Somaliland Cyberspace

Migrant Remittances in the Context of Crisis in Somali Society - A case study of Hargeisa

Anna Lindley
April 2006

Overseas Development Institute
111 Westminster Bridge Road
London SE1 7JD, United Kingdom
Tel. +44 (0) 20 7922 0300, Fax. +44 (0) 20 7922 0399
E-mail: hpg@odi.org.uk, Websites: www.odi.org.uk/hpg
and www.odihpn.org
http://www.odi.org.uk/hpg/papers/BG_Somalia_remittances.pdf

About HPG

The Humanitarian Policy Group at the Overseas Development Institute is dedicated to improving humanitarian policy and practice. It conducts independent research, provides specialist advice and promotes informed debate. Humanitarian Policy Group

Acknowledgements

This research was conducted with the valuable assistance of:

* Hargeisa University, particularly Ahmed Ismail Dahir, Ayaan Hussein Elmi, Dahir Isaaq Jibril,Khadra Abiib Mohamed, Khadra Ahmed mohamed, Mahdi Ahmed Cabdi, and Mohamed Abdillahi Abdirahman ‘Caynab’; and
* Dahabshiil Hargeisa, Financial assistance
* Overseas Development Institute (ODI).
* The African Studies Centre Travel Grant Programme, University of Oxford.

BACKGROUND

Since the 1980s, Somali society has experienced a protracted crisis, or series of crises, with devastating consequences for ordinary people. At the macro level, there are the political problems of factional fighting and an ongoing crisis of state institutions, as well as the economic problems of drought, disruption of the livestock trade by restrictions in foreign markets, and severe local currency depreciations. Within and often linked to these wider political and economic contexts are the personal crises of people’s everyday lives: loss of assets, livelihood deterioration or collapse, and experience of violence, death, displacement, illness and family break up.

After the outbreak of civil war in 1988 and since the collapse of the state in 1991, hundreds of thousands of people have fled Somalia.1 At least one million now live abroad, in neighbouring countries, or further afield. Many of these migrants around the world maintain contact with relatives back home and in the refugee camps and many remit money to support them. Today, these remittances are a major source of external revenue at the macroeconomic level and a significant source of income for a substantial minority of households.

This paper examines how remittances affect households that receive them. The first step was to review relevant literature and available statistical sources on migration and remittances in Somalia. There are several secondary sources, including nationallevel surveys and previous research reports that provide useful contextual information on livelihoods in Somalia, and studies on the Somali diaspora. There are also a few primary research sources specifically relating to remittances. The second step was to undertake exploratory research in Hargeisa, one of the largest Somali cities and the capital of present-day Somaliland. Hargeisa was selected because of the reported importance of remittances in the local economy and the limited security issues confronting research here compared with other locations. The research included consultations with people who work for remittance companies, non-governmental organisations (NGOs) and government bodies; interviews with a small group of Hargeisa residents; and a survey of over 500 people collecting remittances from money transfer outlets in the city.

This paper is divided into two parts. The first part amalgamates available information on remittances, reviewing the role of migration and remittances in Somali society historically and since the collapse of Somalia in 1991, and assessing the scale of, and the significance given in the literature to, remittance flows. The second part is a case study of Hargeisa, establishing the demographic characteristics and socio-economic situation of people receiving remittances, examining remittance patterns, relationships and geography, and exploring the role of remittances in the lives of recipients and the uses to which this money is put. Lastly, evidence on the role of remittances among people coping with crises is explored.

Chapter 1: Introduction

2.1. Historical migration patterns

Ethnic Somalis have traditionally lived in an area in the Horn of Africa that stretches into modern day Djibouti, the Ogaden region of Ethiopia and northern Kenya. Largely nomadic pastoralists or agro-pastoralists, people have moved within and beyond Somalia’s 1960 borders for centuries as part of routine cycles to graze and water livestock, to trade or to work. Colonial occupation by Italy (to the south) and Great Britain (to the north) generated some longer-distance migration paths. For example, young men from pastoralist families in British Somaliland were recruited into the British Merchant Navy, to work in the engine rooms of cargo steamers (Hussein, 1993, p. 54). In 1960, the independent Republic of Somalia was created through the union of the Italian-administrated United Nations Trust Territory of Somalia to the south and the newly independent Somaliland to the north. General Mohamed Siyad Barre came to power in a coup d’état in 1969, supported first by military and development aid from the Union of Soviet Socialist Republics (USSR), and then from the United States (Lewis, 1980).

