 2017 661 financial products, such as savings, insurance, mortgages and credit, which in the end foster economic development (Mashayekhi, 2013). Recipient countries in Sub-Saharan Africa should also consider reducing transaction costs for growth-enhancing remittance flows to be effective. Allowing non-bank financial institutions such as microfinance organisations to undertake distribution and proper supervision can go a long way in reducing remittance transaction costs (Ratha & Riedberg, 2005). On the same note, the removal of taxation remittances can not only lower transaction costs, but can also encourage formal transfers and increase the available fiscal revenue (Chami et al., 2008; World Bank, 2011). It would be advisable for remittance- receiving Sub-Saharan African countries to implement consumption-based taxation in order to reduce the deleterious effects on economic growth. Such a taxation system would also help these countries to greatly reduce distortions induced by macroeconomic stabilisation policies, and also reap the remittance benefits emanating from tax-driven investments. More importantly, perhaps, it would be prudent for remittance-receiving Sub- Saharan African countries to channel their remittance inflows into activities that promote economic development in the long-term, and preserve poverty alleviation efforts in the short-term. This so because there is lack of empirical clarity on whether the two main development objectives of remittances: growth enhancement and poverty reduction are compatible (Ratha, 2013). This being the case, the above governments should strive to strike a delicate balance in allocating remittance receipts between competing growth- promoting activities and poverty alleviating measures. The increased fiscal space generated by remittance inflows certainly allows for this undertaking. International financial institutions can also play a crucial role in assisting recipient Sub-Saharan African countries to leverage the economic impacts of remittances for development. Through engagement and dialogue, these financial organisations can encourage governments to not only implement but also to accelerate the necessary reforms pertaining to remittance inflows. Instead of imposing one-size-fits-all reform prescriptions, international financial institutions should encourage country-specific strategies that are driven by the remittance inflow characteristics of each Sub-Saharan African nation. This may assist these countries to boost the development potential of remittance receipts. All in all, international cooperation between Sub-Saharan African countries and the international community could also be a key to enhancing the AHMR, Vol.3 No1, January-April 2017 662 economic development impact of remittances. This international cooperation agenda should primarily focus on making remittances cheaper, safer and more productive for both migrant source countries and destination countries. Concluding Remarks In conclusion this paper observes that even though remittances have a significant potential to contribute to economic development, many Sub- Saharan African recipient economies have not been able to maximise the developmental benefits from their economic impacts. Therefore, this calls for a better understanding of the realities concerning the remittance-effects on economic development. Without doubt, policy makers intending to channel remittance receipts into economic development need to appreciate that these inflows can have both positive and harmful consequences on the economy. Coupled with this is the fact that remittances do not automatically gravitate towards economic development-enhancing activities, but may sometimes flow into development-inhibiting activities, mainly because of the moral hazards that are associated with them. Against this background, Sub-Saharan African countries that desire to reap the positive economic benefits of remittances need to establish enabling institutional and infrastructural frameworks capable of channeling these inflows into development-enhancing activities through the financial system and private sector. Furthermore, recipient governments need to muster political will in order to efficiently utilise the fiscal revenue generated by remittances in order to judiciously invest only in those public goods and infrastructure that yield the highest returns. To this end, this paper strongly advocates for the development of well-functioning domestic institutions as an important driver for harnessing the positive economic benefits of remittances for development. It is beyond doubt that remittances have a significant potential to facilitate economic transformation, but to translate this possibility into reality, Sub-Saharan African countries need to develop and strengthen their economic and social institutions. References Abdih, Y., R., Chami, J., Dagher, and Montiel, P.2008. Remittances and Institutions: Are Remittances Curse? IMF Working Paper 08/29. International Monetary Fund. Acosta, P., E., Lartey, and Mandelman, F. 2010.Remittances, Exchange Rate Regimes, and the Dutch Disease: A Panel Data Analysis. Review of International Economics, 20(2): 377-395. AHMR, Vol.3 No1, January-April 2017 663 Adams, R.1998.Remittances, Investment and Rural Asset Accumulation in Pakistan. Economic Development and Cultural Change, 47:155-173. Adams, R., and Cuecuecha, A.2010.Remittances, Household Expenditure and Investment in Guatemala. World Development 38 (11): 1626–1641. Adams, R., and Page, J. 2003. International Migration, Remittances and Poverty in Developing Countries. Policy Research Working Paper 3179.World Bank, Washington, DC. Adams, R., and Page, J. 2005. Do International Migration and Remittances Reduce Poverty in Developing Countries? World Development, 33 (10): 1645– 1669. Afreximbank (African Export-Import Bank). 2005 Annual Report. http://www.afreximbank.com (retrieved on 11 December 2016). Ajayi, M., M., Ijaiya, G., Ijaiya, R., Bello, M., Ijaiya, and Adeyemi, S. 2009.International Remittances and Well-Being in Sub-Saharan Africa. Journal of Economics and International Finance, 1(3):78–84. Akinlo, A., and Egbetunde, T. 2010. Financial Development and Economic Growth: The Case of 10 Sub-Saharan Africa Revisited. The Review of Banking and Finance, 2(1): 17-28 Anyanwu, J., and Erhijakpor, A.2010. Do International Remittances Affect Poverty in Africa? African Development Review, 22 (1):51–91. Avendaño, R., N., Gaillard, and Nieto Parra, S. 2009. Are Workers’ Remittances Relevant for Credit Rating Agencies. Development Center Working Paper 282. Organisation for Economic Cooperation and Development, Paris. Azam, J-P. and Gubert, F. 2005. Those in Kayes: The Impact of Remittances on Their Recipients in Africa. Revue Économique, 56 (6): 1331–1358. Azam, J-P., and Gubert, F. 2006. Migrants’ Remittances and the Household in Africa: A Review of Evidence. Journal of African Economies, 15 (AERC Supplement): 426–462. Birdsall, N., and Chuhan, P. 1986. Client Choice of Health Care Treatment in Rural Mali. Unpublished manuscript for the Health, Nutrition and Population Department. World Bank, Washington, DC. Block, S., and P. Webb, P. 2001. The Dynamics of Livelihood Diversification in Post Famine Ethiopia. Food Policy, 26 (4): 333–350. AHMR, Vol.3 No1, January-April 2017 664 Bugamelli, M., and Paterno, F. 2009. Do Workers’ Remittances Reduce the Probability of Current Account Reversals. World Development, 37 (12): 1821– 1838. Bussolo, M., and Medvedev, D. 2007. Do remittances have a flip side? A general equilibrium analysis of remittances, labor supply responses, and policy options for Jamaica. Working Bank Policy Working Paper 4143. Washington D.C. Calero, Carla, Arjun S. Bedi, and Robert Sparrow. 2009. Remittances, Liquidity Constraints and Human Capital Investments in Ecuador. World Development, 37 (6):1143–1154. Catrinescu, N., M., León-Ledesma, M., Piracha, and Quillin B.2006.Remittances, Institutions, and Economic Growth. IZA Discussion Paper No. 2139. Chami, R, A., Barajas, T., Cosimano, C., Fullenkamp, M., Gapen, and Montiel, P.2008. Macroeconomic Consequences of Remittances. IMF Occasional Paper No. 259. International Monetary Fund. Chami, R., Fullenkamp, C.2013.Workers’ Remittances and Economic Development: Realities and Possibilities. In Maximizing the Remittances Impacts on Development, United Nations Commission for Trade and Development (UNCTAD), Geneva. Chami, R., D., Hakura and Montiel, P. 2009. Remittances: An Automatic Stabilizer? IMF Working Paper 09/91. International Monetary Fund. Chauvet, L., F. Gubert, and Mesplé-Somps, S. 2009. Are Remittances More Effective than Aid to Reduce Child Mortality? An Empirical Assessment Using Inter- and Intra-Country Data. DIAL Working Paper DT/2009-11. Développement, Institutions et Analyses de Long terme (DIAL), Paris. Clarke, G., and Wallsten, S.2004. Do Remittances Protect Household in Developing Countries against Shocks? Evidence from a Natural Disaster in Jamaica. World Bank, Development Research Group, Washington, DC Dinbabo, M. and Carciotto, S. 2015. International migration in Sub-Saharan Africa (SSA): A call for a global research agenda. African Human Mobility Review (AHMR), 1(2): 154-177. Dinbabo, M. and Nyasulu, T. 2015. Macroeconomic determinants: Analysis of ‘Pull’ factors of international migration in South Africa. African Human Mobility Review (AHMR), 1(1): 27-52. AHMR, Vol.3 No1, January-April 2017 665 Drabo, A. and Ebeke, C.2010. Remittances, Public Health Spending and Foreign Aid in the Access to Health Care Services in Developing Countries. Working Paper E/2010/04. Centre d’Etudes et de Recherche sur le Développement International (CERDI), Clermont-Ferrand, France. Elbadawi, A., and Roushdy, R.2009.Impact of International Migration and Remittances on Child Schooling and Child Work: The Case of Egypt. Paper prepared for the World Bank’s MENA International Migration Program. Washington, DC. Frankel, J.2011. Are Bilateral Remittances Countercyclical? Open Economies Review, 22 (1): 1–16. Giuliano, P. and Ruiz-Arranz, M. 2009. Remittances, Financial Development and Growth. Journal of Development Economics, 90 (1): 144–152. Gubert, F.2002.Do Migrants Insure Those Who Stay Behind? Evidence from the Kayes Area (Western Mali). Oxford Development Studies, 30 (3): 267–287. Gupta, S., C., Pattillo, and Wagh, S.2009. Impact of Remittances on Poverty and Financial Development in Sub-Saharan Africa. World Development, 37 (1):104–115. Guzmán, J-C., A., Morrison, and Sjöblom, M. 2007. The Impact of Remittances and Gender on Household Expenditure Patterns. In The International Migration of Women, ed. Maurice Schiff, Andrew R. Morrison, and Mirja Sjöblom, 125–152. Washington, DC: World Bank. IMF-International Monetary Fund. 2016. Sub-Saharan Africa Multispeed Growth. World Economic and Financial Surveys: Regional Economic Outlook. Washington, D.C. Ketkar, S. and Ratha, D. (ed.). 2009. Innovative Financing for Development. Washington DC. World Bank. Lachaud, J-P. 1999. Envoi de fonds, inégalité et pauvreté au Burkina Faso. Document de travail 40, Groupe d’Economie du Développement de l’Université Montesquieu Bordeaux IV. Lucas, R. and Stark, O.1985. Motivations to Remit: Evidence from Botswana. Journal of Political Economy, 93 (5): 901–918. Lucas, R.1987. Emigration to South Africa’s Mines. The American Economic Review, 77(3): 313-330. AHMR, Vol.3 No1, January-April 2017 666 Lueth, E., and Ruiz-Arranz, M. 2008.Determinants of Bilateral Remittance Flows. B.E Journal of Macroeconomics, 8 (1), Article 26. Mashayekhi, M.2013.Leveraging Remittances for Development. In Maximizing the Remittances Impacts on Development. United Nations Commission for Trade and Development (UNCTAD). Geneva. McCormick, B. and Wahba, J. 2001.Overseas Work Experience, Savings and Entrepreneurship amongst Return Migrants to LDCs. Scottish Journal of Political Economy, 48 (2):164–178. McKenzie, D., and Rapoport, H.2010. Can Migration Reduce Educational Attainment? Evidence from Mexico. Journal of Population Economics. http://www .springerlink.com/content/95ht846m8m3065l7 (retrieved on 29 November 2016). Mohapatra, S., G., Joseph, and Ratha, D.2009.Remittances and Natural Disasters: Ex-Post Response and Contribution to Ex-Ante Preparedness. Policy Research Working Paper 4972.World Bank. Washington, DC. Nagarajan, S. 2009. Migration, Remittances, and Household Health: Evidence from South Africa. Ph.D. dissertation. Department of Economics. George Washington University. Washington, DC. Odozia, J., T., Awoyemia, and Omonona, B. 2010. Household Poverty and Inequality: The Implication of Migrants’ Remittances in Nigeria. Journal of Economic Policy Reform, 13 (2):191–199. OECD-Organization of Economic Cooperation and Development. 2006. International Migrant Remittances and their Role in Development. International Migration Outlook 2006. Paris Osili, U.2004. Migrants and Housing Investments: Theory and Evidence from Nigeria. Economic Development and Cultural Change, 52 (4): 821–849. Quartey, P. and Blankson, T.2004.Do Migrant Remittances Minimize the Impact of Macro-Volatility on the Poor in Ghana? Report for the Global Development Network, University of Ghana, Legon. Rajan, R. and Subramanian, A. 2005. What Undermines Aid’s Impact on Growth? IMF Working. Paper 05/126.International Monetary Fund and NBER Working Paper 11657 (October). AHMR, Vol.3 No1, January-April 2017 667 Rajan, R. and Zingales, L. 1998.Financial Dependence and Growth. American Economic Review, 88 (3):559–586. Rapoport, H., and Docquier, F.2005.The Economics of Migrants’ Remittances. IZA Discussion Paper No. 1531. Bonn. Ratha, D. 2003.Workers’ Remittances: An Important and Stable Source of External Development Finance in Global Development Finance: 157-175. World Bank. Ratha, D. 2005. Leveraging Remittances for Capital Market Access. Mimeo. World Bank. Washington, DC. Ratha, D. 2007. Leveraging Remittances for Development. Policy Brief (June). Migration Policy Institute. Washington, DC. Ratha, D.2010. Mobilize the Diaspora for the Reconstruction of Haiti. SSRC Feature: Haiti, Now and Next, Social Science Research Council, New York. http://www.ssrc.org/features/pages/haiti-now-and-next/1338/1438. (retrieved on 2 January 2017). Ratha, D. 2013. The Impact of Remittances on Economic Growth and Poverty Reduction. Migration Policy Institute Policy Brief No. 3, September. Ratha, D., and Mohapatra, S. 2007. Increasing the Macroeconomic Impact of Remittances. Paper prepared for the G8 Outreach Event on Remittances. Berlin. November 28–30. http://dilipratha.com/index_files/ G8Berlin.pdf. (retrieved on 24 December 2016). Ratha, D., P., De, and Mohapatra, S. 2011. Shadow Sovereign Ratings for Unrated Developing Countries. World Development, 39(3):295-307. Ratha, D., S. Mohapatra, and Plaza S. 2009. Beyond Aid: New Sources and Innovative Mechanisms for Financing Development in Sub-Saharan Africa” In Innovative Financing for Development, ed. Suhas Ketkar and Dilip Ratha, 143– 83.Washington DC: World Bank. Ratha, D. and Riedberg, J. 2005. On Reducing Remittance Costs. Mimeo. Development Research Group, World Bank, Washington, DC. Rutten, L., and Oramah, O. 2006. Using Commodity Revenue Flows to Leverage Access to International Finance: With Special Focus on Migrant Remittances and Payment Flows. Study prepared by UNCTAD Secretariat AHMR, Vol.3 No1, January-April 2017 668 Sayan, S. 2006. Business Cycles and Workers’ Remittances: How Do Migrant Workers Respond to Cyclical Movements of GDP at Home? Working Paper 06/52, International Monetary Fund, Washington, DC. Sithole, S. and Dinbabo, M. 2016. Exploring youth migration and the food security nexus: Zimbabwean youths in Cape Town, South Africa. African Human Mobility Review (AHMR), 2(2): 512-537. Sorensen, N. 2004. Migrant Remittances as a Development Tool: The Case of Morocco”, IOM Working Paper 2, International Organization for Migration, Geneva. http://bit.ly/2nTm6pE (retrieved on 2 December 2016). Suleri, A. and Savage, K. 2006. Remittances in Crisis: A Case Study of Pakistan. Overseas Development Institute, London. http://bit.ly/2nFkBKC (retrieved on 15 December 2016). Weiss-Fagan, P. 2006. Remittances in Crises: A Haiti Case Study. Humanitarian Policy Group background paper. Overseas Development Institute (ODI), London. http://bit.ly/2oy6JFS (retrieved on 15 December 2016). Taylor, J. and Wyatt, T. 1996. The Shadow Value of Migrant Remittances, Income and Inequality in a Household-Farm Economy. Journal of Development Studies, 32 (6): 899–912. World Bank. 2006. Global Economic Prospects: Economic Implications of Remittances and Migration. Washington, DC: World Bank. World Bank. 2011. Leveraging Migration for Africa: Remittances, Skills and Investments. Washington DC. World Bank. 2016. Migration and Remittances Fact book 2016, 3rd. Washington DC. Wu, T. 2006.The Role of Remittances in Crisis: An Aceh Research Study. Humanitarian Policy Group background paper. Overseas Development Institute (ODI), London. http://bit.ly/2oEyfhX (retrieved on 27 December 2016). Yang, D., and Choi, H-J. 2007. Are Remittances Insurance? Evidence from Rainfall Shocks in the Philippines. The World Bank Economic Review 21 (2): 219–248. AHMR, Vol.3 No1, January-April 2017 669 Annexes Annex 1: Top 10 remittance receiving countries in Sub-Saharan Africa (in gross terms) as of 2015 Recipient countries Gross remittance value (US$ billions) Nigeria Ghana Senegal Kenya South Africa Uganda Mali Ethiopia Liberia Sudan 20.8 2 1.6 1.6 1 0.9 0.9 0.6 0.5 0.5 Source: World Bank (2016) Annex 2: Top 10 remittance sending countries in Sub-Saharan Africa in 2014 Remittance sending countries Gross value of remittances sent (US$ billion) Angola South Africa Liberia Uganda Mozambique Mauritania Kenya Rwanda Tanzania Zambia 1.3 1.1 0.4 0.3 0.2 0.2 0.2 0.1 0.1 0.1 Source: World Bank (2016) AHMR, Vol.3 No1, January-April 2017 670 Annex 3: Recent trends in international financial flows and economic growth in Sub-Saharan Africa 2004-08 2009 2010 2011 2012 2013 2014 2015 2016p Real GDP growth (percentage change) Remittances (US$ billions) FDI inflows (US$ billions) Foreign aid inflows (US$ billions) 6.6 3.9 7.0 5.0 4.3 5.2 5.1 3.4 1.4 36.7 44.9 52.5 57.0 61.9 61.2 63.8 64.6 66.4 42.8 55.1 46.0 49.8 49.7 54.2 49.4 57.5 66.3 39.0 48.0 47.7 51.5 51.1 56.7 54.2 56.4 58.7 Source: International Monetary Fund-IMF (2016), p is projection. AHMR, Vol.3 No1, January-April 2017 671 The Role of Trust and Migrant Investments in Diaspora- Homeland Development Relations Leander Kandilige* Abstract Research into the role of diaspora communities in origin countries’ development is a growing phenomenon. However, there is little understanding of the role of trust in mediating transnational relationships between migrants and recipients of remittances (non-migrants, members of migrant households and community leaders). Using a case study methodology, mixed methods and a comparative approach – in-depth interviews with 40 key informants (20 in the UK and 20 in Ghana), 120 questionnaires administered in the UK and 346 questionnaires administered in Ghana – this paper examines differing conceptualisations of trust among ‘development partners’ in the process of negotiating as well as implementing migrant-funded development projects. It also examines the nature of investments of migrants in the origin country. Migrant respondents are from the Kwahuman Traditional Area and the Upper East Region of Ghana. Ghana is used as a case study to examine this phenomenon both from the perspective of the migrant and that of the origin country partners. Narratives by migrants are examined in order to unearth factors that inform their decision-making and the approaches they adopt to ensure accountability. Survey results are also used to highlight associations between key variables. The findings indicate that the bulk of the expenditure on productive activities by migrants takes place outside of household circles. Consequently, productive uses of migrant remittances are grossly under-reported due to a lack of trust between migrants and beneficiaries in the origin country. Keywords trust, diaspora, development, hometown associations, diaspora- homeland relations. Introduction Globally, migration of people across national borders has increased for a variety of reasons. Recent statistics indicate that approximately 244 million international migrants participated in this process in 2015 (UNDESA, 2016). *Lecturer, Centre for Migration Studies, University of Ghana and Research Associate, Refugee Studies Centre, Department of International Development (QEH), University of Oxford. Email: lkandilige@ug.edu.gh AHMR, Vol.3 No1, January-April 2017 672 This substantial number of international migrants is associated with equally considerable amounts of remittances. Global remittance flows totalled about $601 billion in 2015, with developing countries receiving about $441 billion, according to the World Bank (2016). As migrant populations congregate in common destination locations, some of them coalesce around common symbols of identity and belonging. With time, this strength of association increases as members of migrant communities (from a common origin) feel that they are not fully accepted by the destination community and they develop a nostalgic feeling about their roles in helping develop their communities of origin. These migrant communities, over time, are designated as ‘diaspora’ in recognition of their deliberate decision to assert their rootedness in their societies of origin. Diaspora relations with the homeland are constructed as mostly positive in terms of the potential development resources that could be granted to origin countries (Castles et al., 2014; Kandilige, 2012; Mohan, 2008). However, in some circumstances, diaspora groups are perceived as potential security threats to both the origin and destination countries due to their sometimes non-transparent and activist relations within their homelands (Baser, 2015). These two perspectives highlight the value of trust in the sustenance of mutually beneficial relations between diasporas and their homelands. This paper seeks to examine the shifting bases and prerequisites of trust building and consider the prospects of replicating these structures in a transnational setting. The concept of transnationalism refers to “the process by which transmigrants, through their daily activities, forge and sustain multi-stranded social, economic, and political relations that link their societies of origin and settlement, and through which they create transnational social fields that cross national borders” (Basch et al., 1994: 7). The discussion is framed around the key questions: What is the role of trust in micro and meso level interpersonal relationships? How do ‘lower level’ trust relationships feed into the transnational setting? What are the prospects and/or dilemmas for transnational trust to work in practice? A multiple case study methodology and mixed methods approaches were adopted in the collection and analysis of primary data (Bryman, 2012; Teye, 2012). The paper is arranged into five main sections. After this introduction, the different interpretations of the concept of diaspora are assessed. An attempt is made to define development from both the perspective of human wellbeing and that of traditional market-focused economics. In addition, the concept of AHMR, Vol.3 No1, January-April 2017 673 trust is examined in order to highlight its role in fostering relations between diaspora members and development partners in home countries. Secondly, a review of diaspora-homeland relations in Ghana is presented in order to highlight general perspectives on the role of the diaspora in Ghana’s development. Thirdly, the methodology adopted for study is discussed. Fourthly, findings on the experiences of Ghanaian migrants in the UK are presented in order to highlight the impact of trust deficit in diaspora- homeland relations at the transnational level. Lastly, conclusions are drawn based on the migrants’ narratives and perspectives from partners in the origin country. Conceptualising Diaspora, Development and Trust Defining Diaspora The term diaspora has been subjected to multiple definitions in the social sciences. For instance, it has been used as a figurative designation to describe alien residents, expellees, political refugees, expatriates, migrants and ethnic and racial minorities (Safran, 1991). Commonness of place of origin, source of identity and mode of dispersion (voluntary or involuntary) of “diasporans” (Vertovec, 2006) characterise the Ghanaian diaspora. Place of origin is sometimes defined at different spatial levels by migrants. As a result, the national attribute of ‘Ghanaianess’ serves as a higher identifying characteristic to migrants in a foreign country than their ethnic or clan affiliations. Cohen (1997; 2008) sub-divides diaspora into “victim,” “trade” and “labour” diaspora in an attempt to reflect the different reasons that sometimes inform migration decisions in the first instance. The first wave of Ghanaian emigrants in the mid- 1960s migrated for economic reasons to other West and Southern African countries (Anarfi & Kwankye, 2003) and formed “labour” and “trade” diasporas. The second substantial wave of emigrants fled the country during periods of political upheaval in the late 1970s and 1980s. Over time, these individuals coalesced into a “victim” diaspora in countries such as the United Kingdom, the USA and Canada. Contemporary movements comprise of mostly labour migrants and this cohort of emigrants has bolstered the Ghanaian labour diaspora. Scholars such as Safran (1991: 83-84) insist on specific characteristics that a given society must possess before being described as a diaspora. Going by Safran’s (1991) detailed set of requirements, the Ghanaian migrant communities abroad might not fulfil all of the criteria. However, others such as Clifford (1994) and Dufoix (2008) advocate a more liberal approach. AHMR, Vol.3 No1, January-April 2017 674 Clifford (1994: 305), for instance, acknowledges that “societies may wax and wane in diasporism.” This alludes to the organic nature of feelings of attachment to the homeland, depending on existential factors in migrant communities’ relations with their destination countries. Dufoix (2008: 19-34) also points out that diasporas should not be perceived as pre-existing groups that have static features that meet or do not meet specific academic criteria, but rather that they can be “heterogeneous populations that are self- consciously imagined” and developed into collectives through “the projects of émigrés and states.” Marienstras (1989: 125) introduces a temporal dimension to diaspora formation. In line with this, Koser (2003) refers to the Ghanaian diaspora as a “neo-diaspora,” based on its relative newness compared to others such as the classic Jewish diaspora. Kleist (2008), however, argues against what she perceives as the undue focus on migrant communities defined by dispersion and, rather, proposes that the term diaspora should be conceptualised as “a concept of a political nature that might be at once claimed by and attributed to different groups and subjects” (Kleist, 2008: 308). Ghanaians abroad increasingly claim the label ‘diaspora’ as a political statement of their affinity to a country experiencing socio-economic development partly attributable to the discovery and production of oil and a sustained period of democracy (Wong, 2013). Conversely, the government of Ghana attributes the label ‘diaspora’ to nationals abroad with an aim of attracting development resources. In spite of discrepancies in how diaspora is conceptualised, the commonalities in definitions refer to individuals that form a community outside of their country of origin due to a range of factors and are either unable or unwilling to return ‘home’ on a permanent basis, but hold the prospect of doing so in the future. These are also people who perceive a sense of belonging to and a need to contribute to the development of their origin country. In the case of Ghana, nationals living abroad have routinely attempted to have both virtual and physical presence in the country’s development agenda. However, there is no common understanding of what the ‘development’ they seek to contribute to entails, how this endeavour could be realised and under what conditions. Defining Development The association between diaspora and homeland development could, therefore, vary depending on how ‘development’ is defined. Both scholars and practitioners have subjected the concept of development to multiple AHMR, Vol.3 No1, January-April 2017 675 interpretations. For instance, within the field of migration studies, Basok (2003) defines development to include activities linked to economic growth- related variables – such as the generation of employment and increase in agricultural production – and welfare-related variables, such as reduction of poverty, increase in average income and decrease in inequality. Others note that development needs to be examined closely through the prism of agency-oriented interpretations of human wellbeing (de Haas, 2009; Sen, 1999; Nussbaum, 1992; Griffin, 1986). This conceptualisation is markedly different from definitions by classical development economists such as Rostow (1960) and Lewis (1955), who measure development by economic growth, especially the increase in market activities. On his part, Amartya Sen (1999) argues that the prime focus of development discourses should be on how to maximise substantive freedoms such as access to education, good nutrition, shelter, political participation and healthcare. He perceives these freedoms as basic yet integral to any claims of development. This perspective of development is consistent with that of Dudley Seers (1969) who famously defined development as “the reduction and elimination of poverty, inequality and unemployment within a growing economy.” Ghana has been lower middle-income country since 2011, with an economic growth rate of 14.4 percent in 2011 that made it one of the fastest growing countries in the world, boasting a US$1,580 per capita income and one of the highest GDP per capita in West Africa. Despite this, Ghana continues to experience inequalities, especially between the northern and southern regions of the country (Osei- Assibey, 2013). As de Haas (2009: 5) aptly concludes, development is not only a complex multi-dimensional concept, but can also be assessed and analysed at different levels and has varying meanings within different normative, cultural and historical contexts. How the seemingly uncontroversial concept of development is defined has implications in terms of engendering trust between the promoters and beneficiaries of that development. Diasporas, as agents of development, do not imply a carte blanche relationship with development partners in origin countries. Trust is critical. Defining Trust Just as with the concepts of development and diaspora, trust means different things to different scholars. Mohan (2006), for instance, situates trust within the broader concept of obligation. He examines this within Parekh’s (1996: 264) conceptualisation of obligation as “social actions that the moral agent AHMR, Vol.3 No1, January-April 2017 676 ought to undertake and his failure to do which reflects badly on him and renders him liable to social disapproval.” Within this interpretation of obligation, the moral agent is defined flexibly to include the migrant, non- migrant or former migrant who is expected to be self-critical and conscious of the impact of his or her actions on society. Failure to deliver on their socially prescribed obligations leads to the lost of credibility and incurs social disapproval. Within the Ghanaian context, at the local and community levels, interpersonal relations and social transactions are characterised by mutually dependent social obligations. As such, Mohan (2006) regards trust as central to the mutual exchange of resources and information among ascribed ethnic groupings in Ghana. Lyon (2000: 665) further notes that the sources of these socially prescribed obligations are located in “reputations, sanctions and moral norms.” However, what are the prospects of (re)producing these obligations in a transnational setting? In discussing transnational activities of Cameroonian and Tanzanian home associations, Mercer et al. (2009) cite examples of how the absence of transnational trust sometimes leads to tensions between migrants and home community members, particularly around community development projects. Therefore, the authors perceive trust as key in sustaining both the negotiation and implementation of such projects. More importantly, Smith and Mazzucato’s (2009: 669) work conceptualises trust in transnational relationships as stemming from long-standing relationships created through shared past experiences and reciprocal economic and social investments in one another. They note that relationships of trust established between migrants and friends are freer from social obligations than those with family relations. According to Smith and Mazzucato, when transactions fail, it is easier to apply sanctions on friends than on family members, due to a feeling of entitlement on the part of family members. The different conceptualisations of trust ultimately form a subset of the broader discourses on “social capital” (Putnam, 2000) and are used to guide the Ghanaian case study in order to unearth the particular challenges migrants face in negotiating the preconditions of trust within a transnational setting. Perceptions on Diaspora-Homeland Development The diaspora is increasingly being courted as a potential development partner in Ghana (Kandilige, 2012; Mohan, 2008). As a result, specifically earmarked contributions by the diaspora have been incorporated into recent national development plans in the country (for instance, see: NDPC, 2015; NDPC, 2010; AHMR, Vol.3 No1, January-April 2017 677 NDPC, 2005). At a practical level, diaspora-homeland relations find expression in the political rhetoric, civil society discourses, legislative enactments and initiatives by international development agencies such as the International Organisation for Migration (IOM) and the United Nations Development Programme (UNDP). The aspirations of the political elites in encouraging an increased role for the Ghanaian diaspora can be captured by two important observations. The first one is by a former Ghanaian president, John Agyekum Kufuor: I must acknowledge the contributions made by our compatriots who live outside the country... Many of you do more than send money home, many of you have kept up keen interest in the affairs at home and some of you have even been part of the struggle of the past twenty years (Mohan, 2006). The second is by a former Minister of Finance, Osafo-Maafo: May I humbly invite Ghanaians overseas to use the natural advantage they have over their home based countrymen such as proximity and access to the latest technology, foreign exchange, reliable export markets and partners with know-how to begin to make direct investment into our economy (Mohan, 2008). Both quotes point to an appreciation of the transnational nature of international migration and the possible opportunities that could be exploited by the homeland through its ‘extraterritorial’ citizens (Baubock, 2003; Escobar, 2007; Castles et al., 2014). Within the transnational theory, cash and social remittances are acknowledged as important development tools at the disposal of homelands without any express requirement on migrants to return to their countries of origin on a permanent basis. This is a significant departure from previously popular complaints by leaders of developing countries about the “development of underdevelopment” (Binford, 2003; Lipton, 1980) due to brain drain and the ‘poaching’ of skilled African migrants by the developed ‘core’ countries (Pang et al., 2002; Desai et al., 2002; Voigt-Graf, 2008; Chanda, 2001; Dovlo & Nyonator, 1999). These pronouncements are backed by policy formulations and events that have tended to facilitate diaspora engagement processes.* Ghana drafted a National Migration Policy in 2014 (launched in April 2016) in order to effectively manage migration in a way that yields positive development outcomes. In addition, a Diaspora Support Unit was created in 2012, under the AHMR, Vol.3 No1, January-April 2017 678 auspices of the country’s Ministry of Foreign Affairs and Regional Integration. Its specific responsibility was to identify and provide the needed support to the Ghanaian diaspora purposely to increase their interest in national development. This Unit was later upgraded to a Bureau (Diaspora Affairs Bureau) in 2014 – a possible sign of an even greater appreciation of the role of the diaspora in national development. Another concrete step has been the initiation of the process of drafting a national Diaspora Engagement Policy (started in 2015), drawing on the expertise of the Centre for Migration Studies at the University of Ghana and other development partners such as the IOM, the International Centre for Migration Policy Development (ICMPD) and ECOWAS. The transmission of what Peggy Levitt (1996) refers to as social remittances to Ghana by members of the diaspora has also been hugely significant. These have included the transfer of ideas, behaviours, practices and social capital accrued from destination countries. These transfers have been executed by individual members of the Ghanaian diaspora as well as facilitated by international organisations such as the IOM and the UNDP. Individual social remittances have included ideas on democratic governance, transparency and accountability, human rights, punctuality, work ethics and assertiveness, among others. In addition, social capital derived from their membership of business and epistemic networks abroad have enabled the Ghanaian diaspora to promote transnational investments and collaborations in Ghana (Kandilige, 2012). Equally noteworthy is the role of international agencies (especially the IOM and UNDP) that have initiated projects such as the knowledge transfer programmes for the purposes of bolstering socio-economic development in the country. A classic example is the Migration for Development in Africa (MIDA) initiative launched by the IOM in 2001 to assist in the transfer of critical skills and resources of the African diaspora to their countries of origin. Ghana benefited from the circulation of competencies, expertise and experience of the Ghanaian diaspora (Faist, 2008). Another example is the UNDP’s programme referred to as the Transfer of Knowledge through Expatriate Nationals (TOKTEN). This mechanism allowed for the tapping of professional skills of expatriate Ghanaians through the means of short-term consultancies in Ghana. In terms of cash remittances, Ghana has recorded year-on-year increases in the volume of remittances, which has consistently surpassed the ratio of some AHMR, Vol.3 No1, January-April 2017 679 ‘macro’ variables such as Foreign Direct Investment (FDI) and Overseas Development Aid (ODA) to Gross Domestic Product (GDP). For the period from 1990 to 2003, Bank of Ghana data suggest that private unrequited transfers had a significant impact on the country’s GDP. As a percentage of GDP, remittances increased from about 2 per cent in 1990 to about 13 per cent in 2003, and also increased from 22 per cent to almost 40 per cent as a percentage of total exports earnings (Addison, 2004). There was a similar trend of year-on-year increases in the aggregate volume of cash remittances from 2004 to 2014. The Bank of Ghana recorded a rise in remittances from $1.2 billion in 2004, to over $1.9 billion in 2008, to $2.1 billion in 2014, to $4.9 billion in 2015 (Bank of Ghana, 2015; Kandilige, 2012; Bank of Ghana, 2016). The injection of such cash contributions could support economic development in the origin country. However, as the Human Development Report (UNDP, 2009: 71) notes, “impacts are complex, context-specific and subject to change over time.” For development to be triggered in Ghana as a result of diaspora activities, it depends to a large extent on the internal dynamics in the country. As de Haas (2009: 52) concludes, migrants and their remittances can neither be expected to “trigger take-off development nor be blamed for a lack of development in fundamentally unattractive investment environments.” Methodology and Methods A multiple case study research methodology, involving two different geographical areas in Ghana, was adopted to guide this study. The context and processes involved in the activities of members of two Ghanaian hometown associations (Kwahuman Association and Kasena-Nankana Development League) based in the UK were analysed. This enabled a measurement of their peculiarities as well as similarities. These two hometown associations represent the Eastern and Upper East regions of Ghana, respectively. This kind of research strategy has firm roots in classic studies in other social science disciplines (Burgess, 1983; Cavendish, 1982; Sassen, 2006). Mixed methods were adopted for this study that allowed for the triangulation of results and complementarity of techniques (Bryman, 2012). The selection of respondents from the two regions was actualised by tracing leads from migrants belonging to the two selected hometown associations to their home regions in Ghana. Migrant associations provided information on their contacts and partners in the origin communities and they were interviewed for more information on the nature of and basis for their collaborations with migrants. AHMR, Vol.3 No1, January-April 2017 680 The overall logic of the methodology adopted was to glean quantitative data on frequencies, associations, patterns and proportions in addition to qualitative data that helped provide an in-depth explanation for behaviours, decisions and reactions by respondents. The innovation in this study partly stems from the fact that it was conducted in both the origin and destination communities in order to build a holistic perspective on the concept of trust, rather than the single-sited approach often embraced by most migration researchers (Werbner, 2002; Osili, 2004; Mohan, 2006; Mercer et al., 2009). The first stage of the study was carried out in the Greater London area (especially the boroughs of Southwark, Lambeth, Newham, Hackney, Haringey, Lewisham, Croydon and Brent). The main reason for the choice of these boroughs was that most Ghanaian migrants are based there (COMPAS, 2004). The second phase was conducted in the Upper East and Eastern regions of Ghana. The Upper East Region is among the poorest of the ten regions of Ghana. It is located in the north, and accounts for comparatively fewer migrants. The Eastern Region is much richer and, located in the south-eastern part of the country, it accounts for one of the largest sources of Ghanaian migrants outside of the African continent (Kandilige, 2012). The differences in economic affluence, migration prevalence and geographical location are important because they provide apt comparative parameters. The fieldwork in the Greater London area included in-depth interviews with twenty key informants from the Ghanaian migrant community, participant observation activities, informal conversations and the administration of 120 questionnaires. The aims of these data collection strategies were to gauge, among others, migrants’ participation in group activities, their transnational support to local communities, their main partners in the origin community, challenges they face and the value they place on the concept of trust in their transnational relationships. The Ghana fieldwork included 20 in-depth interviews with local chiefs, District Chief Executives (local government officials), community development leaders and executive members of local hometown associations (located in urban centres). These interviews focused on how development projects are negotiated with migrants abroad (the UK), how they are implemented, the role of the origin communities in local development, their perceptions on the effects of collective remittances on poverty alleviation and income redistribution and possible new areas of collaboration with migrants. All AHMR, Vol.3 No1, January-April 2017 681 interviews were conducted personally by the author and at respondents’ homes, places of work, restaurants or pubs. In order to gauge the perception of the beneficiary communities, 346 questionnaires were administered among heads of migrant households (246 in the Kwahu Traditional Area and 118 in the Upper East Region). Of these, 66 per cent were male and 34 per cent were female. The data presented in this paper are from a bigger project** carried out over fifteen-months. The paper is based on the narratives of respondents from the two hometown associations and their development partners in Ghana as well as some survey statistics. Findings Trust as A Critical Component in Diaspora-Homeland Relations: The Case Of Ghana This section examines the concept of trust between diasporas and their homelands by using the experiences of Ghanaian migrants in the UK and their local partners as a case study. These are analysed under four main sections: knowledge about the types of investments migrants make, engaging family/relatives to carry out projects, migrants’ individual experiences and hometown associations’ group experiences of initiating and executing migrant-funded projects. Knowledge about Investments in General Empirical