From 1973, many people went to work in the booming economies of the Gulf States as construction labourers, drivers, house cleaners and maids, and guards, and some worked in skilled occupations, for example as carpenters, cooks, electricians, masons, and heavy truck and machinery drivers (Sheikh Ali, 1997). Particularly after the prolonged drought (dabadheer) of 1974, people were keen to seek better economic opportunities (Xasan Gaani, 2005). By the mid- 1980s, most city dwellers had a close relative in the Gulf, mainly Saudi Arabia (Simons, 1995). The majority was men, most often young, single and expecting to return home after a period—the Saudi regime did not allow permanent settlement (Mohammed, 2000). Urban wages in Somalia more or less stagnated throughout the 1970s, but inflation doubled between 1970 and 1978. In 1980, however, Somalis working in the Gulf countries were earning at least five to six times the average Somali wage, remitting perhaps around one-third of this (Jamal, 1988, p. 212). Jamal (1988) argues that remittances and trading activities, rather than waged employment, met the basic needs of most urban families. Thus, by the end of the 1980s, the overseas population was already playing a significant role in the economy.

2.2. Migration since the war

Emigration patterns were dramatically affected by the outbreak of civil war in 1988 and the collapse of the state in 1991. With dwindling foreign aid and a shrinking formal economy to milk, the regime was sinking into economic and political crisis. The 1980s saw the rise of armed opposition to Barre in the northeast, then the northwest, and then in the central and southern regions (World Bank, 2005). The Somali National Movement (SNM), drawing on the Isaq clans for support and financed largely by disgruntled elements in the diaspora, initiated an insurgency movement in the northwest.3 The government responded in 1988 with savage reprisals against the local civilian population, destroying Hargeisa, killing more than 50,000 people, and generating larger-scale displacement internally and into Ethiopia and Djibouti (Gundel, 2003). In 1989, the United Somali Congress (USC) and the Somali Patriotic Movement (SPM) launched rebel movements against the regime. In January 1991, the USC ousted Barre from Mogadishu and appointed an acting president and interim government. The SPM and the SNM rejected their authority, and as the main factions failed to reach an agreement on a new government, the civil war continued, and state institutions collapsed (Jama, 1997). In many places, militia looted businesses, and robbed, raped and killed people associated by clan with opposing groups or those belonging to minority groups. Where possible, civilians fled within Somalia to their clan homelands or over the borders.

The northwest seceded in 1991 as the Republic of Somaliland, claiming the previous boundaries of the British protectorate. It overcame civil strife in 1992 and 1994–96 to build state institutions and establish some basic public services, and to hold elections. While Somaliland is not recognised by any other government, and the eastern region of Sool and Sanaag are disputed, its territory has been largely stable, has received large numbers of returnees from the Ethiopia refugee camps and even has attracted some returnees from the Middle East and beyond. In 1998, the regional state of Puntland was formed in the northeast, as part of the former Republic of Somalia.

Puntland weathered a political crisis caused by the failure to transfer power at the end of the first administration’s term, and remains relatively stable compared with the central and southern region (World Bank, 2005). In some other regions, localised politics, based on traditional clan-based, municipal, Islamic and local business influences, have provided a degree of stability, but high levels of violence continue in central and southern Somalia. The Transitional Federal Government, formed in Kenya in 2004, has partially relocated to a temporary seat in Jowhar, Middle Shabelle, but cleavages among factions hamper its progress.

Chapter 2: Background: crisis, migration and remittances in Somali society 2

Somalia today remains in serious political crisis. Key economic and human development indicators are shown in Table 1. Since the beginning of the conflict, many people have fled to neighbouring countries or further afield: although the refugee population has declined gradually since 1995, it remains one of the largest in the world. With at least 389,272 refugees in 2004, the Somali diaspora is spread across a range of countries (see Table 2).

2.3. The contemporary Somali diaspora

The refugee figures only capture a portion of Somalia’s overseas population: it is reasonable to assume that at least one million people from Somalia now live abroad (UNDP, 2001). These people can be divided into three broad geographical groups. Immigration status and occupational patterns in each geographic group shape their remittances.

First, many people, possibly the majority, live in neighbouring Djibouti, Ethiopia or Kenya. In these countries, there were already some long-established ethnic Somali communities: some new arrivals received support from family members. Thousands eke out an existence on meagre humanitarian aid in the refugee camps. Others have tried their luck in the cities, often leading an economically precarious existence (Human Rights Watch, 2002). Some people who moved for business reasons before the war, or who have been particularly successful since, are able to send remittances regularly to support family members in Somalia. However, research among refugees in the Kenya camps, Nairobi’s Eastleigh district and in Cairo, suggests that they are much more likely to receive than send remittances (Horst, 2003; Al-Sharmani, 2004; Lindley, forthcoming).

Second, many people live in the Middle East: mainly in Saudi Arabia, the United Arab Emirates (UAE) and Yemen. Those in the latter nation have limited economic opportunities, many living in refugee camps. When the state collapsed, some migrant workers in the Gulf unable to return home stayed on, sometimes joined by family members, although the Gulf States did not officially accept refugees. Yet, as the labour market in the oil-producing countries tightened in the 1990s, many people were forced to move on, to Europe, or back to Somalia. In 1987, it was estimated that there were 375,000 Somali migrant workers abroad, mainly in the Gulf (Ahmed, 2000). Some estimates suggest that 20,000 people still live in Saudi Arabia and 25,000 in the UAE (Pérouse de Montclos, 2003). There is very little research on the socio-economic situation of Somali migrants in the Gulf today. It seems that, with a growing population in the West, the importance of the Middle East as a source of remittances has subsided since the heydays of the 1980s.

Third, a portion of the overseas population lives in Europe and North America. Some sought asylum, while others acquired a visa to reunite with family members who were citizens or were resettled by agencies on international programmes. There are high levels of unemployment among Somali populations in some northern European countries (see, for example, Greater London Authority, 2005). Employment rates in Italy may be higher, with more women working, often informally, as housekeepers, babysitters and care workers with local families (Sørensen, 2004). In London, people often work in public transport, local government, adult education, community and care work , and as taxi drivers or in small retail businesses, particularly internet cafes (Lindley, forthcoming). Many in the diaspora in the West send remittances on a regular basis to family members in Somalia and elsewhere in the Horn of Africa.

Table 1: Contemporary Somalia: key statistics, 2002

Population size 6.8 million
Average household size 5.8
Proportion of population living in urban areas 34%
Per capita household income USD 226 (urban USD 291; rural USD 195)
Proportion of population living in extreme poverty4 43% (urban 24%; rural 53%)
Adult literacy rate 19%
Gross primary school enrolment rate 17%
Table 2: Refugee population originating in Somalia by
major countries of asylum, 2004
Country of asylum End of 2004
Kenya 153,627
Yemen 63,511
UK 36,700
US 31,110
Djibouti 17,331
Ethiopia 16,470
Netherlands 11,239
Denmark 7,788
South Africa 7,118
Norway 6,242
Source: UNDP Somalia and the World Bank, 2003. Source: UNHCR, 2005 (Table 4)

These patterns of migration are socio-economically differentiated. People reaching the Gulf or the West have tended to be from cities or nomadic families, better off and better educated than the average (Gundel, 2003). There are substantial economic costs involved in travelling to Europe and North America, which are increasing as regulation of immigration in these countries tightens. Family reunion mechanisms privilege social groups that already have members abroad. For example, Somalis living in the United Kingdom before the war were mainly from the Issaq and Darod clans and originated in the north, so family reunion rules initially tended to perpetuate the regionalised and clan dimension of migration to the UK (El-Solh, 1991; Lewis, 1994). Subsequently, new ‘chains’ of migration developed and the clan base is now much more diverse (Pérouse de Montclos, 2003). However, Gundel (2003) notes that, compared with majority clan members, those from weaker clans and minority groups in Somalia,5 although particularly badly affected by the violence, have not migrated in any substantial numbers beyond Africa. These groups were unlikely to be able to afford the costs of emigration to richer countries (as they were originally socio-economically disadvantaged), more vulnerable during the conflict (lacking arms to defend themselves, and less likely to survive, and more likely to lose all their assets), and had more limited support networks on which to draw to finance further migration (smaller groups, and more sedentary, with mainly local networks). Some minority groups, though, have moved beyond Africa, for example through the resettlement of groups of Somali Bantu people in the US. The situation of disadvantaged and minority groups in Hargeisa is explored later in this paper.