research among 364 migrant households in Ghana indicates that the bulk of remittances received, according to heads of migrant households, were predominantly used to pay for education, daily upkeep and solving family emergencies (see Figure 1). Almost 44 per cent and 34 per cent of respondents in Kwahu Traditional Area and the Upper East Region, respectively, reported using remittances for education purposes. This is consistent with international research, which also suggests a positive relationship between remittances and educational attainment and enrolment (Rapoport & Docquier, 2005; Cox- Edwards & Ureta, 2003). AHMR, Vol.3 No1, January-April 2017 682 n=246 Figure 1. Purpose of remittance. There was, however, a discrepancy between what migrants themselves and heads of migrant households claimed about how remittances were actually spent. On the one hand, heads of migrant households generally claimed that the bulk of remittances were spent on ‘consumptive’ expenditures (Connell & Conway, 2000). On the other hand, migrants indicated that only a small percentage of their total remittances were sent to migrant households for such ‘consumptive’ expenditures. So what accounts for this apparent disjuncture? There could be a myriad of reasons for this disparity but the most common refrain by migrants was that they did not trust their families or relatives in Ghana to run their businesses honestly and properly and, as a result, failed to declare such projects to them. This suggests that some heads of migrant households were either not informed about some of the business ventures and investments (‘productive’ expenditures) their migrant relatives own in Ghana, or were aware but not involved in the running of such ventures. To test this assertion, an analysis of the number of migrants who had funded, already set up or were in the process of setting up businesses in Ghana was carried out. Where 32 per cent of respondents (migrants in the UK) reported having funded, set up or were in the process of setting up businesses in Ghana, less than 9 per cent of heads of migrant households in Ghana reported that their migrant relatives had made such ‘productive’ investments. This implies that 0 5 10 15 20 25 30 35 40 45 50 P e r c e n t a g e Kwahu Upper east AHMR, Vol.3 No1, January-April 2017 683 the bulk of the expenditure on ‘productive’ activities by migrants takes place outside of household circles. Furthermore, when a direct question was posed on investment in ‘productive’ activities in general, up to 84 per cent of respondents (heads of migrant households) in the Upper East Region and 67 per cent in the Kwahu Traditional Area believed that their transnational relatives had neither businesses nor stores in Ghana (see Figure 2). This marked disparity suggests the need for empirical research into the relationship between remittances and development to consider both the perceptions of the receiving households or communities and the views of the migrants themselves. This is critical because a lop-sided examination of the extent of migrant investments in homelands, only from the perspective of heads of migrant households without a matched sample from migrants, is likely to underestimate the magnitude of ‘productive’ investments. n= 353 Figure 2. Perceived investment habits of UK-based Ghanaians by heads of migrant households). Knowledge about Investments in Housing and Residential Arrangements of Migrants International empirical research suggests that a large proportion of remittances are spent on housing-related expenditure in migrants’ home countries, generally. This has been the case in countries such as Morocco, where between 71 per cent (de Haas, 2003) and 84 per cent (Hamdouch, 67 20 9 4 0 84 12 3 0 0 0 10 20 30 40 50 60 70 80 90 No Yes, set up a business Yes, opened a store Don't know Other Kwahu Upper east AHMR, Vol.3 No1, January-April 2017 684 2000) of remittances have been spent on housing, and Egypt where 54 per cent of remittances are spent on housing (Adams Jr., 1991). Also, Osili (2004) reported that a large proportion of remittance income to Nigeria is spent on housing. He concludes that a ten (10) per cent increase in remittance income in Nigeria raises the probability of investing in housing by three (3) percentage points. Consistent with these high percentages, the study found that over 81 per cent of respondents in the UK had investments in either private housing or real estate development as a business (see Table 1). Additionally, more than a third of the UK respondents have either set up or were in the process of setting up a business or a store in Ghana. Table 1. Investment patterns of Ghanaian transnationals in the UK. Sector Frequency Percentage Housing only 59 49% Housing plus business and stores 38 32% Total housing 97 81% No investments 16 13% No response (refused to answer) 7 6% Total 120 100% Beyond some productive investment interests of Ghanaian migrants being concealed from members of their immediate families, there also appear to be gaps in knowledge about the residential arrangements of migrants in the origin community. There has been an appreciable chunk of migration literature on the apparent conspicuous nature of migrants’ investments in housing, especially second houses (Mohan, 2006; de Haas, 2007; Fadloullah et al., 2000 in de Haas, 2009 and Van der Geest, 1998). These residential edifices are preserved as a sign of prestige by migrants and only occupied very occasionally when they are on holiday to the origin community or when they are attending special events such as funerals, weddings, anniversaries or religious and cultural celebrations. A classic example exists in the residential AHMR, Vol.3 No1, January-April 2017 685 arrangements of one migrant (Kojo) from the Kwahu Traditional Area in the Eastern Region of Ghana, who owns a spacious six-bedroom house (see Figure 3) on an exclusive migrants’ residential enclave. This property contains three washrooms, two living rooms, a mini bar and two garages, among other amenities. Interestingly, up to ten months in a year, a caretaker occupies this ‘mansion.’ Kojo’s UK-based nuclear family only visits Kwahu over the Easter festive period to participate in paragliding and some cultural celebrations. In contrast, Kojo’s mother, eight siblings, nephews and nieces all live in a rundown mud house (see Figure 4) located two towns away from the migrants’ residential enclave. He indicates that his family in Ghana is unaware of the existence of his opulent house and that his decision to keep it a secret stems from his anxiety over likely excessive demands for money by members of his left-behind household (see Mohan, 2008; Henry & Mohan, 2003), requests from them to occupy his property on a ‘temporary’ basis and his fear of envious neighbours and family members killing him using juju (voodoo). This is an extreme example of trust deficit in migrant-homeland relations. However, it is another example of the difficulty of sustaining trust relations in a transnational context. Kojo’s account challenges Smith and Mazzucato’s (2009) conceptualisation of trust since in spite of the long-lasting relationship between Kojo and his family in Ghana, which should have informed a reciprocal economic and social investment from him, he fails to replicate this at the transnational level. He, however, trusts an outsider (caretaker) over his family, in line with Smith and Mazzucato’s views that social obligations are freer with outsiders than with family members and sanctions are easier to apply on outsiders. Within Parekh’s (1996) conceptualisation of obligation, Kojo’s actions have the propensity to be judged negatively by his community and to incur social disapproval. AHMR, Vol.3 No1, January-April 2017 686 Figure 3. Posh migrants’ mansions (mostly occupied by caretakers). Figure 4. Rundown accommodation occupied by parents and left-behind household members of a migrant. Trust Deficit in Supervisory Roles Misgivings expressed by migrants about informing members of their left- behind households of their investment interests are further corroborated by survey results among migrant households in Ghana in three main areas (running of businesses on behalf of migrants, supervising building projects and clearing goods from harbours). These areas are selected as important in testing the level of migrants’ trust in their families and relatives in Ghana, because they are activities that involve large amounts of capital or cash transactions. Less than 20 per cent of heads of migrant households reported AHMR, Vol.3 No1, January-April 2017 687 any involvement in the running of businesses or clearing of goods on behalf of their relatives in the UK. The supervision of building projects, whilst slightly higher than the other two activities, still accounts for less than half of all cases (see Table 2). These total figures are slightly lower than in cases of city-based migrant household members where up to half are involved in supervising housing construction for migrants, according to previous research (Mazzucato, 2011: 460). The increasing reliance on friends, associates and former work colleagues to supervise projects and to conduct business transactions on behalf of migrants is again consistent with Smith and Mazzucato’s (2009) assertion that obligations with ‘outsiders’ are freer than those with family members. However, beyond the absolute percentages of heads of migrant households who are involved in the three key activities, inferential statistics (chi-square test) is used to determine whether there is a statistically significant difference between the two regions in terms of the level of trust. The results (Table 2) show that there is a statistically significant association between migrants’ trust in relatives to supervise their buildings and the hometown of the migrant, χ2 (1, N= 363) = 13.425, p < 0.001). However, the results (Table 2) indicate that there is no statistically significant association between migrants’ trust in relatives to run a business for them and the hometown of the migrant, χ2 (1, N= 364) = 1.012, p > 0.001). Furthermore, the results demonstrate that there is no statistically significant association between migrants’ trust in relatives to clear goods for them and the hometown of the migrant, χ2 (1, N= 362) = 1.533, p > 0.001). Table 2: Level of involvement in capital-intensive activities on behalf of migrants. Supervise a building (Dependent Variable) Home town (Independent Variable) No Yes Eastern Region (Kwahu Traditional Area) 189 (73.5 %) 57 (53.8 %) Upper East Region 68 (26.5 %) 49 (46.2 %) AHMR, Vol.3 No1, January-April 2017 688 Total 257 (100.0 %) 106 (100.0 %) χ2 statistic = 13.425, df = 1, N = 363, *p = 0.000 < 0.001 Run a business Home town (Independent Variable) No Yes Eastern Region (Kwahu Traditional Area) 214 (68.6 %) 32 (61.5 %) Upper East Region 98 (31.4 %) 20 (38.5 %) Total 312 (100.0 %) 52 (100.0 %) χ2 statistic = 1.012, df = 1, N = 364, p = 0.198 > 0.001 Clear goods Home town (Independent Variable) No Yes Eastern Region (Kwahu Traditional Area) 201 (66.3 %) 44 (74.6 %) Upper East Region 102 (33.7 %) 15 (25.4 %) Total 303 (100.0 %) 59 (100.0 %) χ2 statistic = 1.533, df = 1, N = 362, p = 0.138 > 0.001 Source: Author’s fieldwork. AHMR, Vol.3 No1, January-April 2017 689 Trust: Migrants’ Personal Experiences Several Ghanaian migrants in the UK genuinely feel let down by their own family and relatives and have resorted to relying on help from friends and former colleagues to run and manage their businesses and building projects (see also Smith & Mazzucato, 2009: 667-669). This seems to account for the marked disparity in the perceptions of members of migrant households and the reality on volumes of ‘productive’ investments undertaken by migrants. Personal accounts help contextualise the levels of mistrust. For instance, ‘Ibrahim’s’ own sister misappropriated his funds: When I started [a housing project], I gave my sister 30 million [Cedis] to start but when the old man died and I went home, I asked her where’s the house and she said she spent the money... My own sister, one mother one father (sic). No! So I don’t allow them to go even near my things (56 year-old male migrant, Upper East Region). ‘Yaw’ was also cheated and betrayed by his relative: That has been a problem honestly ... it started off with my sister-in-law she actually bought the land and you know Ghana the way they are, I even realised that they inflated the price of the land like five times. Then she told me the boyfriend was a contractor. When I was sending the money they were not even using the money to do the project, she was now using it for her own thing. So I had to get rid of that contractor, get another person to do it, so the initial money I spent was just wasted (61 year-old female migrant, Eastern Region). These two quotes demonstrate the sense of frustration and despair that characterise some relations between migrants and their kin. Family members and relatives are expected by society to observe relations that Parekh (1996) conceptualises as relations of obligation. Failure to adhere to these socially constructed obligations reflects badly on them and attracts social disapproval. Accounts of outright disregard for the investment capital of migrants, some of whom have experienced incidents of discrimination and racism in the host country (see Herbert et al., 2006), have engendered a feeling of resignation on the part of some migrants. This despair is evident in what a female migrant from the Kwahu Traditional Area had to say: You can’t help Ghanaians. That’s one thing you just cannot. Honestly they take too much advantage and they think we don’t know what goes on, we do. Even though we’ve been away from the country for a long time, AHMR, Vol.3 No1, January-April 2017 690 we still know what goes on (Secretary of the Kwahuman Association, UK). Similar expressions of doubt and mistrust discourage some migrants from investing in the origin country altogether. Some of those migrants who invest in private accommodation units reported either using the services of private real estate developers to construct their houses or purchasing already built houses in an attempt to avoid embezzlement of funds by family relations. That is the difficult side of it. As I said you send them money and they embezzle it so it is difficult for us people living over here to put up houses because your own brother or sister or best friend that you trust to do something for you will clean you out (58-year old female migrant, UK). Trust: Experiences of Migrant Collectives Versus Homeland Partners Feelings of mistrust are a double-edged sword and examples of such feelings also exist from the perspective the origin country. In-depth interviews with chiefs, community development leaders and members of local-based hometown associations in Ghana suggest both cordial relations with migrants in the process of negotiation and implementation of migrant-funded community-based projects, but also tensions between migrants and home communities. On the one hand, instances of prior discussions between migrants and local stakeholders as well as collaborative work have been recorded. These were more prominent among the Kwahuman Association members and their Ghanaian partners. Examples include detailed discussions of proposed projects in origin communities between migrants and traditional leaders during migrant-funded overseas trips by traditional chiefs. They sometimes invite Nana [the chief] and his elders to visit the UK but especially Holland and the USA. They pay for all their travel expenses so that Nana and his elders will go over to brief them on what is going on, on the ground. They then fundraise and send the money to support whatever projects they agree on (Linguist to the Paramount Chief, Abene, Eastern Region). Such cordial deliberations have helped cement relations between members of the Ghanaian diaspora and development partners from their origin communities. These negotiations form a basis for the transfer of both cash and in-kind collective remittances towards community development. For that matter, projects supported are not dissimilar to those reported from other research among the Cameroonian diaspora (Mercer et al., 2009), Pakistani AHMR, Vol.3 No1, January-April 2017 691 diaspora (Werbner, 2002), Mexican diaspora (Smith, 2003) and Moroccan diaspora (de Haas, 2007). These mostly include the donation of used medical equipment, educational materials, street lighting, potable water and computing equipment, the renovation of old public buildings and the setting up of scholarships for local students. Two examples of donations to health facilities are provided: They [migrants] used to send clinic or hospital equipment especially beds, mattresses, wardrobes and other materials that are being used at the hospital. The other day they brought some one or two containers full of hospital equipment alone (sic) ... They brought incubators, bicycles for people with cardiovascular problems and so on and they are all at the hospital now (Chief of Abetifi, Kwahu, Eastern Region). They brought about 50 sets of beds. The chief gave them a place to store them and when some are broken then they go and replace them from the stores. Secondly, an electric plant was donated to the clinic, a generator to the clinic so that when the lights go off they can use it. When the machine arrived they called Nana [the chief] and everyone in the town and he inaugurated it. We the local association here in Abetifi have also built a shed to cover it in order to protect it from the elements (Sub-Chief of Abetifi, Kwahu, Eastern Region). On the other hand, mistrust is manifested, transnationally, in differences in the value placed on remittances by origin partners as opposed to diaspora members. Origin partners complain of over-estimation of the value of collective remittances sent by members of the diaspora and a lack of appreciation for the magnitude of contributions made by local-based hometown associations towards community development. Diaspora members are accused of placing unreasonable demands on local counterparts. This feeling is demonstrated in two quotes from representatives of two migrant communities in Ghana. I actually run into problems with the people in the UK. If you send $100 or £100 and you think that it is a lot, and that’s for the year. When I gave an example of how much my wife and I alone have contributed to development of our town, they took offence but I was just doing some analysis. When you compare their earnings and ours, they should be doing more. So they [migrants abroad] should not feel that if they send £100 that is so much money (Chairman of a local hometown association, Upper East Region of Ghana). AHMR, Vol.3 No1, January-April 2017 692 I think they [migrants abroad] brought in $2000. At that time it was the equivalent of about 16 Million Cedis. When you hear 16 Million Cedis it sounds big but when it goes to the ground it can’t do much (sic). So sometimes that is the problem we have with them [migrants] because they find it difficult to understand why they bring in the money and they don’t see what it has been used for. Because they could not understand this, they refused to top it up and since we were also having our different projects going on, we were not be able to raise extra money to do what they wanted us to do (Community Development Leader, Kwahu, Eastern Region). Migrant collectives (hometown associations), however, insist that their incredulities or suspicions are based on actual experiences of cases of embezzlement of collective remittances by some local counterparts. There are also accounts of lack of transparency in the selection of community representatives and the refusal by others to publicly acknowledge receipt of collective remittances (see Mercer et al., 2009). A representative of the one of the selected hometown associations in the UK aptly portrays these claims in the following statement: Yes, that was a good plan to build a big roof supermarket…So we started with stage one which according to them the government of Ghana had given them about 350 million Cedis so we donated 170 million Cedis which, you know, will come to half a billion to start the project. They were rushing us. Our people ordered me to go and present to them the 170 million Cedis, which was the equivalent of £10,000 at that time and the money is gone astray! (Treasurer, Kwahuman Association, UK). Members of migrant collectives also hold their origin country partners to a high standard of openness and accountability in line with socialised values, probably cultivated in the host country (the UK). Some migrant groups demand legitimacy and representativeness of local community groups as a condition for continued funding. This assertion is partly borne out by demands such as these: I demanded certain guarantees from them because I needed to make sure that the election of people onto that committee was fair and that whatever they did they had a constitution that guided them as to, you know, what they wanted to work on, and I wasn’t really impressed with the fact that they were just hand picking people to sit on it. So I decided to kind of step back for now … I’ve worked in this kind of area for many years here in the UK so I am fairly aware of what can go wrong if you don’t get the group or groups set up properly with AHMR, Vol.3 No1, January-April 2017 693 the right representation. You could cause lots of problems (52-year old female migrant, Upper East Region). In addition, migrants request evidence of receipt and use of collective remittances as a monitoring tool but also as a useful advertisement to future donors. Forms of evidence range from audio clips of radio broadcasts, footage of TV coverage, photographs in print media or on the internet and formal acknowledgement of receipt in writing on headed paper by development partners. Failure to deliver on these requests sometimes leads to mistrust and frustration on the part of migrants. These socially prescribed obligations are located in “reputations, sanctions and moral norms” (Lyon, 2000: 665). This is captured in the remarks made by one fundraiser for the Kwahuman Association: The reason I haven’t continued to fundraise is, you know, when you collect this fundraising money these people [British] they want evidence to see that you haven’t spent the money. So initially I sent £500 to them [local partners] so now I’m waiting for them even to send me a picture to prove we’ve done the foundation or we’ve done this or that. Every time I phone, I don’t get any word from them (sic) (61-year old female migrant fundraiser, Eastern Region). Cases of mistrust emanating from lack of publicity on migrants’ collective remittances are not limited