2.4. Insights into the scale and significance of remittance transfers

There are two main methods to estimate total remittance flows to Somalia. The top-down approach uses data provided by remittance companies. There has been no official data on remittances since 1990, and even before the state collapsed, remittances were mainly travelling through channels not monitored by the Central Bank. As the vast majority of remittances today pass through Somali remittance companies, some of which are very large operations with central clearinghouses, attempts have been made to estimate the total volume of remittances to Somalia based on company records. Estimates range from USD 375 million (Sheikh Ali, 1997) to USD 800 million–1 billion (Omer, 2002). Top-down estimates have several limitations: they omit flows through other systems (for instance, traders or hand carrying); there is sometimes unwillingness to cooperate on the part of remittance companies (as information on transaction levels is commercially sensitive); and it can be difficult to disaggregate family remittances and other types of investments and trade transfers.

The bottom-up approach, meanwhile, is based on information on average transfers and numbers of migrants, or data derived from household surveys in Somalia. The United Nations (UN) estimated in 2003 that remittances made up USD 360 million of household income (UNDP Somalia and the World Bank, 2003). Pérouse de Montclos (2003) came up with an estimate of ‘over $120 million a year’ in remittances sent by Somalis in major host countries to Somali family members in the Horn of Africa, based on limited information regarding diaspora population and remittance habits.6 Obvious shortcomings to using this bottomup approach in the diaspora context are the lack of reliable data on population size and composition, income, and the incidence and patterns of remittance behaviour among Somali migrants in different host countries (or the time-consuming process of obtaining existing data from multiple national sources).

Based on a combination of top-down and bottom-up methods, on information collected from remittance companies, households and business people in Somalia, the United Nations Development Programme (UNDP) estimated that in 2004 remittances amounted to roughly USD 700 million–1 billion in Somalia, of which USD 360 million was directed at the subsistence of families, with the remainder going to business and investment (Omer and El Koury, 2004). In 2005, the World Bank estimated that the total volume of remittances to Somalia was more than USD 825 million (World Bank, 2005). Even the lower end of these approximations supports the conclusion that remittances are significant at the macroeconomic level. According to International Monetary Fund (IMF) trading partner data, in 2004, export and import values were USD 265 million and USD 461 million, respectively (World Bank, 2005). International aid from UN and bilateral and multilateral donors was USD 170 million in 2002, and USD 272 million was pledged for 2003 (Somalia Aid Coordination Body, 2003).7

In 2002, total household income for Somalia was estimated at USD 1.5 billion (UNDP Somalia and the World Bank, 2003). Thus, taking USD 360 million as a conservative estimate of ‘family remittances’, remittances may constitute 25 per cent of household income, compared with 50 per cent made up by self-employment and 14 per cent made up by wage employment. However, given the unreliability and variation in remittance flow estimates, comparisons with approximations of other macroeconomic flows should be made with caution. Existing reports make several points regarding remittances in Somalia. At the macroeconomic level, it is argued that remittances have contributed to the ‘dollarisation’ of the economy. The US dollar was widely circulated in Somalia under Barre; since 1989, this has increased (UNDP, 2001). Major transactions and savings are now in dollars and minor transactions in Somali shillings or Somaliland shillings (SlSh). Shillings are bulky and impractical for making large payments, and holding money in dollars protects against inflation.

Remittances have put dollars in the hands of more people. The migrant pays to convert his or her local currency (outside the 5 Groups considered to be minorities include: Bantu Somalis or Wa Gosha, Benadari (Hamari, Bravanese, Bajuni), Boni, Eyle, Gaheyle, Galgale, Midgan, Tumal and Yibir (Minority Rights Group International, 1997).

US) to dollars, and the recipient is paid in dollars at the other end, although he/she may choose to convert it immediately into Somali or Somaliland Shillings at the remittance company or through another trusted money exchanger. Remittances inflows also increase demand for imported goods—from cereals to cars to cement to televisions—encouraging trade (Ahmed, 2000). Moreover, the demand by hundreds of thousands of migrants wanting to send money home, along with the demand by traders for an efficient means of moving their funds, encouraged the development of a thriving financial sector, in the form of Somali remittance companies—the financial arteries of Somalia since the collapse of banking institutions in 1991.

The political effects of remittance inflows have also been touched on in the literature on Somalia in recent years. Diaspora groups have been mobilised to support various factions: ‘The diaspora has, during certain periods, been a critical source of funds for clan militias, which have raised funds (sometimes via coercion) from clan members living abroad’ (UNDP, 2001, p. 132). More indirectly, by allowing ordinary people to cope in the context of political crisis, some suggest that remittances have reduced political accountability: ‘One negative aspect of remittances is that they have allowed the factions to survive without delivering anything to the people’ (Marchal et al., 2000, p. 14). The larger political formations have also garnered support from the diaspora. There have been attempts in recent years to raise funds in the diaspora for Abdullahi Yusuf, then the leader of Puntland. In Somaliland, many would agree with the government’s Foreign Minister, Edna Adan Ismail, that ‘the disapora has brought Somaliland to where it is today’.8 Patriotic feeling and civic pride have been converted into valuable material support in the form of diaspora donations, investment and activism on behalf of political formations and communities in Somaliland and Puntland.

Reports on livelihoods in Somalia have also raised the issue of remittances, suggesting that those receiving money from abroad are usually in the middle and better-off income groups as opposed to poor or destitute—possibly reflecting the educated, urban origins of many migrants to the West (Adams et al., 2003). According to UNDP, ‘Many households would suffer significant setbacks in food security and access to private medical care and private schools if remittance flows were interrupted. Remittances are … an important mechanism for meeting human development needs at the household level’ (UNDP, 2001, p. 131). There is even evidence suggesting that remittances can help people to access international aid, where community contributions are required. Reportedly, for example, the 2001 closure of Al-Barakaat, the largest remittance company, ‘affected the ability of rural communities and families to participate in development initiatives, such as those that Oxfam has been supporting’ (Bradbury, 2002, p. 24).

As illustrated, several papers and reports refer to the significance of remittances in Somali society, providing information that helps to paint a rough picture. However, knowledge of the dynamics and effects of remittances on individual recipients, their households and communities and on the society as a whole remains rather limited, perhaps surprisingly so given the suggested significance of the phenomenon. The rest of this paper presents the findings of research in one specific urban context: Hargeisa.

3.1. Hargeisa: background

Described by I.M. Lewis (1980, p. 65) as ‘a haven of peace in a turbulent area, strategically placed at the intersection of the caravan routes leading from the coast to the Ogaden’, Hargeisa was founded in the nineteenth century by the religious leader Sheikh Maddar, in a northern Somali area mainly populated by the Habar Awal, an Isaq clan, but the town included members of various clans.9 The town grew considerably after Britain relocated its colonial administration here in 1941.

The 1988 violent repression of civilians (particularly those belonging to Isaq clans who Barre viewed as a threat) and the bombing of Hargeisa by government forces in the face of the SNM’s advance virtually emptied Hargeisa. People of all social groups fled from the city to their rural homelands, to Ethiopia and Djibouti, and further afield. When the state collapsed in 1991, and the independence of Somaliland was declared, people began returning to Hargeisa. Civil strife broke out between Isaq groups in 1994–97, which displaced more people into Ethiopia. Subsequently, the United Nations High Commissioner for Refugees (UNHCR) began voluntary repatriations to Somaliland (Lewis, 1980; WSP International, 2005). Since then, Somaliland has held a constitutional referendum (2001), local elections (2002), presidential elections (2003) and parliamentary elections (2005).

Today, the population of Hargeisa, the capital of Somaliland, is well over 300,000 and may reach 500,000. Most of the people of Hargeisa work in some kind of trade or service or for public institutions or NGOs, as shown in Table 3. Very few residents rely directly on pastoralism for income or food, although some are involved in the trade of livestock. People still emigrate from Hargeisa to other countries, generally to seek out better economic and educational opportunities.

Some officials estimate that inhabitants of Hargeisa receive around USD five million each month in remittances, two-thirds of which goes directly to meeting the expenses of households— the remaining one-third is invested in construction or related business (King, 2003). Medani (2000) found that in Hargeisa, nearly 40 per cent of residents (excluding those living in returnees on the outskirts of the city) received remittances, and for practically one-quarter, remittances were the main source of income. Ahmed (2000) found in 1998 in a survey of 116 recipient households in Hargeisa that an average of USD 4,170 was received per year, representing nearly two-thirds of their total income (as will be shown, our larger sample indicates a lower annual average). An urban economy assessment in 2003 suggested that remittances are an important source of income mainly among those in middle-income groups (see Table 3).