to the Ghanaian context. An instructive example exists in Mercer et al.’s (2009: 154) account on Cameroonian and Tanzanian home associations. According to them, the Bali Cultural and Development Association, UK (BCDA-UK) cut its links with the Bali Nyonga Development and Cultural Association (BANDECA) in the homeland because the BANDECA’s water department had failed to acknowledge the BCDA-UK’s donation of 500,000 CFA francs (about £500) in its published list of donors. Conclusion This paper has highlighted the limited involvement of familial relations in the execution of productive investments that are funded by diaspora members. While a substantial proportion of remittances are directed at funding the cost of education, healthcare and daily upkeep of migrant households in origin communities, the paper demonstrates that the bulk of expenditure on ‘productive activities’ takes place outside of household circles. Even though ‘consumptive expenditure’ could generate long-term multiplier effects beyond the immediate recipients (de Haas, 2005), ‘productive expenditure’ yields direct, immediate effects on job creation and improvements in living AHMR, Vol.3 No1, January-April 2017 694 standards. This questions the scope of New Economics of Labour Migration theorists’ interpretations of the household as the most appropriate unit of analysis of migration as a livelihood strategy. Lack of trust towards family members has negative implications on the potential of remittances as household poverty alleviating resources. The paper also finds that there are some similarities in the preconditions and basis of constructing trust in social relations both within the local and transnational contexts. As Lyon (2000: 665) notes, fundamental sources of socially prescribed obligations are located in “reputations, sanctions and moral norms.” The risk of reputational damage, imposition of social sanctions (real or perceived) and the ascribed normative culture form the basis of trust within the local sphere. These prerequisites are equally valid in the transnational context but geographical proximity and the attendant immediacy of effects of social sanctions on the social actor engender greater intensity in trust relations locally compared to transnationally. Also, Parekh’s (1996: 264) conceptualisation of obligation as “social actions that the moral agent ought to undertake and his failure to do which reflects badly on him and renders him liable to social disapproval,” has limited application in the transnational context. While social actions around the provision of consumptive goods are critical in migrants’ transnational interrelationships, substantial productive investments are broadly based on “long-standing relationships created through shared past experiences and also reciprocal economic and social investments in one another” (Smith & Mazzucato, 2009: 669). These transnational relationships do not have to be familial. To this extent, how much of the estimated $601 billion remitted globally in 2015, out of which $441 billion went to developing countries (World Bank, 2016), was actually directed at productive uses? This and the level of involvement of migrant households in managing productive investments ensuing from these remittances are difficult to ascertain by conducting single-sited empirical research only among migrant households in origin countries. The Ghana example is instructive as well as illustrative of the nuances that international/intergovernmental development agencies, such as the UNDP, the IOM, the World Bank, the European Union (EU) governments and others, need to bear in mind when advocating for a greater role for diasporas as development partners in the developing country context. A firm appreciation of the complex dynamics in establishing and sustaining trust relationships AHMR, Vol.3 No1, January-April 2017 695 transnationally is critical to the success and viability of institutional interventions within the migration-development nexus framework. Ultimately, there is a need for a reconceptualisation of the role of migrants beyond the narrow and undifferentiated prescriptions attributed to them by intergovernmental organisations and international financial institutions. References Adams, R. Jr. 1991. The Effects of International Remittances on Poverty, Inequality and Development in Rural Egypt. Research Report 86. Washington, DC: International Food Policy Research Institute. Addison, E.K.Y. 2004. The Macroeconomics Impact of Remittances, Bank of Ghana. AITE (2004) Consumer Money Transfers: Powering Global Remittances. Anarfi, J. and Kwankye, S. 2003. Migration from and to Ghana: A Background Paper. University of Sussex: DRC on Migration, Globalisation and Poverty. Bank of Ghana. 2016. Remittances and unrequited transfers for 2015. Research Department. From: (retrieved November 20 2016). Bank of Ghana. 2015. Remittances and unrequited transfers for 2014. Research Department. From: (retrieved November 5 2015) Baser, B. 2015. Diasporas and Homeland Conflicts: A Comparative Perspective. Surrey: Ashgate Publishing Ltd. Basch, L., Schiller G.N. and Szanton-Blanc, C. 1994. Nations unbound: Transnational projects, postcolonial predicaments, and deterritorialized Nation-States. Langhorne, PA: Gordon & Breach. Basok, T. 2003. Mexican seasonal migration to Canada and development: A community based comparison. International Migration, 41(2): 1-26. Baubock, R. 2003. Towards a political theory of migrant transnationalism. International Migration Review 37(3): 700-723. Binford, L. 2003. Migrant remittances and (under) development in Mexico. Critique of Anthropology, 23: 305-336. Bryman, A. 2012. Social Research Methods. 4th Edition. Oxford: Oxford University Press. http://www.bog.gov.gh/ http://www.bog.gov.gh/ AHMR, Vol.3 No1, January-April 2017 696 Burgess, R.G. 1983. Inside Comprehensive Education: A Study of Bishop McGregor School. London: Methuen. Castles, S., de Haas, H. and Miller M.J. 2014. The Age of Migration: International Population Movements on the Modern World. Houndmills Basingstoke: Palgrave Macmillan. Cavendish, R. 1982. Women on the Line. London: Routledge & Kegan Paul. Chanda, R. 2001. Trade in health services. Paper WG4: 5, Commission on Macroeconomics and Health. (retrieved 10 May 2014). Clifford, J. 1994. Diasporas. Cultural Anthropology, 9(3): 305-6. Cohen, R. 1997. Global Diasporas: An Introduction. London: UCLA Press. Cohen, R. 2008. Global Diasporas: An Introduction. Abingdon: Routledge. COMPAS Report Team. 2004. The contribution of UK-based diasporas to development and poverty reduction. Oxford, ERSC Centre on Migration, Policy and Society (COMPAS). Connell, J. and Conway, D. 2000. Migration and remittances in island microstates: A comparative perspective on the South Pacific and the Caribbean. International Journal of Urban and Regional Research, 24(1): 52-78. Cox Edwards, A. & Ureta, M. 2003. International migration, remittances, and schooling: Evidence from El Salvador. Journal of Development Economics, 72(2): 429-61. De Haas, H. 2009. Mobility and Human Development. Human development research paper, UNDP Human Development Reports. UNDP. De Haas, H. 2007. The Impact of International Migration on Social and Economic Development in Moroccan Sending Regions: A Review of Empirical Literature. Working Paper No.3, University of Oxford: International Migration Institute (IMI). De Haas, H. 2005. International Migration, Remittances and Development: Myths and Facts. Third World Quarterly, 26(8): 1269-1284. De Haas, H. 2003. Migration and Development in Southern Morocco: The Disparate Socio-Economic Impacts of out-migration on the Todgha Oasis Valley. Ph.D. Thesis, Catholic University of Nijmegen. AHMR, Vol.3 No1, January-April 2017 697 Desai, M.A., Kapur, D. & McHale, J. 2002. Sharing the spoils: taxing international human capital flows. Updated version of the paper presented. Paper No. 02-06, Weatherhead Center for International Affairs, Harvard University. Dovlo, D. & Nyonator F. 1999. Migration of graduates of the University of Ghana Medical School: A preliminary rapid appraisal. Human Resources for Health Development Journal, 3(1): 34-37. Dufoix, S. 2008. Diasporas. Berkeley, CA & London: University of California Press. Escobar, C. 2007. Extraterritorial political rights and dual citizenship in Latin America. Latin American Research Review, 42(3): 43-75. Faist, T. 2008. Migrants as transnational development agents: an inquiry into the newest round of the migration-development nexus. Population, Space and Place, (14): 21-42. Griffin, J.P. 1986. Well-Being: Its Meaning, Measurement and Moral Importance. Oxford: Clarendon Press. Henry, L. and Mohan, G. 2003. Making homes: The Ghanaian diaspora, institutions and development. Journal of International Development, 15: pp. 611-622. Hamdouch, B. 2000. The remittances of Moroccan emigrants and their usage. In: OECD. 2005. The Development Dimension: Migration, Remittance and Development. Paris: OECD Publishing. Herbert, J., Datta, K., Evans, Y., May, J., McIlwaine, C. & Wills, J. 2006. Multiculturalism at Work: The Experiences of Ghanaians in London. London: Department of Geography, Queen Mary, University of London. Kandilige, L. 2012. Transnationalism and the Ghanaian Diaspora in the UK: Regional Inequalities and the Developmental Effects of Remittances at the Sub- NationalLlevel. DPhil. Thesis, University of Oxford. Kleist, N. 2008. Mobilising ‘the diaspora’: Somali transnational political engagement. Journal of Ethnic and Migration Studies, 34(2): 307-323. Koser, K. 2003. New African diasporas: An introduction. In: Koser, K. (Ed.). New African Diasporas. London: Routledge, pp. 1-16. AHMR, Vol.3 No1, January-April 2017 698 Levitt, P. 1996. Social Remittances: A Conceptual Tool for Understanding Migration and Development. From (retrieved 12 May 2009). Lewis, W.A. 1955. The Theory of Economic Growth. London: George Allen and Unwin. Lipton, M. 1980. Migration from the rural areas of poor countries: The impact on rural productivity and income distribution. World Development, 8: 1-24. Lyon, F. 2000. Trust, networks and norms: The creation of social capital in agricultural economies of Ghana. World Development, 28: 663-68. Marienstras, R. 1989. On the notion of diaspora. In: Chaliand, G. (Ed.). Minority Peoples in the Age of Nation-States. London: Pluto, pp. 119-125. Mazzucato, V. 2011. Reverse remittances in the migration–development nexus: two-way flows between Ghana and the Netherlands. Population, Space and Place, 17(5): 454-68. Mercer, C., Page, B. & Evans, M. 2009. Unsettling connections: Transnational networks, development and African home associations. Global Networks, 9(2): 141-161. Mohan, G. 2006. Embedded cosmopolitanism and the politics of obligation: The Ghanaian diaspora and development. Environment and Planning A., 38: 867-883. Mohan, G. 2008. Making neoliberal states of development: The Ghanaian diaspora and the politics of homelands. Environment and Planning D: Society and Space, 26: 464-479. NDPC. 2015. Ghana Shared Growth and Development Agenda (GSGDA) II, 2014-2017, medium-term national development policy framework. Republic of Ghana. NDPC. 2010. Guidelines for the preparation of district medium-term development plans: Ghana Shared Growth and Development Agenda (GSGDA I) 2010-2013. Republic of Ghana. NDPC. 2005. Growth and Poverty Reduction Strategy (GPRS II) (2006-2009). Republic of Ghana. Nussbaum, M.C. 1992. Human functioning and social justice: In defence of Aristotelian essentialism. Political Theory, 20: 202-246. AHMR, Vol.3 No1, January-April 2017 699 Osei-Assibey, E. 2013. Inequalities in Ghana: nature, causes, challenges and prospects. The heart of the Post-2015 development agenda and the future we want for all. Final draft report for the Post-2015 Development Agenda: Ghana Inequalities Group Team. Osili, U.O. 2004. Migrants and housing investments: Theory and evidence from Nigeria. Economic Development and Cultural Change, 52(4): 821-849. Pang, T., Lansang, M.A. & Haines, A. 2002. Brain drain and health professionals: A global problem needs global solutions. British Medical Journal, 324: 499-500. Parekh, B. 1996. Citizenship and political obligation. In: King, P. (Ed.). Socialism and the Common Good. London: Frank Cass, pp. 259-289. Putnam, R. 2000. Bowling Alone: The Collapse and Revival of American Community. New York, NY: Simon and Schuster. Rapoport, H. & Docquier, F. 2005. The Economics of Migrants’ Remittances. IZA Discussion Paper no. 1531. Rostow, W.W. 1960. The Stages of Economic Growth: A Non-Communist Manifesto. Cambridge: Cambridge University Press. Safran, W. 1991. Diasporas in modern societies: myths of homeland and return. Diaspora, 1: 83-99. Sassen, S. 2006. Territory, Authority Rights: From Medieval to Global Assemblages. Princeton NJ: Princeton University Press. Seers, D. 1969. The meaning of development, Institute of Development Studies Communications 44. Sen, A. 1999. Development as Freedom. New York: Anchor Books. Smith, R.C. 2003. Migrant membership as an instituted process: Transnationalization, the state and the extra-territorial conduct of Mexican politics. International Migration Review, 37(2): 297-343. Smith, L. & Mazzucato, V. 2009. Constructing homes, building relationships: Migrant investments in houses. Tijdschrift voor Economische en Sociale Geographie, 100(5): 662-673. Teye, J.K. 2012. Benefits, challenges and dynamism of positionalities associated with mixed methods research in developing countries: Evidence from Ghana. Journal of Mixed Methods Research 6(4), 379-391. AHMR, Vol.3 No1, January-April 2017 700 UNDESA. 2016. Monitoring global population trends. United Nations Department of Economic and Social Affairs, Population Division. UNDP. 2009. Overcoming barriers: Human mobility and development. Human Development Report 2009. New York: UNDP. Van der Geest, S. 1998. “Yebisa Wo Fie”: Growing old and building houses in the Akan culture of Ghana. Journal of Cross-Cultural Gerontology, 13: 333-359. Vertovec, S. 2006. Transnational Challenges to the 'New' Multiculturalism. University of Oxford: Oxford Centre on Migration, Policy and Society (COMPAS). Voigt-Graf, C. 2008. Migration, labour markets and development in PICs. Paper presented at the Conference on Pathways, Circuits and Crossroads in Wellington, June 9-11. ‘New Research on Population, Migration and Community Dynamics.’ Werbner, P. 2002. Imagined Diasporas among Manchester Muslims. Santa Fe, NM: SAR Press. Wong, M. 2013. Navigating return: the gendered geographies of skilled return migration to Ghana. Global Networks, 14 (4): 438-457. World Bank. 2016. Migration and Remittances Factbook 2016. 3rd Edition. World Bank Group. Endnotes * The Homecoming Summit in 2001; the Dual Citizenship Act [Act, 591, Republic of Ghana, 2002]; establishment of the Non- Resident Ghanaian Secretariat in 2004; the Ghana Investment Promotion Centre [promotion of Tourism] Instrument, 2005 [L.I. 1817]; the Representation of the Peoples [Amendment] Act [Act 699, Republic of Ghana, 2006] and the Investment Summit in 2007). ** This project was among members of the Ghanaian diaspora in the UK and heads of migrant households in Ghana, community leaders and Ghanaian political elites, the Bank of Ghana, Ghana Statistical Service and the Department of National Archives. The UK component of the bigger study comprised of interviews with 20 key informants, a survey of 120 Ghanaian migrants, participant-observation activities and library research. The Ghana AHMR, Vol.3 No1, January-April 2017 701 component involved a survey of 346 heads of migrant households, interviews with 20 key informants (community leaders and political elites), data from the Bank of Ghana, Ghana Statistical Service and the Department of National Archives. AHMR, Vol.3 No1, January-April 2017 702 Attitude, Risk Perception and Readiness of Ethiopian Potential Migrants and Returnees Towards Unsafe Migration Abebaw Minaye* and Waganesh A. Zeleke** Abstract In Ethiopia, where there is high prevalence of migration to the Middle East and Europe, a multitude of studies have focused on the relationships between the role of smugglers, push and pull factors, and illegal migration. However, only a fraction of studies have examined the context from the individual and collective mind-set perspective. The process of the decision to migrate may be influenced by individuals’ beliefs about illegal migration. This study examined the attitudes, levels of readiness and risk perceptions of potential and returnee migrants towards illegal migration to the Middle East. Survey data were collected from 1,726 Ethiopian returnee (n=991) and potential migrants (n=735). Results indicated that here was a significant difference between potential and returnee migrants in holding a positive attitude towards unsafe and illegal migration, t (1260) =-8.474, p=0.000. Potential migrants favour illegal migration more so than returnees. The level of risk perception of returnee migrants and the level of readiness for migration to the Middle East of both potential and returnee migrants was found to be below the expected mean score. Gender differences in the level of risk perception and readiness were also observed. Female participants tended to see the risks associated with unsafe migration less than male returnee migrants; female participants indicated a lower level of readiness than their male counterparts. Results imply a need to work on the level of behaviour change communication and to focus on attitude and practice change rather than mere awareness-raising. The results also imply the need to create actual jobs that can keep people from choosing unsafe migration. Keywords Ethiopia, migrants, attitude, risk-perception, readiness. *Assistant Professor, School of Psychology Addis Ababa University. **Assistant Professor, Department of Counselling, Psychology and Special Needs Education, Duquesne University. AHMR, Vol.3 No1, January-April 2017 703 Introduction Migration is a global occurrence, however, information on what shapes individuals’ decisions towards unsafe migration in developing countries such as Ethiopia is scant. Although it is well-documented that poverty, system failure (Abebaw & Waganesh, 2015; Adams, 2011; Dinbabo & Carciotto, 2015; Fernandez, 2013), actions of organised criminal groups like traffickers, smugglers and brokers (Fernandez, 2013; Friesender, 2007; Gozdziak & MacDonnel, 2007; UNODC, 2010; US Department of State, 2016), family and peer pressure (Abebaw, 2013; Biniam, 2012; Marina, 2016), and other push forces (Animaw, 2011) act as causes for the high prevalence of legal and illegal migration, a similar level of attention is needed to understand what shapes migrants’ decisions at the individual level. One evidence of the focus on organised crime groups is the billions of dollars invested by leading anti- trafficking nations or regions, such as the USA and Europe, in convicting traffickers and illegal migrants rather than supporting migrants. Internal dynamics within the individual migrant, like attitude and risk-perception, seem to have received little attention. This lack of understanding might undermine the agency of migrants in the decision-making process, which in turn has implications for our understanding of the migration decision and shifts the target of intervention to external forces rather than the migrants themselves (Gong et al., 2011; Sharma, 2005). The literature regarding migrants’ attitudes and risk-perceptions of migration to the Middle East and South Africa in particular, is dearth. The Middle East is becoming one of the biggest hubs for migrants (Saudi Arabia and the United Arab Emirates (UAE) alone host 17 million migrants) from South East Asia (mainly the Philippines, Sri Lanka, India and Bangladesh) and Eastern Africa and Horn of Africa (mainly Ethiopia, Somalia, Kenya, Uganda and Tanzania). Of the top ten countries with the highest migrant population relative to their ‘native’ population in the 2004 Human Development Report, the first four are Middle Eastern countries (GMDAC, 2015). For example, the UAE, Qatar, Kuwait and Bahrain have migrants that constitute more than 50% of their population (IOM, 2014). South Africa is also a top destination (2.2 million in a 2015 estimate) within Africa for migrants from different regions of Africa (West, East and South) (Horwood, 2009) and other parts of the world (mainly South East Asia, i.e.: India, China, Pakistan, Bangladesh and Thailand) (Statistics South Africa, 2014). AHMR, Vol.3 No1, January-April 2017 704 Factors related to the influx of migrants to the Middle East and South Africa should be investigated in more detail, especially because there are many safety concerns identified in various studies (Waganesh et al., 2015; Meskerem, 2011) and reports by international aid and human rights agencies such as Human Rights Watch, ILO and IOM. There is a dire need for empirical evidence to better understand what contributes to the decisions people make to migrate to the Middle East and South Africa despite the potentially unsafe situations in these destinations. Ethiopian Context Ethiopia is a source country for migration, smuggling and trafficking (Fernandez, 2013; Jureidini, 2011; UNICEF, 2005; US Department of State, 2016). According to the Ethiopian Ministry of Labour and Social Affairs (MOLSA, 2013), in 2011/12 alone, 198,667 Ethiopian women legally (their migration processed by the Ethiopian Ministry of Labour and Social Affairs) migrated to Saudi Arabia, the UAE and Kuwait for domestic work. On top of this, 60-70% of migrants migrate illegally, that is through smuggling or being trafficked (US Department of State, 2016). In 2011/12, the illegal migration was estimated at 300,000-350,000, making the total sum close to 550,000. The Regional Mixed Migration Studies (2014) reported the smuggling of 334,000 Ethiopians through Yemen alone from 2006-2014. The issue was given attention as a major social, economic, political and national security problem following the 2013 deportation of over 163,000 Ethiopian irregular migrants when the Kingdom of Saudi Arabia (KSA) cracked down on irregular migration (IOM, 2014). Studies that address the Ethiopian migration context emphasise factors such as push and pull forces (Blagbrough, 2008; Kangaspunta, 2007; Yoseph et al., 2006)and the plight of migrants, such as sexual abuse, physical abuse, salary denial, labour exploitation, confinement and denial of access to health services (Human Rights Watch, 2008; US State Department, 2016;Abebaw, 2013). The movement of populations has been associated with sexual health risks and increased transmission of HIV/AIDS (Bastide, 2015; Weine & Kashuba, 2012; Camlin, et al., 2010). There are also reports of the negative mental health consequences of unsafe