Table 3: Urban Household Economy Assessment, Hargeisa, 2003

Wealth group Daily household Food and income sources Daily household expenditure for a income* household of seven
Better off Over 80,000 SlSh – Large-scale business (e.g. import–export, Around 20% of income on food. Give gifts 10–15% of ~over $12 shops). in cash and kind to poorer relatives (rural households – Senior employees. and urban) and neighbours.
– Investments in property and vehicles (taxis, buses and trucks).
Middle 25–80,000 SlSh – Skilled labour (e.g. masons, carpenters). Around 50% of income on food. Give gifts 40–60% ~$3.75–12 – Mid-level government, NGO and (see above). private-sector employment.
– Mid-level petty trade (qat, clothes, larger quantities of vegetables and milk).
– Remittances. Poor 15–25,000 SlSh – Petty trade (vegetables, milk, Around $3 per day on food and non-food 20–35% ~$2.25–3.75 prepared food). items. More money allows more spending
– Casual unskilled daily labour (domestic on water, health care, clothing, qat, work, construction, market portering). education.
– Low-waged employment (domestic work, Very poor Under 15,000 SlSh shop and restaurant assistants, bus Around 13–14,000 SlSh on food and Remaining ~$2.25 drivers and conductors). essential non-food items. Purchase food households – Self-employment (donkey cart). daily in small quantities. School/health
– Gifts from neighbours and relatives. spending minimal.
– Begging.
– Children’s work in the poorest households.
* Note that the value of the Somaliland Shilling (SlSh) fluctuates: amounts correct as of 2003. Source: King, 2003.

Chapter 3: Case study of Hargeisa

3.2. Methodology

It is clear that a substantial minority of people in Hargeisa today receive remittances from migrant relatives. Who are they? How often and how much do they receive and from where? How has this shaped their resettlement in the city, and the lives they have subsequently led? Quantitative and qualitative data were collected using three methods:

First, key people were consulted, including government officials, remittance company staff, locally based researchers, and NGO workers, to obtain general information and expert views on emigration from Hargeisa and remittances.

Second, several local residents were interviewed (eight women and five men) to obtain more detailed information on the role of remittances in households. Topics discussed included: family, and household context; livelihood activities remittance patterns, and the roles of remittances in relation to the recipient in the wider community. These people were contacted through personal acquaintances, NGO workers, schoolteachers and members of the survey team, belonged to various clans and sub-clans, and had an assortment of occupations; five were under 30 years of age. The aim was not to build a representative sample of people living in Hargeisa, but to provide some detailed examples of remittance relationships. Interviews were held in people’s homes, offices and cafes: sometimes relatives were present or one member of the survey team accompanied the researcher to interpret. Interview notes were transcribed from tape or typed from handwritten notes shortly after the interview.

Third, a survey was conducted of people receiving remittances, to paint a reliable quantitative picture of who is receiving remittances, how much, how often and from whom, and how the money is used. By far the most common method for sending money to the city is through a Somali money transfer company. Dahabshiil, the longest-established company, is believed to be the largest international money transfer company serving the Somali regions. It is also the longest-established company in Hargeisa, has many more branches than any other operator, and is deemed to bring the majority of remittances into this city.10 The assumption was made that people collecting remittances from Dahabshiil are broadly representative of the remittance receiving population of Hargeisa as a whole. The company management agreed to cooperate with the research.

With the assistance of the University of Hargeisa, six men and women were recruited who had recently graduated in business administration. The week prior to the survey was devoted to training in survey skills and questionnaire preparation and practice in Somali. The survey was conducted over six days, from 31 August–6 September 2005—surveying at the beginning of the month maximised the sample size as this is when most remittances reach Hargeisa, after the people sending money from abroad receive their wages. Three pairs of interviewers surveyed people in a total of 12 branches, covering remittance collection points throughout the city, in areas where various clans/subclans dominated. The interviewers were organised to cover their own local areas where possible. Offices are usually divided into a back office, separated by a counter from a waiting room, where interviewers approached people coming in to collect money.

Respondents and interviewers were often the only people present; in the busier branches there was less privacy, which may have affected responses, although we have no evidence to suggest this happened. Only people who lived in Hargeisa and were collecting remittances were eligible to participate (some people visited the branches for other purposes or were from outside the city). Ninety-two per cent of eligible people approached agreed to participate—538 people in total. Information was gathered from respondents in Somali language via a four-page questionnaire, including individual, household and housing details, economic activities and sources of income. Detailed information was also collected on remittances received during the past 12 months by sender. In addition, respondents were asked if other members of the household received remittances, and how much. The researcher entered and analysed the data using the Statistical Package for the Social Sciences (SPSS).

There were three main methodological limitations:

* First, the person collecting remittances was not necessarily the person best positioned to give us information on the household in general, so we had to accept that certain responses (such as remittances to other family members) would be rough rather than exact.
* Second, as we carried out the survey at the beginning of the month, monthly recipients may have been over-sampled: if the survey had been conducted throughout the month or throughout the year, there might have been a slightly greater proportion of recipients who receive remittances infrequently or only in emergencies—and these may be the poorer recipients.
* Third, as respondents were usually busy going about their daily tasks, the questionnaire was designed to be conducted in 10 minutes or less, limiting the data collected.11 Bearing in mind these limitations, and given the steps outlined above, the sample is not precisely representative, but the 10 While some of the other companies and traders serve particular groups or carry transactions from a few host countries to Hargeisa, Dahabshiil has a broad customer base in the diaspora and serves people throughout the city. As a leading financial services company, it also offers banking facilities, including savings and cheque accounts, to individuals, businesses and international organisations in Somalia. More information on the company is available at http://www. dahabshiiil.com.

results may be taken as painting a broadly indicative picture of the remittance recipient population of Hargeisa as a whole.12 The survey and interview findings have been combined here to provide an account of: the profile of people who receive remittances; remittance patterns; the roles of remittances with regard to recipients; and insights into the role of remittances in the context of crisis in Hargeisa.

3.3. Remittance patterns, relationships and geography

Amounts received in the past 12 months by individuals ranged from only USD 50, a one-off transfer, through to around USD 14,000, averaging about USD 2,300 per year (or USD 190 per month). One-quarter of respondents lived in households where another member also received remittances. To keep the questionnaire short, a crude definition of the household, people who usually live with the respondent, was used. As people may not know for sure whether other family members receive remittances or how much, the value of total household remittances calculated here must be treated as an estimate, probably an underestimate. Total estimated house-hold remittance receipts are shown in Figure 1: the average was around USD 2,600 per year (roughly USD 215 per month), and the most common range was USD 1,000–1,499 (about USD 80–120 per month).

While the majority of respondents, around 80 per cent, received money from only one relative in the past 12 months, approximately 15 per cent received money from two people, and a very small proportion received money from three people. Over three quarters of respondents had received transfers each month or more frequently in the past year. There are some regular seasonal variations in remittance flows. Throughout the year, remittances peak at the turn of the month, after migrants receive their wages. One of the five pillars of Islam is Sako (Zakaat), the obligatory yearly giving of alms, representing 2.5 per cent of one’s assets above a minimum threshold, to the poor and needy. In Somali society, alms are often given to close relatives and friends (Ibrahim Cadbi, 2005). Thus, throughout the holy month of Ramadan, a fasting period, the volume of transfers to Somalia is increased, until the Eid festival.13 One interviewee, whose family regularly received USD 200 each month from relatives in the UK, said that his family often receives three times this amount before Eid, to help pay for clothes, food, children’s activities, social events et cetera. People who rarely receive remittances are more likely to receive them at Eid as distant relatives remember them. During the summer, some citizens of European and North American countries visit Hargeisa on holiday. There is a common expectation that migrants will be generous: and many migrants put aside money specifically to give to relatives and old friends who inevitably come calling.

Information was collected on 660 migrants from all over the world who had sent remittances to our respondents in the past 12 months. Around 60 per cent were men, and approximately 40 per cent were women. The women migrants sent slightly less money on average than the men. Most migrants were of working age, in their thirties or forties. Average amounts remitted were similar across age groups. Many people consulted during the fieldwork remarked that women play an important role in sending remittances, indeed several suggested that today, women are the biggest senders. This is not borne out by this study, but it is clear that throughout Somali society and in diaspora communities, women are increasingly economically active in new realms. The economic role played by women in remitting to their own families is not expected in traditional culture and therefore is widely noted.