migration on victims (Anbesse et al., 2009; Bhugra, 2004; Mirsky, 2009; Waganesh et al., 2015). AHMR, Vol.3 No1, January-April 2017 705 Unsafe and stressful migration affects migrants, their families and their governments in general. Consequences of unsafe migration can have an impact on people physically, psychologically, socially and economically. For example, a qualitative study involving female Ethiopian migrants who were employed in the Middle East found themes of exploitation, enforced cultural isolation, undermining of cultural identity and thwarted expectations (Anbesse et al., 2009). Some women in this study alluded to being sexually assaulted and dehumanised. Women shared their experiences; one noted: “They don’t see you as a human being.” Another said: “There were times when I would take food from the wastebasket and eat” (Anbesse et al., 2009: 560). The aforementioned themes revealed major concerns for the mental and physical health of Ethiopian migrants in the Middle East. Anbesse et al. (2009) remarked that migrants initially left Ethiopia hoping to work and improve their lives. But potential migrants may not have accurately evaluated the reality of potential risks related to migrating. In spite of this, how migrants perceive potential risks was not well studied, at least in the Ethiopian context. Another important issue related to better understanding the attitudes of migrants is the belief that some migrants espouse that their fate is in the hands of God, which may help them to overcome the fear and risks associated with migrating (Bastide, 2015). This is a critical point to understand about the attitudes of some migrants, because if they believe that they must endure their hardships as part of their predestined lives, their conceptualisation of a “risk” may be less relevant to them (Bastide, 2015). Paradoxically, some migrants may view leaving their birth country as a way of challenging their destiny; they do not necessarily see themselves as passive recipients of a predetermined life (Bastide, 2015). Though these beliefs may vary by individual, region and religion, these conceptualisations are important to consider when working to understand how potential migrants evaluate risks and how migrating may affect their lives. As tenants of social judgment theory (SJT) and cognitive behavioural theory (CBT) show, cognitive processes are related to subsequent behaviour. So, if migrants believe that their fate is already predetermined, they may not be concerned about the decisions they make and the safety implications of those decisions. Research on Sudanese youth indicated that their attitudes towards international migration were influenced by their families, friends and current circumstances in their home country, often related to unemployment (Yaseen, 2012). This study asserted that “the majority of youth had positive attitudes AHMR, Vol.3 No1, January-April 2017 706 towards migration” and believe it to be the “best solution to improve” their economic circumstances, specifically “unemployment and low incomes” (Yaseen, 2012: 108). It seems that unfavorable current circumstances may lead potential migrants to overlook the dangers of migrating. In most Ethiopian cases, the initial migration decision is made by the migrants’ own free will (Abebaw & Waganesh, 2015; Addis Ababa University, 2015; Sharma, 2005). They are often given misinformation regarding the positions and circumstances awaiting them in the host countries by brokers, agencies, smugglers and traffickers (Lansink, 2006; UN, 2000; Marina & Medareshaw, 2015). Misleading information by traffickers shapes the knowledge and beliefs of migrants. In spite of theoretical support regarding the role of an individual’s attitude in decision-making of any sort, studies on attitudes of migrants towards unsafe migration is missing. The only study which addresses attitudes of Ethiopian migrants is the Regional Mixed Migration Study of 2014, entitled “Blinded by Hope: Attitude, Knowledge and Practice of Ethiopian Migrants.” To reduce the impact of unsafe and illegal migration, we must improve attitudes and increase readiness of migrants for potential risks. This potential is substantiated by research that suggests that modifying attitudes also changes behaviours (Sheeran et al., 2016). Given the importance of individuals’ attitudes towards migration, it is necessary to examine factors that shape these attitudes. In this study, we proposed a conceptual framework for examining the factors that influence migrants’ and potential migrants’ attitudes. The framework includes both antecedents and consequences of migrant returnees and potential migrants’ attitudes. The antecedents can be grouped into personal factors (e.g. gender, age and religion), social factors (e.g. mass media, family influence and peer groups), previous exposure to migration, and other environmental factors (e.g. one’s level of education and economic independence). This model could help us understand how migrants’ and potential migrants’ attitudes towards migration are formed and maintained, which could assist in our efforts to shape those attitudes and prevent unsafe migration that may happen as a result of individuals’ attitudes and misguided perceptions of risks. Perceptions of risks are sometimes overshadowed by potential migrants’ hopes for personal change for the better and a positive future for their families. This cognitive minimisation of risks might increase positive attitudes towards unsafe migration. In order to positively influence the attitudes of potential AHMR, Vol.3 No1, January-April 2017 707 migrants, it is necessary to first understand the level of attitude and risk perception, as well as the variables that influence these attitudes and perceptions. Theoretical and Conceptual Frameworks Attitude is a key component in judgment and subsequent courses of action. Attitude can be understood as the internal psychological tendency one uses when evaluating an idea or potential action (Eagly & Chaiken, 1993). For instance, if a person is weighing whether to leave his or her birth country in hopes of obtaining better circumstances for his/her family, his/her attitude about that hope and the likely negative consequences associated with the migration (risk perception) are part of this individual’s processing of the decision alternatives. There are cognitive, affective and behavioural components to attitudes which relate to beliefs, feelings and actions, respectively (Eagly & Chaiken, 1993; Garrett et al., 2003). SJT offers a paradigm to better understand how people’s attitudes contribute to their mindset and how attitudes can change. In their theory of reasoned action, Aijen and Fishben (1980 cited in Southey, 2011) have indicated the link between attitude and behaviour and identified intermediaries that affect the strength of this link, like specificity and strength of the attitude. Stemming from Brunswik’s functionalist psychology (Doherty & Kurz, 1996), SJT posits that an individual’s attitudes are an integral part of the perception and evaluation of new ideas (Sherif & Hovland, 1980). Attitudes are part of a person’s identity (Sherif et al., 1965), which could be based on a person’s current state and past experiences, and they inform their decision-making process. Therefore, it is important to take into account the context within which a migrant returnee or potential migrant functions, and the systems that affect him/her, when trying to better infer his/her attitudes and decision- making processes. Another important piece of SJT is persuasiveness (Sherif & Sherif, 1968) and the study of the likelihood of someone to change his/her mind or attitude about something, which includes consideration of what contributes to that process. Specifically, ego-involvement, or how personally affected by or involved someone is in a particular issue, is highly relevant to how likely someone is to change his/her attitude (Sherif & Sherif, 1968). CBT also offers useful information when attempting to decipher how people come to their decisions to make changes. CBT refers to the interconnectedness of cognitive processes, emotions and behaviours. This theory is based upon a model of cognition that entails the aforementioned components. Beck’s (2006) AHMR, Vol.3 No1, January-April 2017 708 work in rational emotive behaviour therapy indicates that systematic bias may contribute to irrational thoughts. The application of CBT can be useful when attempting to better understand a person’s mindset. One can apply CBT to the migrant’s mindset and attitudes as part of cognitive processes as well as perceptions of risk. Given the interrelatedness of cognitions and behaviours, it is worthwhile to examine the attitudes of migrants to understand their decision-making processes to leave their birth countries. Attitudes create a potentially systematic bias that contributes to a migrant’s decision to leave their country and choice of where to go. Although attitudes might be strong enough in informing the decisions of people, individuals may not be conscious of their attitudes. Some migrants may not realise the inherent danger in some of the decisions they make. So it is worthwhile to better understand their attitudes and perceptions about migration and the risks associated with it. Taken together, SJT and CBT offer some models to aid in understanding how people come to make judgments about their lives; however, more information is needed about the migrants’ mindset specifically, so they can be informed about potentially unsafe decisions they are making. Sometimes because of having strong attitudes, migrants may not be fully aware of the conditions that await them after they migrate and the dangers associated with migrating to specific regions. This leads to an attitude not based on sound knowledge or on irrational belief. The present study examined the attitudes and risk perceptions of Ethiopian returnees and potential migrants towards migration to the Middle East and South Africa. It examined various beliefs of the study participants that influenced their decision-making processes to commit to or consider unsafe or illegal migration. It reviewed variations in belief systems based on participants’ demographic characteristics. Understanding the belief systems of Ethiopian returnees or potential migrants can promote culturally and contextually relevant outreach and treatment (Chang-Muy & Congress, 2009). Specifically, the study addressed four research questions regarding returnee and potential migrants’ belief systems towards migration: 1)What is the level of readiness, risk perception, and attitude of Ethiopian potential migrants and returnees towards migration to the Middle East and South Africa?2) Is there a significant relationship between readiness, risk perception and attitude among returnee and potential migrants? 3) Is there a significant difference in risk perception and attitude between returnees and potential migrants? And AHMR, Vol.3 No1, January-April 2017 709 4) What is the effect of readiness and risk perception on the attitudes of returnees and potential migrants towards migration? Method A. Site of the Study This study was conducted in six regions identified as hotspots for migration from Ethiopia to the Middle East and South Africa. The six regions consist of eight sites identified as hotspots for migration based on previous studies (Animaw, 2011; Abebaw, 2013), Ethiopian media reports and consultation with the Ethiopian Ministry of Labour and Social Affairs. The sites included two zones in Oromia region, two zones from the Southern Nations and Nationalities region, one zone in Amhara region, one zone in Tigray, and Addis Ababa and Dire Dawa city administrations. B. Participants Using a purposive sampling technique (convenient and snowball), study participants were individuals who either returned from the Middle East and South Africa or those who were preparing to migrate to the Middle East or South Africa illegally. Data was collected from February-April2014. Of the 1,771 participants, 60.5% (n=1036) were returnee migrants from the Middle East or South Africa and 39.5% (n=735) were potential migrants who planned to illegally migrate to the Middle East or South Africa. Both returnees and potential migrants completed a survey questionnaire having demographic, readiness, attitude and risk perception items/scales. C. Instrumentation At the time of this writing, there were no standardised or empirically validated scales that measure the constructs proposed in this study (readiness, attitude and risk perception of migrants to unsafe migration). Also, there were no instruments available in Amharic, the primary language of Ethiopia. Consequently, this survey was developed by a group of psychology and social work faculty members at Addis Ababa University. Based on a comprehensive review of the risks and challenges associated with migration and a review of the literature on Ethiopian and African migration, the researchers developed the survey instrument. The instrument was organised into five sections: (a) AHMR, Vol.3 No1, January-April 2017 710 demographic data, (b) antecedents, (c) readiness scale, (d) perception of risk scale and (e) attitude scale. Demographic questions included sex, age, employment status, educational level, religion, region of residency and migration status of respondents. The antecedents section asked information about the role of social context like family, brokers, agencies, friends, relatives and mass media. It also asked about personal reasons and motivations that shaped decision-making. D. Readiness Measure This was a five items scale, which measured participants’ levels of readiness to work in the cultural context and understandings of the nature of the job circumstances in the destination countries. Items included: How aware are/were you about the nature of the job in the destination? How familiar are/were you about the culture of the destination country? How aware are/were you about your rights and responsibilities in the destination country? How confident are/were you in your coping skills if you face challenges? How much do/did you know about the communication style in the Middle East/South Africa? Scores were computed by taking the sum of the 5 items (each measured on a 1 to 4 Likert scale), with 1 being “totally unaware” and 4 being “very aware.” Scores for this variable ranged from 5 to 20. E. Risks Perception Scale The Perception of Risk Scale was a three point Likert scale which measured participants’ perceptions of the risks involved in migrating to the Middle East and South Africa. The items included: Level of fear of being cheated by broker or agency or employer; Level of concern about facing difficult circumstances in the destination country; Level of determination to go to the Middle East or South Africa even through illegal channels if the legal process doesn’t work; and Level of concern about safety while traveling through the illegal migration route. The scale score is computed by taking the sum of the 4 items measuring the knowledge of risks, each measured on a 1 to 3 scale with 1 being ‘I know much,’ 2 being ‘I know some’ and 3 being ‘I don’t know.’ The scores for this variable range from 3 to 12, and the scores were reverse-coded so that high scores reflected more knowledge and low scores reflected little knowledge regarding risks. AHMR, Vol.3 No1, January-April 2017 711 F. Attitude Scale The attitude scale was a six item scale developed to measure participants’ attitudes towards unsafe or illegal migration. The scale consisted of items such as: Life is predetermined, not affected by whether you migrate or not; I prefer working low status jobs overseas than in Ethiopia; I don’t believe working in Ethiopia changes my own or my family’s life for the better; I believe media reports on the problems of illegal migration are exaggerated; I prefer to work overseas with no dignity than living under poverty in Ethiopia; and I believe that Ethiopian youth can change their own life and their family’s lives only by working overseas. The scale score was computed by taking the sum of the 6 items, each measured on a 1 to 4 Likert scale, with 1 being ‘strongly disagree’ and 4 being ‘strongly agree. ’The scale score ranged from 6 to 24. Since the items were negatively-worded, the scores were reverse-coded so that low scores reflected negative attitudes and high scores reflected positive attitudes towards unsafe migration to Saudi Arabia and South Africa. The measures were developed and administered in Amharic. Internal consistency estimates (Cronbach’s Alpha) for each subscale were computed to confirm the reliability of the subscales used in this study. The reliability estimates for all the subscales were as follows: attitude (a=. 79), risk perception (a=.54) and readiness (a=.72). The reliabilities of the measures were, therefore, quite satisfactory (Panayides, 2013) and indicate consistency. G. Data Collection Procedures With approval from Addis Ababa University’s Office of the Vice President for Research and Technology Transfer, 21 data collectors (three in each of the six zones and three more for Addis Ababa and Dire Dawa) and 7 supervisors (one for each of the six zones and one for Addis Ababa and Dire Dawa) were hired and given a day-long training in the data collection process. The data collectors included graduate students and faculty members at higher institutes in Ethiopia from the fields of Psychology, Social Work, Law, Economics and Health. The supervisors were faculty members from Addis Ababa University. They were assigned to assist and monitor the data collection process. The data collectors explained the purpose of the study to the study participants before administering the questionnaire. Participation in the study was voluntary. Returnees and potential migrants who chose to participate completed a paper- AHMR, Vol.3 No1, January-April 2017 712 and-pencil questionnaire. Those who could not read were assisted by the data collectors in completing the questionnaire. H. Data Analysis Descriptive statistics were conducted to report participants’ demographic characteristics and the associated factors involved in participants’ decision- making about migration. A factorial MANOVA was conducted to evaluate the differences due to Migration Experience, Gender, Religion and Education on the three dependent variables of readiness to work in destination context, attitude towards migration and perception of risk. Pre-screening of data was conducted to identify outliers and missing data, and to collapse the response categories of the independent variables with small sample sizes in cases where it made conceptual sense. For example, for the variable Education, the categories of ‘Certificate/diploma’ and ‘First degree and above’ in the original data were combined into one category, ‘Certificate/diploma and above.’ I. Variables The main independent variables in this study were demographic variables (see Table 2); migration experiences (returnee vs. potential), gender, religion and educational level. The dependent variables were readiness to adjust to the destination context, attitude towards migration and perception of risks (see Table 1). Table 1. Dependent variables. Minimum Maximum Expected mean Readiness to adjust to destination context 5.00 20.00 12.50 Attitude towards migration 6.00 24.00 15.00 AHMR, Vol.3 No1, January-April 2017 713 Perception of risks 4.00 12.00 8.00 Research Findings and Analysis Demographic Characteristics of Participants The sample consisted of 1,726individuals of which 991 were returnee migrants from the Middle East and South Africa and 735 were potential migrants who were considering migrating to the Middle East or South Africa. A demographic profile of the sample is presented in Table 2. Returnee Migrants The returnees were between the ages of 16 and 60 years (M=27.55; SD 6.34). Approximately 55% (n=545) were female and 44.9 % (n=446) were male. Over 82.96% (n=789) returned from the Middle East, while 17.04% (n=162) returned from South Africa. Over half (51.2%) of the returnees resided in an urban setting (n=518) and 48.8% were from rural parts of Ethiopia (n=498).In terms of participants’ education, 6.4% (n=66) did not have any formal education, 8% (n=88) could write and read, 8.8%(n=90) attained grade 1-4 levels of education, 25.9% (n=268) had grade 5-8 levels of education, 30.8%(n=316) completed grade 9-10, 12.9% (n=132) completed grade 11-12 and7.3% (n=74) had a certificate, diploma or above. The prior migration job profile of returnees indicated that 49% (n=464) did not have a specific job other than supporting their families. Over half 51% (n=479) reported that they were employed as domestic workers, daily labourers, government employees or other types of low-class jobs. Potential Migrants The average age of potential migrant participants was 25.3 years. Approximately, 57% were female (n=417) and 43.3% were male (n=325). Over half (53.9%) of the participants resided in urban settings (n=398) and 46.1% were from rural parts of Ethiopia (n=340). In terms of education, 5% (n=37) did not have any formal education, 5.4% (n=40) said they could read and write, 7.9%(n=59) attained a grade 1-4 level of education, 35.9% (n=268) had a grade 9-10 level of education, 27.7%(n=206) completed a grade 5-8 level of education, 9.8% (n=73) completed a grade 11-12 level of education and the AHMR, Vol.3 No1, January-April 2017 714 8.5% (n=63) had a certificate/diploma or above. In terms of employment, 63.6% (n=478) did not have a job at the time they planned to migrate, whereas 31.4% (n=236) had a job. Table 2. Demographic characteristics of participants. Returnee migrants (n=991) Potential migrants (n=735) n % n % Sex Male 446 45.00 325 43 Female 545 55.00 417 57 Religion Orthodox 301 29.60 244 32 Muslim 512 50.40 412 55 Protestant 175 17.20 81 11 Catholic 18 1.80 12 1.6 Education No education 66 6.40 37 5.0 Reading and writing 88 8.00 40 5.40 Grade 1 to 4 90 8.80 59 8.00 Grade 5 to 8 268 25.90 268 36.00 Grade 9 to 10 316 30.80 206 28.00 Grade 11 to 12 132 12.90 73 10.00 Certificate/diploma or above 74 7.40 64 8.50 Residence Urban 518 52.20 395 53.70 Rural 498 48.80 340 46.30 Marital status Married 449 44.80 226 32.00 Single 507 50.00 457 64.70 Divorced 43 4.30 23 3.30 Employment status Employed 479 51.00 236 31.40 Not-employed 464 49.00 478 63.60 Destination Middle East 789 82.96 639 86.90 South Africa 162 17.03 92 13.10 AHMR, Vol.3 No1, January-April 2017 715 *Please note, due to missing data, sample size for some variables may not add up to 1036 for returnees, or 735 for potential migrants. Associated Factors that Lead to Migration or the Decision to Migrate The motivation for migration for 64.4% of returnees (n=646) and 66.4% of potential migrants (n=499) came from themselves, while 20% of returnees (n=201) and 13.4% of potential migrants (n=104) reported that their friends encouraged them to migrate. About 13.8% (n=138) of returnees and 14.8% (n=111) of potential migrants reported that families/relatives initiated the decision to migrate. Less than 1.8% (n=20) of returnees and 5.4% (n=39) of potential migrants reported the initiative came from brokers and agencies. Concerning source of information, 42.1% of returnees reported that they got the information on how to go to the Middle East from brokers, 39.1% got it from friends, 14.3%indicated family as their source of information and 4.1% of returnees reported they received the information from the media. Related to the major reasons for migrating, almost half, 49.7%, of returnees and 45.8% of potential migrants indicated poverty, while 38.5% of returnees and 31.5% of potential migrants indicated a lack of opportunity for employment in Ethiopia and a desire to be economically independent. The data on source of information, reasons for migration and motivation for migration is presented in Table 3. Table 3. Frequency and percentage of source of information, reasons and motivation for migration. Returnee Migrants Potential Migrants n % n % Source of information Broker 423 42.10 291 40.81 Family 144 14.30 107 15.00 Friend 397 39.50 286 40.12 Media 41 4.10 29 4.07 Reason to migrate Poverty 502 49.70 344 45.80 Others’ Influence 110 10.90 63 8.40 Economic Influence 389 38.50 274 31.50 AHMR, Vol.3 No1, January-April 2017 716 Who initiated migration Self 46 64.40 499 66.40 Friends 201 20.00 104 13.80 Family/Relatives 138 13.80 111 14.80 Others 18 1.80 19 2.40 Risk Perception of Unsafe Migration The results show that the returnee migrants’ score (M=10.307, SD=1.81) was lower than the expected average (M=12.5). These results suggest that returnee migrants assumed lesser risk related to unsafe migration. The independent t- test indicated that there was a significant difference between male (M=10.57, SD=1.75) and female returnees (M=10.0984, SD=1.842) on their levels of risk perception, t (649) =3.43, p=0.001. This suggests that female returnee migrants tend to acknowledge the risks associated with unsafe migration less than male returnees. Participants’ Readiness Both returnee (M=6.13) and potential migrants’ (M=7.04) scores were below the average (8.00) on the readiness scale. An independent sample t-test was conducted to compare the difference between returnee (M=6.1304, SD=1.80) and potential migrants (M=7.04, SD=1.80) in terms of their levels of readiness, t (1276) =-8.93, p=0.23. These results suggest that both returnee and potential migrants’ levels of readiness to migrate to the Middle East and South Africa are less than the expected mean. The independent t-test result revealed that there was a significant difference between male (M=6.87, DS=1.97) and female participants (M=6.37, SD=1.75) with regards to level of readiness, t (1263) =4.59, p=0.04. These results suggest that females were, or are, less prepared for illegal migration than their male counterparts. Table 4. Independent sample t-test: Readiness. Returnee Migrants Potential Migrants T Df Readiness to Migrate M 6.130 7.033 -8.930 1276 SD 1.806 1.801 Gender AHMR, Vol.3 No1, January-April 2017 717 Note * = p <0.05 Attitude towards Migration As Table 5 shows, 49% of the returnees believe that life is pre-determined, showing an undermining tendency of vulnerability. Over 60% prefer working a low-paying job abroad than doing the same job in Ethiopia, indicating a favorable attitude towards migration. Close to 55% of the returnees do not believe that working in Ethiopia will help to change their own and their family’s life for the better, rather 65% of the returnees believe that Ethiopian youth can change their life for the better by working abroad. Over 53% of the returnees believe that reports by the media, specifically the Ethiopian Television, about the problems of illegal migration are exaggerated. All of this shows that almost half of the migrants had a positive attitude towards migration of any sort (legal or illegal). The general attitude scale result shows that 540 (57.1%) of the returnees have a positive attitude toward migration in any form. This reveals that despite their plight and expulsion by force and after facing several problems in the destination country, more than half of these returnees are in favor of migration. They added that if they got the chance they would be willing to re- migrate even if it was risky (with the likelihood of being trafficked). The one sample statistic for returnees shows that there is no significant difference between the expected and obtained means, indicating that returnees are not significantly in favour of or against the expected unsafe migration despite what they have experienced. Female Male M 6.374 6.869 4.594* 1263 SD 1.759 1.972 AHMR, Vol.3 No1, January-April 2017 718 Table5. Frequency and percentage, n (%), of attitude of returnees towards migration. The independent t-test results indicated a significant difference between returnee migrants (M=15.318, SD=5.15) and potential migrants (M=17.39, Attitudes Strongly Disagree Moderately Disagree Moderately Agree Strongly Agree Life is predetermined, not affected by whether or not you migrate. 368(36.5) 146(14.5) 184(18.3) 309(30.7) I prefer working low status jobs overseas than in Ethiopia. 252(24.9) 164(16.2) 251(24.8) 346(34.2) I don’t believe working in Ethiopia improves my own or my family’s life. 292 (29.1) 164(16.3) 246(24.5) 303(30.1) I believe that Ethiopian Television reports on the problems of illegal migration are exaggerated. 336(33.7) 132(13.2) 282(28.3) 247(24.8) I prefer to work overseas with no dignity than living under poverty in Ethiopia. 367 (36.7) 182 (18.2) 174(17.4) 278 (27.8) I believe that Ethiopian youth can change their own and their families’ lives only by working oversees. 188 (18.7) 141(14.0) 242 (24.1) 435( 43.2) AHMR, Vol.3 No1, January-April 2017 719 SD=4.65) on the level of attitude towards illegal or unsafe migration, t (1260) =-8.47, p=0.000. These results suggest that potential migrants have a more favourable attitude towards unsafe migration than returnee migrants. The difference between female (M=16.21, SD=4.94) and male (M=16.124, SD=5.202) attitudes towards migration, t (1511) =-0.332, p=0.74, was found to be insignificant. Table 6. Independent sample t-test: Readiness. Returnee Migrants Potential Migrants T Df Attitude towards migration M 15.3083 17.3929 - 8.474* 1260 SD 5.152 4.652 Gender Female Male M 16.208 16.124 -0.332 1511 SD 4.943 5.202 Note * = p<0.01 Relationship Between Participants’ Ages and the Dependent Variables Age has a high negative correlation with attitude, showing that younger participants have more positive attitudes towards migration. Readiness has a strong positive correlation to both risk perception and attitude (p < 0.01), showing that better readiness is associated with higher risk perception but positive attitude to migration. AHMR, Vol.3 No1, January-April 2017 720 Table 7: Correlations among study variables. Variables Age Readiness Attitude Risk Perception Age of the respondents 1 0.05 -0.71* Readiness 1 0.26** 0.19** Attitude 1 -0.07 Risk perception 1 Analysis of Variance: Education, Attitude and Readiness Analysis of variance (ANOVA) yielded significant differences between levels of education among participants and readiness, F(7,591)=4.62,p<0.001; level of education and attitude towards migration (F(7,692)=2.10, p<0.05); gender and awareness (F(3,1260)=7.58, p<0.000); and gender and attitude (F(3,1260)=3.40, p<0.001. These results suggest that participants’ with a higher level of education seem to have better readiness (M=8.57, SD=2.38) and less favorable attitudes (M=15.79, SD=0.492) towards migration to the Middle East and South Africa than participants with less education (M=6.03, SD=1.47) and lower levels of readiness (M=17.71, SD=5.055). AHMR, Vol.3 No1, January-April 2017 721 Table 8. Test of ANOVA in readiness and attitude based on educational background. Variables Groups Sum of Squares Df Mean Square F Sig. Readiness Between Groups 134.37 3 19.20 7.58 .000 Within Groups 4252.40 1260 3.36 Total 4386.78 1272 Attitude toward migration Between Groups 601.35 3 85.91 3.40 .001 Within Groups 41132.386 1260 25.25 Total 41733.740 1636 Discussion Overall, the current study indicated that both potential and returnee migrants have a positive attitude towards migration to the Middle East and South Africa. On the other hand, their levels of readiness and risk perception towards unsafe migration were found to be below the expected average. These findings are of particularly great concern because they indicate the lack of informed decision- making by participants about illegal or unsafe migration. More than the action of organised groups or the push forces, returnees’ and potential migrants’ decisions to migrate are a product of their attitudes and risk perception about migration. The results further indicated that, in general, the experience of migrating may influence risk perceptions and attitudes about migration, as returnees saw themselves as less prepared to re-migrate than potential migrants. The potential migrants had more favourable attitudes about migration than the returnees. This could be because the returnees had AHMR, Vol.3 No1, January-April 2017 722 witnessed the challenges of unsafe migration and were more likely to have an objective evaluation of the risks, whereas potential migrants could have been more optimistic as they had not directly experienced the challenges. In terms of readiness, the potential migrants and returnees significantly differed from each other. The potential migrants reported having greater readiness. This could also be associated with returnees’ emic experience of what is required during harsh travel and working in the destination. They knew their gaps better. They may have struggled to adjust to the work and social environment in the destination and in transit. Whereas potential migrants might have been guided by their optimistic bias about what was actually demanded. Many Ethiopian domestic workers suffer in the Middle East, partly because of their poor preparation for the work and lifestyle (Waganesh, 2015). They assume that domestic work is similar in Ethiopia and in the Middle East. Returnees perceiving themselves as less prepared for migration than potential migrants may speak with the clarity of hindsight about the reality of migration. Returnees understand that they did not have much readiness for what the destination cultural, language and work-nature required. Supporting this inference, it was found that the participants’ experiences related to migration had an effect on the level of his or her perception of risk and attitudes toward migration. The returnees indicated a higher level of risk perception and a less positive attitude towards migrating compared to potential migrants. This further indicates that the experiences of migrating have some influence on subsequent views on migration and possibly on willingness to migrate again as well as where to migrate. This, in particular, is in line with previous findings related to the potentially dehumanising and unfavorable situations Ethiopian migrants found themselves in when they migrated to the Middle East (Anbesse et al., 2009). The results also suggest an association between socio-demographic characteristics such as age, sex and educational level, the perception that illegal migration is low risk, and an individual’s favorable attitude towards migration. This provides some support for the argument that public education and knowledge-building may be effective strategies in addressing illegal migration issues. It is not surprising that there was a direct positive association between education level and a relatively better readiness to migrate. Education level is inversely associated with a favorable attitude to migration which makes sense in that the better the understanding level (education), the lesser an individual may favour unsafe migration. In terms of AHMR, Vol.3 No1, January-April 2017 723 sex, female participants were less ready to migrate. Additionally, readiness and risk perception were significantly related such that increased readiness of an individual to migrate is associated with increased perceived risk related to unsafe migration. Notably, attitude about migration was negatively correlated with readiness, age and risk perception, seeming to further indicate that as one gains experience, by aging or otherwise, his or her attitudes about migration seem to become less favorable. Consistent with and building upon previous research on youth having positive attitudes toward international migration (Yaseen, 2012), the current study found that younger individuals had more positive attitudes about migration than older individuals, possibly indicating that they would rather work low-paying jobs outside of Ethiopia, even if their dignity was compromised, in pursuit of a better life. Additionally, perhaps the older migrants have gained more experience with the hardships associated with migrating, which may account for the more negative attitudes of the older participants in this study. On top of this, there are more opportunities for work for young people as the demand is higher for young labour. Previous studies have shown that trafficking and smuggling victims are largely children and women because of the growth of the service and sex industry (Lutz, 2011) and because of their hard work, compliance and submissiveness, as well as the assumption that they can be easily misguided about opportunities (IOM, 2006). The more favourable the attitude towards migration, the less perceived risk is associated with migrating. What is arguably one of the most important points of these results is the finding that many returnees still have a favourable view of migration, even if their experiences were negative. Some indicated that they would re-migrate, even knowing that they could be trafficked. They were still in favour of migrating. What is notable about these findings is that even knowing the hardships associated with migration, many Ethiopians are willing to consciously put themselves in potentially dangerous situations in order to leave Ethiopia. Such preferences may speak to broader concerns related to conditions in Ethiopia. Looking at the individual’s context and work culture in Ethiopia may be of value when attempting to understand these results. One example is the tradition of ascribing a ‘brave son or daughter’ to youngsters who risk their lives for the well-being of their family. A second example is the long tradition of viewing offspring as property in which children are expected to work beginning from around age six to contribute to the family’s economic betterment, working within the household or in other households for a salary that goes to the family. Referencing back to SJT and CBT, perhaps the returnees AHMR, Vol.3 No1, January-April 2017 724 are of the mindset that they would rather work abroad, even under similar poverty-stricken circumstances, than remain in Ethiopia. Thus, their behaviour could be in line with their surmised attitude that migration will not change their destiny, and it is better to leave Ethiopia than to stay. Another possible explanation is the one given by Marina and Medareshaw (2015), who propose that returnees may regret being deported before tasting the good side of migration and hence desire to try again. Another seemingly contradictory finding is the higher the readiness, the less individuals favour migrating. This may be partly related to the relationship between readiness and education. Migrants who have readiness and an understanding of the work and lifestyle of an illegal migrant may be more concerned about the possibility of abuse and, hence, they may be cautious in their attitudes. Although there is no specific study on attitude, risk perception and readiness of migrants, the current study is consistent with one previous study by the Regional Mixed Migration Secretariat (2014) that documented Ethiopian migrants being blinded by hope. In line with previous research, potential migrants often consider leaving their birth countries due to economic reasons, and they hope to have increased opportunities elsewhere (Anbesse et al., 2009). The present study has some new information about the perspectives of migrants and potential migrants. A particularly salient point from the current study is that almost half of the participants believe that life is predetermined and that migrating will not ultimately affect their lives or destinies; this particular finding is consistent with some previous literature regarding some migrants’ beliefs of God as being in control (Bastide, 2015). This mindset may lead some potential migrants to overlook risks and vulnerabilities associated with migrating, relating to the aforementioned systematic bias that can be associated with cognitive theory or CBT. When we reference systematic bias, we are not indicating that this “bias” is inherently a “bad” thing, rather, we are pointing out that a systematic bias, or basis from which they operate, namely believing their lives to be predetermined, likely permeates all areas of their lives and should be taken into account. Conclusions The current study examined attitudes, risk perception and readiness to migrate to the Middle East or South Africa from Ethiopia. The data was gathered from Ethiopians who were either returnee migrants (n=1036) or considering migration to the Middle East and South Africa (n=735). To our AHMR, Vol.3 No1, January-April 2017 725 knowledge, this is the first time that attitudes toward illegal migration to the Middle East have been studied in the Ethiopian context. The main finding here is that the majority of respondents believe that migrating to the Middle East by any means, including illegal migration, is better than living in Ethiopia as there is little chance for a better life in Ethiopia. The majority perceive the risks associated with illegally migrating to the Middle East and South Africa to be very low. The level of readiness, such as being aware of what kinds of job they could do or having some knowledge and experience of the language and lifestyle of the destination country, was found to be low. The study also explored other influences that potentially shape attitudes toward migration. This included characteristics such as age, gender and educational level. The results suggest that a mix of different influences shape the perception of migration among Ethiopian returnees and potential migrants. Migrants hold the key in the beginning of the migration decision in the form of initiating the migration, being driven by lowered risk perception and not being ready as a result of lower risk perception/optimistic bias. Of course, their risk perception and more positive attitude may be shaped by external forces like family, peers and brokers. Factors such as desire for economic independence, hearing of successes, desiring to support one’s family and the lack of job opportunities shape their attitudes. Thus, a more practical action to reduce unsafe migration would be working on risk perception, readiness and attitudes of migrants. But there is also a need to work on what leads to the positive attitude and lowered risk perception, which may include issues of socialisation agents (family, peers, neighborhood/community, societal culture and media) and push forces for migration like the evaluation of opportunities in one’s own country and in a foreign country. Therefore, future studies may further investigate the factors that shape migrants’ attitudes. Hence, by first focusing on the inward, we move to the outward. Looking inward helps researchers to understand the dynamics of migration decisions. For example, there is evidence that poverty may not necessarily drive the decision for unsafe migration. At times we may also wonder why people migrate under the likelihood of trafficking. Part of the explanation is related to the attitudes and risk perceptions of migrants. So risk perception and attitudes are vital for judgment and decision. There is also a need to disentangle the mindset of migrants. AHMR, Vol.3 No1, January-April 2017 726 A first key implication is that actors in preventing unsafe migration in Ethiopia have to focus on the attitude/mindset of migrants to bring meaningful change. Second, as findings suggest, some Ethiopians are willing to put themselves in harm’s way. So, working to change their attitudes that contribute to these dangerous decisions is important. However, it is also imperative to respect a person’s culture while attempting to raise awareness. For example, if a person holds the belief that their life is in God’s hands and is predestined, it is important to acknowledge that paradigm from which they operate and not dismiss it. Carl Rogers’ (1957; 1959) person-centered approach may be of value as it acknowledges that in order for change to occur, there needs to be a mismatch between a person’s experience and awareness, and that person needs to feel to to be understood. Translating this idea to the context of migration, to raise awareness about the potential dangers of migration, we may need to point out the differences between a person’s experience and their attitudes while making sure we acknowledge and validate the culture and context they function within. Finally, more studies are needed to examine both what shapes Ethiopians’ attitudes toward illegal migration and how their attitudes also (re)shape the decisions they make about migration. References Abebaw, M. 2013. Experiences of Trafficking Returnee Domestic Workers from the Gulf States: Implications for Policy and Intervention.PhD Dissertation, Addis Abeba: School of Social Work, Addis Ababa University. Abebaw, M. & Waganesh, Z. 2015. Re-conceptualizing human trafficking: The experiences of Ethiopian returnee migrants. Journal of Trafficking Organized Crime and Security, 1(1):9-23. Adams, C. 2011.Re-trafficked victims: How a human rights approach can stop the cycle of re-victimization of sex trafficking victims. The George Washington International Law Review, 43: 201-234. Addis Ababa University. 2015. Managing the socio-cultural, health, legal and economic dimensions of migration: Emphasis on Ethiopian migrants in the Middle East and South Africa. Abebaw, M. & Abera, T. (Eds.). Unpublished thematic research report, Addis Ababa. AHMR, Vol.3 No1, January-April 2017 727 Anbesse, B., Hanlon, C., Alem, A., Packer, S. & Whitley, R. 2009. Migration and mental health: A study of low-income Ethiopian women working in Middle Eastern countries. International Journal of Social Psychiatry, 55(6): 557-568. Animaw, A. 2011. Trafficking in Persons Overseas for Labor Purposes: The Case of Ethiopian Domestic Workers, Addis Ababa: ILO Country Office. Bastide, L. 2015. Faith and uncertainty: Migrants’ journeys between Indonesia, Malaysia and Singapore. Health, Risk & Society, 17(3-4): 226-245. Beck, A.T. 2006. How an anomalous finding led to a new system of psychotherapy. National Medicine, 12: 1139-1141. Bhugra, D. 2004. Migration, distress and cultural identity. British Medical Bulletin, 69:129-141. Biniam, T. 2012. Ethiopian Maids in a Severe Arab World. (retrieved 16 September 2012). Blagbrough, J. 2008. Child domestic labor: A modern form of slavery. Children in Society, 22: 179-190. Camlin, C.S., Hosegood, V., Newell, M., McGrath, N., Barnighausen, T. & Snow, R. 2010. Gender, migration and HIV in rural KwaZulu-Natal, South Africa. PLos One, 5(7): 115-139. Chang-Muy, F. & Congress, E.P. 2009. Social Work with Immigrants and Refugees: Legal Issues, Clinical Skills, and Advocacy. New York: Springer. Dinbabo, M. & Carciotto, S. (2015). International migration in Sub-Saharan Africa (SSA): A call for a global research agenda. African Human Mobility Review (AHMR). 1(2): 154-177. Doherty, M.E. & Kurz, E.M. 1996. Social judgment theory. Thinking & Reasoning, 2: 109-140. Eagly, A.H. & Chaiken, S. 1993. The Psychology of Attitudes. London: Harcourt Brace College Publishers. Fernandez, B. 2013.Traffickers, brokers, employment agents, and social networks: The regulation of intermediaries in the migration of Ethiopian domestic workers to the Middle East. International Migration Review, 47(4): 814-843. AHMR, Vol.3 No1, January-April 2017 728 Friesender, C. 2007. Pathologies of security governance: Efforts against human trafficking in Europe. Security Dialogue, 38(3): 379-402. Garrett, P., Coupland, N. & Williams, A. 2003. Investigating Language Attitudes: Social Meanings of Dialect, Ethnicity and Performance. Cardiff: University of Wales Press. Global Migration Data Analysis Center. 2015. Global Migration Trends Factsheet 2015. Berlin: GMDAC. Gong, F., Xu, J., Fujishiro, K., Takeuchi, T.D. 2011. A life course perspective on migration and mental health among Asian Immigrants: The role of human agency. Social Science & Medicine 73: 1618-1626. Gozdziak E.M. & MacDonnel, M. 2007. Closing the gaps: The need to improve identification and services to child victims of trafficking. Human Organization, 66(2):171-184. Horwood, C. 2009. In Pursuit of the Southern Dream: Victims of Necessity, Assessment of the Irregular Movement of Men from East Africa and the Horn to South Africa. Geneva, Switzerland: IOM. Human Rights Watch. 2008. Sinai Perils Risks to Migrants, Refugees, and Asylum Seekers in Egypt and Israel. New York: Human Rights Watch. International Organization for Migration. (IOM). 2006. National Counter- Trafficking Capacity Building Training. Manual for Prosecutors. IOM Special Liaison Mission in Ethiopia. International Organization for Migration (IOM). 2014. Global Migration Trends: An Overview, Geneva: IOM Migration Research Division. Jureidini, R. 2011. Mixed Migration Flows: Somali and Ethiopian Migration to Yemen and Turkey. Final report prepared for the Mixed Migration Task Force, Center for Migration and Refugee Studies, Cairo: American University. Kangaspunta, J. 2007. Collecting data on human trafficking: Availability, reliability, and comparability of trafficking data. In: Savona, E.U. & Stefanizi, S. (Eds.). Measuring Human Trafficking: Complexities and Pitfalls.Italy, International Scientific and Professional Advisory Council: Springer Science, Business Media LLC, pp.27-36. AHMR, Vol.3 No1, January-April 2017 729 Kapiszewski, A. 2006. Arab versus Asian migrant workers in the GCC.UN expert group meeting on international migration and development in the Arab Region, UN-DESA, Beirut, May 15-17 2006. Lansink, A. 2006. Human rights focus on trafficked women: An international law and feminist perspective. Agenda, Empowering Women for Gender Equity, 20(70):45-56. Lutz, H. 2011. The new maids: Transnational women and the care economy. Zed Books, London Marina, D.R. 2016. Time to look at girls: Adolescent girls’ migration in Ethiopia. Research report. Terre Des Hommes, International Federation. Marina D.R. & Medareshaw, T. 2015: Deported before experiencing the good sides of migration: Ethiopians returning from Saudi Arabia. African and Black Diaspora: An International Journal, DOI: 10.1080/17528631.2015.1083178. Meskerem, M.2011. Psychosocial and Economic Experiences of Gulf States Returnee Ethiopian Women Domestic Workers. MA Thesis, Addis Ababa: Institute of Gender Studies, Addis Ababa University. Mirsky, J. 2009. Mental health implications of migration: A review of mental health community studies on Russian-speaking immigrants in Israel. Social Psychiatry and Psychiatric Epidemiology, 44: 179-187. MOLSA. 2013, April. Labor market information bulletin. Addis Ababa: Employment Service promotion Directorate, Federal Democratic Republic of Ethiopia Ministry of Labor and Social Affairs, 73(8): 1169-1177. Panayides, P. 2013. Coefficient alpha: Interpret with caution. Europe’s Journal of Psychology, 9(4): 687-696. Regional Mixed Migration Secretariat. 2014. Blinded by hopes: Attitude, knowledge and practice of Ethiopian migrants. Mixed migration research series, Kenya, Nairobi, Westlands. Regional Mixed Migration Studies. 2014. The letter of the law: Regular and irregular migration in Saudi Arabia in a context of rapid change. Mixed migration research series, explaining people on the move, Nairobi, Kenya. Rogers, C. 1957. The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology, 21(2): 95-103. AHMR, Vol.3 No1, January-April 2017 730 Rogers, C. 1959. A theory of therapy, personality and interpersonal relationships as developed in the client-centered framework. In: Koch, S. (Ed.).Psychology: A Study of a Science. Vol. 3 Formulations of the Person and the Social Context. New York: McGraw Hill. Sharma, N. 2005. Anti-trafficking rhetoric and the making of a global apartheid. NWSA Journal of States of Insecurity and the Gendered Politics of Fear, 17(3): 88-111. Sheeran, P., Montanaro, E., Bryan, A., Miles, E., Maki, A., Avishai-Yitshak, A., Klein, W.M.P., & Rothman, A.J. 2016. The impact of changing attitudes, norms, and self-efficacy on health-related intentions and behavior: A meta-analysis. Health Psychology, 35(11): 1178-1188. Sherif, M. & Hovland, C.I. 1980. Social Judgment: Assimilation and Contrast Effects in Communication and Attitude Change. Westport: Greenwood. Sherif, M.S. & Sherif C.W. 1968. Attitude as the individuals’ own categories: The social judgment-involvement approach to attitude and attitude change. In: Sherif, M. & Sherif,C.W. (Eds.).Attitude, Ego-involvement, and Change. New York, NY: John Wiley & Sons, pp. 105-139. Sherif, C.W., Sherif, M.S. & Nebergall, R.E. 1965. Attitude and Attitude Change. Philadelphia: W.B. Saunders Company. Southey, G. 2011. The theories of reasoned action and planned behaviour applied to business decisions: A selective annotated bibliography. Journal of New Business Ideas & Trend, 9(1): 43-50. Statistics South Africa. 2014. Documented Immigrants in South Africa. Statistics South Africa, Private Bag X 44, Pretoria 0001. UNICEF. 2005. Trafficking in Human Beings, especially Women and Children, in Africa. (2ndEd). The UNICEF Innocenti Research Centre: United Nations Children’s Fund. UNODC. 2010. Organized Crime Involvement in Trafficking in Persons and Smuggling of Migrants: Issue Paper. United Nations Office for Drugs and Crime. United Nations (UN). 2000. Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children, Supplementing UN Convention against Transnational Organized Crime. Geneva: United Nations. AHMR, Vol.3 No1, January-April 2017 731 US Department of State. 2016. Trafficking in Persons Report. Washington DC: United States Department of State. Waganesh, W., Abebaw, M. & Kangangn, G.2015. Mental health and somatic distress among Ethiopian migrant returnees from the Middle East. International Journal of Mental Health &Psychiatry, 1(2): 1-6. Weine, S.M. & Kashuba, A. 2012. Labor migration and HIV risk: A systematic review of the literature. AIDS and Behavior, 16(6): 1605-1621. Yaseen, H. 2012. Sudanese youth attitudes towards international migration: Case study of some Universities in Khartoum State. Ahfad Journal, 29(2): 108- 109. Yoseph, E., Mebratu, G. & Belete, R. 2006. Assessment of trafficking in women and children in and from Ethiopia. IOM Special Liaison Mission, Addis Ababa, Ethiopia. AHMR, Vol.3 No1, January-April 2017 732 The Impact of Migration on the Welfare of Households in Ghana: A Propensity Score Matching Approach Abdul-Malik Abdulai*, Louis Boakye-Yiadom** and Peter Quartey*** Abstract Individuals usually migrate to improve their economic conditions and those of their family members left behind. However, less is understood about the impact of internal migration on the well-being of family members left behind. This study contributes to the debate by offering an empirical assessment of the welfare difference between migrant and non-migrant households by analysing data recently collected by the Centre for Migration Studies, University of Ghana. Results from the analysis are inconclusive. Using two indicators of measuring welfare (the World Bank Welfare Index and the number of children enrolled in school at the time of the survey), the latter shows higher welfare in favour of non- migrant households whilst the former indicates equality in welfare for all households. On the other hand, econometric estimations using the Propensity Score Matching (PSM) did not find any significant difference between the welfare of households that sent out migrants and those that did not. Keywords impact, internal migration, households’ welfare, Ghana. Introduction In developing countries, it is fairly common to see one or more household members migrate to either urban areas or neighbouring countries, attempting to increase their income and improve the economic conditions of their households. While remittances from migrant workers have been shown empirically to improve the economic conditions of family members left behind (Ackah & Medvedev, 2010; Quartey, 2006; Yang, 2005), less understood is the impact of migration on the well-being of family members left behind, more specifically, households with internal migrants. Since most migrant workers *Department of Mathematics, University for Development Studies, Navrongo, Ghana. **Department of Economics, University of Ghana, Legon. ***Department of Economics, University of Ghana, Legon. AHMR, Vol.3 No1, January-April 2017 733 migrate alone, and in some cases migrate with spouses,1 their children and other dependents (the aged) are left in the care of others. As such, a question worth investigating is: To what extent is the well-being of the left-behind affected by this migration? It is evident that though migration is prevalent in Ghana and is also not a new phenomenon, very few studies have rigorously examined its welfare impact empirically. While several causes might account for this gap in the Ghanaian literature, it does appear that data limitation and modelling difficulties are some of the main reasons. Individuals normally participate in migration to improve their well-being, whether those decisions are made at the individual or household levels (Lipton, 1980; de Haan, 1999). A household is likely to send a migrant when the expected value of the migrant’s remittances exceeds that individual’s net contribution to household welfare prior to migration. That is to say that in the ideal situation, migration would always be welfare enhancing. In the welfare literature, however, there are mixed findings regarding the likely impact of migration on welfare. While some studies find a higher welfare for households with migrants (see Ackah & Medeleve, 2010; de Brauw et al., 2012), others did not find any significant impact of migration on household welfare (see Boakye-Yiadon, 2008; Litchfield & Waddington, 2003; Farrington & Slater, 2006; Lloyd-Sherlock, 2006; Sahn & Alderman, 1996). The current study contributes to the debate by offering a novel empirical assessment of the welfare difference between migrant and non-migrant households by drawing on and analysing data recently collected by the Centre for Migration Studies, University of Ghana and funded by the Sussex University, U.K. Methodology of the Study Data used for this study is sourced from the Migrating Out of Poverty (MOP) dataset collected by the Centre for Migration Studies, University of Ghana, in collaboration with Sussex University, U.K., from March to May 2013. This dataset is unique2 over the census datasets, which are predominantly used by most studies in Ghana. A multi-stage sampling technique was used to select 315 households in the Northern Region, consisting of 231 migrant and 84 non- migrant households for interview. The first stage was a purposive selection of 1 This is more specific to rural-rural migrants who migrate for the purposes of farming. 2 Given that the survey was specifically designed for migrant source regions with high incidences of poverty, it was more detailed and comprehensive. AHMR, Vol.3 No1, January-April 2017 734 11 districts consisting of 21 Enumeration Areas (EAs) selected from a list of EAs in the 2010 National Population Census results. The number of EAs selected from each district is proportional to the total number of out-migrants from each district. These districts are the Tolon, Savelugu-Nantong, Kumbungu, Sagnarigu, Central and East Gonja, West and East Mamprusi, Kariga, Yendi, and Tamale Metropolis. The selection of these districts was informed by the high proportion of out-migrants produced by those districts recorded in the 2010 National Population Census results. This was followed by a listing of households in these selected EAs. The second stage involved a systematic sampling with a random start to select migrant and non-migrant households from each of the selected EAs. The migrant households were further stratified into three groups: seasonal migrants, return migrants and absent out-migrants. A total of 15 households were purposively selected from each EA, consisting of 4 households from the non-migrant households and 11 households from the migrant households. The final stage was the purposive selection of an adult household member, irrespective of gender, who had more knowledge about the household for interview. Estimation Procedure Given that individuals or households undertake migration to improve on their well-being as espoused by economic theories, the natural questions that arise are: Why are some people failing to migrate? What, if any, are the welfare differences between migrant and non-migrant households as a result of migration? Clearly, these questions fall under impact evaluation. One major challenge in empirical studies regarding impact evaluation is the issue of a counterfactual, that is, a hypothetical scenario of what the welfare profile of households would have been in the absence of the activity (migration). The counterfactual is hypothetical because it is impossible for the researcher to know or observe exactly what the welfare levels of households would have been had one or more of their members not migrated or if the activity had not occurred. Methods of welfare impact can be classified into two broader types: experimental and non-experimental methods. Experimental methodologies randomly select a control group prior to the application/onset of the activity and, therefore, individuals or households belonging to this group are then exempted from the intervention/activity (see Burtless, 1995; World Bank, 2007; Galasso et al., 2001). This approach is suitable for evaluating AHMR, Vol.3 No1, January-April 2017 735 interventions for which participation can be controlled by the researcher or for which participation is randomly and naturally determined. Migration flow is not an activity that can be assessed experimentally in terms of its impact on welfare since participation is neither random nor subject to a researcher’s influence. Non-experimental approaches, on the other hand, consist of a wide range of techniques that construct a control group to facilitate comparisons with a treatment group (see World Bank, 2007; Ravallion, 2001; Moffitt, 1991). These techniques include propensity score matching (PSM), instrumental variable (IV) and the double difference (difference in difference) methods. The first two techniques will be used in the current study for comparison to deal with the issue of selection bias. The last technique cannot be used as it requires data on treatments and controls before and after migration. Instrument variables are variables that matter to participation but not outcomes. That is, a variable (instrument) that is related to migration but not to welfare is chosen. The instrumentals are first used to predict participation in activity (migration) and then one sees how the outcome indicators vary with the predicted values conditional on other characteristics. PSM, on the other hand, is a statistical matching technique that attempts to estimate the effect of a treatment, policy or other intervention (in our case, migration) by accounting for the covariates that predict receiving the treatment. The PSM approach tries to capture the effects of different observed covariates X on participation in a single propensity score. Then, outcomes of participating and non-participating households with similar propensity scores are compared to obtain the program’s (participating in migration) effect (see Mensah et al., 2010). That is, propensity score: P(X) = Pr (T = 1/X). In using the PSM, two assumptions need to be made. These are the conditional independence and presence of common support (Rosenbaum & Rubin, 1983). The conditional independence states that given a set of observable covariates X that are not affected by treatment; potential outcomes Y are independent of treatment assignment T. If YT represents outcomes for participants and YC outcomes for non- participants, conditional independence implies: (YT, YC) ┴ Ti/X. The second assumption – common support or overlap condition: 0 < P (Ti = 1/Xi) < 1. AHMR, Vol.3 No1, January-April 2017 736 This condition ensures that treatment observations have comparison observations ‘nearby’ in the propensity score distribution (Heckman et al., 1999). In matching participants to non-participants, different matching criteria can be used to assign participants to non-participants on the basis of the propensity. Primarily, there are four matching methods. These include the Nearest-neighbour matching (NNM), Radius matching (RM), Caliper Matching (CM) and Stratified matching (SM).3 In this study, the NNM technique with replacement and a caliper imposed to avoid poor matching is used. In the NNM, the absolute difference between propensity scores in the treated and control groups is minimised. The control and treatment subjects are randomly ordered where a treated subject is matched with a control subject with a closest propensity score. C(Pi) = Minj│Pi - Pj│ Where C(Pi) is the group of control subjects j matched to treated subjects i on the estimated propensity scores. Pi is the estimated propensity score for the treated subject i. Pj is the estimated propensity score for the control subject j. Having estimated the propensity scores, the next thing to do is to examine the impact of program participation (migration) on potential outcomes (in this case, welfare). The impact of a treatment for an individual i, noted δi, is defined as the difference between the potential outcomes with treatment and without treatment. Thus, δi = Y1i −Y0i. The mean impact of treatment is obtained by averaging the impact across all individuals in the population. This estimate is called Average Treatment Effect (ATE). Implies ATE = E(δ) = E(Y1-Y0). A quantity of interest is the Average Treatment Effect on the Treated (ATT), which measures the impact of participation in migration on the welfare of those households left behind. 3 See Khandker et al. (2010) for a more detailed discussion of the various methods. AHMR, Vol.3 No1, January-April 2017 737 i.e ATT = E(Y1-Y0/D=1). Finally, the Average Treatment on the Unmatched (ATU) measures the impact that the treatment would have had on those who did not receive the treatment (control group). In welfare literature generally, there are two main ways of measuring welfare. These are the consumption and income approaches. Each of these approaches has advantages and disadvantages. Consumption is usually measured better than income in survey data; it is less noisy and better reflects long-term household well-being. It is also a standard measure of welfare, allowing for cross-country comparisons with the results of other studies (Bontch- Osmolovski, 2009). The use of income, on the other hand, may be comparable across space (de Brauw et al., 2012). This study adopted the consumption approach of measuring welfare because of the reasons cited above by Bontch- Osmolovski (2009). To assess the welfare difference between migrants and non-migrants in a more robust framework, the methodology of Beegle et al. (2011) was followed. The model is specified as follows: ∆ln𝐶 = 𝛼0 + 𝛼1𝑀 + 𝛼2𝑋 + 𝛼3𝐻 + 𝜀 …………………(1) Where ∆𝑙𝑛𝐶 is the change in the logarithm of monthly per capita consumption for a given household, X represents individual characteristics, M is an indicator variable for an individual who migrates out of the household or if a household had a migrant at the time of the survey, and H represents a household-specific fixed effect. An advantage of this model is that one can control for individual differences through the vector X by including categorical variables for age, gender, marital status and educational status in X. Results and Discussion A comparison of mean consumption expenditure as a measure of welfare across districts in the study area revealed that Tamale metropolis and Sagnarigu district have the highest mean welfare values (GH¢9,419.12 and AHMR, Vol.3 No1, January-April 2017 738 GH¢9,499.12, respectively). 4 This result is expected, as these two areas constitute the capital of the Northern Region of Ghana where, all being equal, living standards are relatively high and more job opportunities are available. Savelugu district recorded the least mean welfare in the region. Using the welfare index, a common approach of welfare measure (World Bank, 2008) expressed as the ratio of the mean consumption of migrants to that of non-migrants, gives an index of 0.9786. 5 This indicates equality in welfare between migrant and non-migrant households. One other measure of non- financial welfare used widely in the literature is the proportion of children enrolled in school by households (see Dudwick et al., 2011; Molyaneth, 2012). Among migrant and non-migrant households, we estimated the proportion of children between the ages of 5 and 15 years (inclusive) who were at school at the time of the survey. The result indicates that there is a welfare difference between migrant and non-migrant households, with the latter households enjoying higher welfare (having higher numbers of children enrolled in school). As high as 81 percent of children in non-migrant households were enrolled in school, compared to 27.3 percent in migrant households. Perhaps the low enrolment in migrant households could be explained by the absence of migrant parents to send their left-behind children to school in cases where nobody is fully responsible for their wards at origin. This is consistent with the conventional knowledge in the literature that presumes negative consequences of parental migration for the children who are left behind at origin (see Dinbabo & Nyasulu, 2015; Xin & Chikako, 2015; Liang et al., 2008; Xiang, 2007). The acquisition of household consumer durables by households, which is an indicator of household welfare, was also examined. The study noted that a higher percentage of non-migrant than migrant households acquired new houses and agricultural lands. Approximately 25 percent of non-migrant households acquired new houses compared to 23 percent of migrant households. Also, about 20 percent of migrants’ households acquired more agricultural lands, compared to 25 percent of non-migrant households. The low patronage of agricultural lands by migrant households could be explained perhaps by the identification of other more productive investment sources 4 Equivalent to US$ 2,943.48 and US$ 2,968.48, respectively, at an exchange rate of US$1 to GH¢3.2 as at January 2015. 5 An index closer to 1 indicates greater equality between groups on the welfare measure (Dudwick et al., 2011). AHMR, Vol.3 No1, January-April 2017 739 than acquiring agricultural lands and might therefore be diversifying. This result is at odds with the findings noted in the literature that through remittances migrant households are more able to improve their agricultural land holding than non-migrant households. As expected, more migrant than non-migrant households acquired household electronic goods and other electronic appliances. Also worth noting is the small number of households6 that reported acquiring household ‘white goods’ (fridges and washing machines). This could be explained by the fact that some of the communities in the study area were not connected to the national grid or, if connected, might have had difficulty paying for the exorbitant electricity bills. In recent times, sanitation has been considered important not only for healthy living but also for ensuring a filth-free environment. This is undoubtedly welfare enhancing. Availability of places of convenience in the study area was therefore examined. The result shows that regardless of the migration status of households, almost all households (99 percent) in the study area have bathrooms. On the other hand, the majority of households (81.40 percent) do not have toilet facilities. 7 This figure is about four times higher than the national average of 19 percent, but slightly lower than the regional average of 86 percent (GSS, 2012). However, it is observed that a higher percentage of migrant than non-migrant households (85 versus 81 percent) do not have toilet facilities, indicating that non-migrant households do better in constructing toilet facilities at homes and enjoy higher welfare than migrant households. This finding is consistent with Awumbila et al.’s (2014) study that reports that close to 94 and 63 percent of migrant households in Old Fadama and Nima, respectively, do not have toilet facilities in their residence. Sundari (2003) noted similar findings about Chennai slums. To assess the impact of migration on households’ welfare in the study area and to correct for selectivity bias, a regression equation was estimated using the Heckman two-step selection model. The first step was to choose an instrument that is correlated with migration as an independent variable and is not correlated with per capita consumption expenditure as the dependent variable. Ethnicity of individuals was chosen as an instrument and it was argued that people are likely to migrate if they are identifiable by certain 6 0.71 and 0.70 percent, respectively, for non-migrant and migrant households. 7 This high deficit of toilet facilities registered by households is because most households use the bush as a place of convenience or use public toilets where available. AHMR, Vol.3 No1, January-April 2017 740 ethnic groups. For example, in most of the other regions, certain ethnic groups8 have replicas of their chieftaincy institutions and, therefore, people who travel to those destinations and are identified with these institutions are accommodated and given the necessary assistance to adjust. The same argument cannot be made for consumption expenditure. To test the relevance of the instrument, the authors regressed the ethnicity variable on per capita consumption expenditure as proposed by Mckenzie and Sasin (2007). The result shows no association between the two variables. The second step was to regress the log of per capita consumption expenditure on the regressors, including the migration status using the Heckman two-step selection to counter the selectivity bias. The Heckman two-step selection involves, first, estimating a probit model to drive the inverse mills ratio. The second step is to estimate the outcome equation including the inverse mills as a regressor. Five welfare variables were included in the equation as explanatory variables and these are the human asset variables, the house composition variables, the location variables, the employment variables and the physical asset variables. Results from the estimation indicate that most of the parameter estimates are statistically significant and rightly signed, as shown in Table 1. Consistent with prior expectations and existing literature, educational qualification of household head increases with welfare. Implicitly, the more a household head is educated, the more likely he/she increases the earnings as predicted by the human capital model, therefore increasing the household consumption and welfare. Households increase their consumption expenditure by 6.3 percent as their heads increase their educational status. Age of head of household (Ageheadhh) is positive and significant, which means that household welfare is higher when the head is younger. This finding is consistent with studies by Ackah and Medvedev (2010). 8 For example, we have the Dagomba, Gonja and the Mamprusi Chiefs in most of the other regions in Ghana. AHMR, Vol.3 No1, January-April 2017 741 Table 1: Migration effect on welfare using the Heckman Two-Step Selection Model Variables LnConexp Migstathd Coef. Z Coef. Z Educquhead 0.0627*** 5.180 Ageheadhh 0.0181*** 4.830 Sexheadhh -0.5573*** -3.610 Ethnicindiv -0.114*** -7.740 Hsehdsize -0.0775*** -4.780 0.0599*** 5.530 Childdeprat -0.0195 -0.160 Maturindex -1.0585** -2.030 Lansize 0.0153*** 2.810 Ocupatn -0.1140*** -4.960 Distrt -0.0546*** -2.950 Migstathds 0.0955 0.730 Sexindiv -0.0776 -1.020 Ageindiv 0.0163** 1.960 Educindiv 0.0119*** 4.370 Comtyfacil 0.3829*** 8.590 Hseownship -0.2510 -1.220 Agesqur -0.0002* -1.680 _cons 7.0710*** 15.870 -1.074*** -3.550 Mills lambda -0.7839*** -3.830 Rho -0.6330 Sigma 1.2384 Lambda -0.7839 AHMR, Vol.3 No1, January-April 2017 742 Number of Observations 1207 Wald Chi2 140.73 Prob > Chi2 0.000 Note: ***, ** and * imply the level of significance at 1%, 5% and 10%, respectively. Source: Authors’ estimation from the MOP dataset. Sex of household head significantly and negatively relates to welfare as indicated by sexheadhh variable. This means that households with male heads have less welfare than their female counterparts. This can be explained perhaps by the practice of polygamy by male heads. Polygamous practice may result in larger household size9 and subsequently lead to a reduction in per capita consumption. This finding is at variance with those of Nwaru et al. (2011) and Sakiru (2013), which indicate that male-headed households have higher welfare than female-headed households in Nigeria. Having a larger household and a higher maturity index negatively affect household welfare, while owning land of greater size increases household welfare. The strong negative correlation between household size and welfare is consistent with Lanjouw and Ravallion’s (1995) findings, which show a negative correlation between household size and consumption per person in developing countries. There were spatial differences in terms of welfare as shown by the district variable (Distrt). Somehow at odds with literature and the widely held view that living standards and welfare are highest in cities, the findings of the current study show that households in all other districts in the study area have higher household welfare than those in the Tamale metropolis. Of much interest to the study is the effect of migration on household welfare as indicated by the Migstathd variable. The parameter estimate shows increases in household welfare with increasing migration, though not statistically significant. This confirms our earlier findings about the welfare and migration linkage using simple tabulation. The coefficient of the Mills ratio (λ) is negative and significant which lends support to the hypothesis of selectivity bias and, therefore, the model could not have been estimated using the OLS. Individual educational level and access to community facilities enhance household welfare. 9 Male-headed households tend to have larger household sizes using cross tabulation with a Chi- Square value of 152.09 at p < 1%. AHMR, Vol.3 No1, January-April 2017 743 An alternative method to the instrumental variable approach in dealing with the selectivity bias is propensity score matching. The authors used this method to also ascertain if there is a welfare difference between migrant and non- migrant households. The results from the average treatment effects, indicators of the quality of propensity score matching, estimation of propensity scores using the logit model, covariate balance as well as robustness checking or sensitivity analysis are presented below in Tables 2, 3 and 4, respectively. The summary statistics in Table 2 clearly show a difference in mean welfare between migrant and non-migrant households. That is, migrant households increase their consumption expenditure by Gh¢ 22.47 over non-migrant households. This mean difference in welfare may not just be the result of the treatment (migration) but could also be due to differences in individual characteristics. Table 2: Summary statistics of mean welfare difference before matching Group Observation Mean Std.Err Non-Migrant 1583 346.46 17.72 Migrant 209 323.99 36.36 Combined 1792 343.84 16.22 Diff 22.47 50.54 t = 0.4445 diff = Mean (Migrant) - Mean (Non-Migrant) Source: Authors' estimate from MOP dataset. To ascertain if individual characteristics explain the welfare difference, the propensity scores using the logit model were estimated. Results from the scores indicate that most of the variables used were rightly signed and significantly influence the probability to migrate. Sex of individual significantly and negatively associated with the probability to migrate (see Table 3), suggesting that males are less likely to migrate than females. This finding is at variance with the widely held notion in the literature that males are more likely to migrate than females (see Richter & Taylor, 2006). Marital status, household size and having access to community facilities negatively and significantly correlate with the probability to migrate. The negative effect of access to community facilities on the probability to migrate suggests that having access to electricity and pipe-borne water, among others, reduces the AHMR, Vol.3 No1, January-April 2017 744 likelihood of a household sending out a migrant (see Ackah & Medvedev, 2010). Table 3: PSM logistic regression result Migstathds Variables Coeff Z-stat Sexindiv -0.612*** -2.68 Ageindiv 0.314*** 4.97 Agesqur -0.005*** -4.99 Ethnicindiv 0.030 0.94 Maritindiv -0.439* -1.7 Ocupatn 0.191*** 2.88 Sexheadhh 0.626 1.59 Ageheadhh 0.011 1.19 Educquhead 0.011 0.36 Hsehdsize -0.075** -2.24 Hseownship -0.029 -0.07 Childdeprat 0.190 0.59 Maturindex 0.113 0.09 Comtyfacil -0.251** -1.89 Distrt 0.022 0.49 _cons -5.662*** -3.93 Number of obs 784 LR chi2(15) 88.36 Prob > chi2 0.00 Pseudo R2 0.13 Log likelihood -284.43 AHMR, Vol.3 No1, January-April 2017 745 Source: Authors’ estimate from MOP dataset. Examining the welfare difference between migrant and non-migrant households, the average treatment effect was estimated, as shown in Table 4. The result indicates a welfare difference between households with and without migrants. Focusing on the ATT, which is the average treatment effect on the treated, it is noted that migrant households increase their consumption expenditure (welfare) by GH¢ 14.17 higher than the non-migrant households with the difference not statistically significant. This result is at odds with our earlier result on welfare using the number of children enrolled in school, but is consistent with the finding of Awumbila et al. (2015). Table 4: Welfare difference between migrant and non-migrant HH after matching Variable Sample Treate d Control s Differenc e S.E T- stat Conexppcapi ta Unmatche d 319.14 310.63 8.51 52.16 0.16 ATT 340.21 326.04 14.17 68.65 0.21 ATU 323.16 444.21 121.05 118.8 5 1.02 ATE 105.18 108.6 9 0.97 Source: Authors’ estimate from the MOP dataset. We further conducted a balancing test between the covariates to ascertain if the level of heterogeneity between the treated (migrant households) and the control (non-migrant households) is eliminated after matching. Both the standard bias before and after matching indicates that the PSM using the nearest neighbour algorithm eliminates most of the bias between the treated and the control groups. Indeed, we do not observe any significant difference in means of all the covariates included in the model after matching. The proportion of bias reduction for most of these variables is at least 55 percent. The Pseudo R2 value after matching has also significantly reduced, suggesting that the overall results from the matching procedure have satisfactorily AHMR, Vol.3 No1, January-April 2017 746 created a balance between the covariates of the treated and the control groups (Sianesi, 2004). Conclusions and Recommendations The study examined the welfare difference between migrant and non-migrant households. A comparison of mean consumption expenditure (a proxy for welfare) across source districts shows that Tamale metropolis and Sagnarigu district have higher welfare values than the other districts studied in the region. Welfare difference between migrant and non-migrant households was not conclusive. While there is a higher welfare difference between migrant and non-migrant households, with the latter enjoying higher welfare using the number of children enrolled in school, there is equality in welfare between the households using the World Bank Welfare Index. On the contrary, an econometric assessment (PSM) shows higher welfare levels for migrant households, though not statistically significant. Regarding policies, it would be useful for government and other non- governmental organisations (NGOs) considering policies to streamline migration and improve on the welfare of the people to give more priority to education in terms of access and affordability. Also, government and other stakeholders should consider providing job opportunities and extending access to public facilities such as schools, electricity, health posts and potable drinking water to the people. Furthermore, for the recent government policy dubbed ‘sanitation day’10 (which, of course, is a welfare policy) to achieve the desired goal, district assemblies should enforce their by-laws that mandate every house to have a toilet facility as most households did not have toilet facilities. 10 This is a day reserved for every household to clean around its environs in the country, which falls on every first Saturday of the month. AHMR, Vol.3 No1, January-April 2017 747 References Ackah, C. and Medvedev, D. 2010. Internal Migration in Ghana: Determinants and Welfare Impacts. World Bank 2010 Ghana Poverty Assessment Paper. World Bank. Awumbila, M., Owusu, G. and Teye, J.K. 2014. Can Rural-Urban Migration into Slums Reduce Poverty? Evidence from Ghana. Migrating Out of Poverty Research Program Consortium Working Paper 13. Awumbila, M., Teye, J.K., Litchfield, J., Boakye-Yiadom, L., Deshingkar, P. and Quartey, P. 2015. Are Migrant Households better off than Non-migrant Households? Evidence from Ghana. Migrating Out of Poverty Working Paper 28. Beegle, K., De Weerdt, J. and Dercon, S. 2011. Migration and economic mobility in Tanzania: Evidence from a tracking survey. The Review of Economics and Statistics, 93(3): 1010-33. Boakye-Yiadom, L. 2008. Rural-Urban Linkages and Welfare: The Case of Ghana's Migration and Remittance Flows. Philosophy PhD Thesis, University of Bath. Bontch-Osmolovski, M. 2009. Work-Related Migration and its Effect on Poverty Reduction and Educational Attainment in Nepal. PhD Thesis, University of North Carolina at Chapel Hill. Burtless, G. 1995. International Trade and the Rise in Earnings Inequality. Journal of Economic Literature, 33(2): 800-816. de Brauw, A., Mueller, V. and Woldehanna, T. 2012. Does Internal Migration Improve Overall Well-Being in Ethiopia? ESSP Working Paper 55. de Haan, A. 1999. Livelihoods and poverty: The role of migration: A critical review of the migration literature. Journal of Development Studies, 36(2): 1-47. Dinbabo, M. and Nyasulu, T. 2015. Macroeconomic determinants: Analysis of pull factors of international migration in South Africa. African Human Mobility Review (AHMR), 1(1): 27-52. Dudwick, N., Hull, K., Katayama, R. Shilpi, F. and Simler, K. 2011. From Farm to Firm: Rural-Urban Transition in Developing Countries. World Bank, Washington D.C. AHMR, Vol.3 No1, January-April 2017 748 Farrington, J. and R. Slater 2006. Introduction: Cash transfers: Panacea for poverty reduction or money down the drain? Development Policy Review, 24(5): 499-511. Galasso, E., Ravallion, M. and Salvia, A. 2001. Assisting the Transition from Workfare to Work: Argentina’s Proempleo Experiment. A report on ex-post evaluation of the World Bank’s Social Protection III Project in Argentina, September 2001. Ghana Statistical Service 2012. 2010 Population and Housing Census Summary Report, Accra, Ghana. Heckman, J.J., Robert, L. and Jeffrey, S. 1999 .The Economics and Econometrics of Active Labor Market Programs. In: Ashenfelter, O. and Card, D. (Eds.). Handbook of Labor Economics, 3. Amsterdam: North- Holland, pp. 1865-2097. Khandker, S.R, Koolwal, G.B and Samad, H.A 2010. Handbook on Impact Evaluation. Quantitative Methods and Practices. The International Bank for Reconstruction and Development, World Bank. Lanjouw, P and Ravallion, M. 1995. Poverty and household size. Economic Journal, 105:1415-34. Liang, Z., Guo, L. and Duan, C.C. 2008. Migration and the well-being of children in China. The Yale-China Health Journal, 5:25-46. Lipton, M. 1980. Migration from rural area of poor countries: the impact on rural productivity and income distribution. World Dev. 8(1): 1-24. Litchfield, J. and H. Waddington 2003. Migration and Poverty in Ghana: Evidence from the Ghana Living Standards Survey. Sussex Migration Working Paper no. 10. Sussex Centre for Migration Research. London: MacMillan Pres ltd. Lloyd-Sherlock, P. 2006. Simple transfers, complex outcomes: The impacts of pensions on poor households in Brazil. Development and Change 37(5): 969- 995. McKenzie, D. and Sasin, M.J. 2007. Migration, Remittances, Poverty, and Human Capital: Conceptual and Empirical Challenges. World Bank Policy Research Working Paper 4272. AHMR, Vol.3 No1, January-April 2017 749 Mensah, J. Oppong, J.R. and Schmidt, C.M. 2010. Ghana’s national health insurance scheme in the context of the health MDGs: An empirical evaluation using propensity score matching. Health Economics, 19: 95-106. Moffitt, R. 1991. Program evaluation with non-experimental data. Evaluation Review, 15(3): 291-314. Molyaneth, H. 2012. Revisiting Poverty-Migration Nexus: Causes and Effects of Cambodia-Thailand Cross-Border Migration. PhD Thesis, Nagoya University. Nwaru, J. Iheke, C., Onwuchekwa, R. and Onyenweaku, C.E. 2011. Impact of migrant remittances on the welfare of arable crop farm households in South Eastern Nigeria. Human Ecology Review, 18(2). Quartey, P. 2006. The impact of migrant remittances on household welfare in Ghana. Research Paper 158, AERC. Ravallion, M. 2001. The mystery of the vanishing benefits: An introduction to impact evaluation. The World Bank Economic Review, 15(1): 115-140. Richter, S. and Taylor, J.E. 2006. Gender and Determinants of International Migration from Rural Mexico Over Time. (retrieved 17 March 2017). Rosenbaum, P.R. and Rubin, D.B. 1983. The central role of the propensity score in observational studies for causal effects. Biometrika 70(1): 41-55. Sahn, E. D. and Alderman, H. 1996. The Effect of Food Subsidies on Labor Supply in Sri Lanka. Economic Development and Cultural Change, 45(1), 125- 145. Sakiru, O.A. 2013. Profiles and determinants of poverty among urban households in South-West Nigeria. American Journal of Economics. 3(6): 322- 329. Sianesi, B. 2004. Evaluation of the active labour market programmes in Sweden. The Review of Economics and Statistics, 86(1): 133-155. Sundari, S. 2003. Quality of life of migrant households in urban slums. In: Martin, J., Bunch, V., Suresh, M. and Vasantha Kumaran, T. (Eds.). Proceedings of the Third International Conference on Environment and Health, Chennai, India. pp. 537-552. World Bank 2007. Impact Evaluation. (retrieved 17 March 2017). http://bit.ly/2o9ofg9 AHMR, Vol.3 No1, January-April 2017 750 World Bank 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. Xiang, B. 2007. How far are the left-behind left behind? A preliminary study in rural China. Population, Space and Place, 13:179-191. Xin, M. and Chikako, Y. 2015. Children of Migrants: The Impact of Parental Migration on their Children’s Education and Health Outcome. IZA Discussion Paper, No. 9165. Yang, D. 2005. International Migration, Human Capital and Entrepreneurship: Evidence from Philippine Migration‘s Exchange Rate Shocks. Policy Research Working Paper, World Bank. AHMR, Vol.3 No1, January-April 2017 751