Various relations send money to Hargeisa. Brothers, sisters, husbands, sons and daughters of the individuals we interviewed were the most common benefactors. Figure 4 shows the proportion of remittances received from different relatives. Brothers and sisters of the respondents sent 40 per cent of the volume of remittances. The sibling chain of assistance is particularly strong in Somalia: it is traditional for older siblings living in towns to help younger brothers and sisters in rural areas. Supporting elderly parents is also common, and is seen as an important moral and social duty. With many families affected by war deaths, separation or divorce, and many spouses reunited or married in the diaspora, marital bonds appear to be less critical with regard to remittances today than during the 1970s and 1980s.

The largest proportion of remittances arrived from Europe, although the Middle East and North America are also important (see Figure 3). Very small amounts of remittances trickled in from family members in neighbouring countries. By country, the largest number of migrants remitting to Hargeisa lives in the UK and the largest volume of remittances came from the UK, followed by the US, the UAE, Canada and Saudi Arabia (see Figure 4). As noted above, patterns of emigration are often differentiated according to clan or region. Given Britain’s colonial history in Somaliland, and preferable immigration conditions in the late 1980s,14 when the conflict in the northwest was at its height, it is not surprising that the UK is the most important source of remittances. In contrast, in other Somali cities, such as Baidoa, Bossasso or Mogadishu, one would expect relatively more people to have left for Italy, the Middle East and the US.

As my Somali is limited and I was not conducting the survey interviews myself, I used several strategies to keep up to date with issues arising during the research process and to check the quality and accuracy, including listening to the interviewers going through questionnaires with respondents, checking questionnaires, and organising regular debriefing sessions.

Figure 1: Total estimated remittances received by respondents’ households in the past 12 months
120 100 80 60 40 20 0
$6000> $5500–5999 $5000–5499 $4500–4999 $4000–4499 $3500–3999
$3000–3499 $2500–2999 $2000–2499 $1500–1999 $1000–1499 $500–999
$0–499

3.4. Profile of recipients

In demographic terms, the recipients surveyed were similar to the general adult population. The age and gender profile of respondents is shown in Figure 5. The most marked finding is that women form the majority of remittance recipients, and receive approximately two-thirds of the total volume of remittances. The age distribution of remittance recipients by amounts received was broadly similar to Somaliland’s urban adult population as a whole, suggesting that remittances do not benefit any particular age group much more than another. It is worth noting, though, that relatively few recipients were in their twenties.15

Figure 2: Proportion of total remittance volume transferred by different relatives

Other 17%
Paternal
Uncle 4%
Son 12%
Daughter 8%
Husband 16%
Mother 4%
Sister 17%
Brother 22%

Figure 4: Volume of remittances by countries of origin

Other 25%
US 12%
Canada 7%
Saudi Arabia 6%
UK 41%
UAE 8%

Figure 5: Age and gender of respondents

100  80 60  40  20  0
<20 20–29 30–39 40–49 50–59 60>

Figure 3: Volume of remittances by regions of origin

Other regions 1%
Africa 2%
North America 20%
Middle
East 17%
Europe 60%

The system employed by the remittance companies means that a named recipient collects the money transferred. The individual collecting remittances has a greater degree of control over the money, but in the majority of cases, they are routinely used to help satisfy general household needs. Households ranged in size from 1–25 people, but the average household size was 7.7. While most respondents (71 per cent) were living with their own child or children, and just over one-half were living with their spouse(s), wider relationships were also very common, including those built around parents, siblings and other relatives. (Male respondents were slightly more likely to be living with their spouse(s) than female respondents, who were considerably more likely to be living with their own children than were male respondents.) Around one-quarter of respondents were working for money or for a family business. Main paid work activities are shown by gender in Figure 6: nine per cent of respondents have regular work as an employee; five per cent does casual work; and 13 per cent is selfemployed.

A further 14 per cent is looking for work. The rest of those surveyed is primarily looking after the family/home; some are too old or ill to work, are students or do not need to work. Most (around 40 per cent) of those not working outside the home for money were women occupied with looking after children and the home. At the household level, around one-half of individual recipients live in households where there is some alternative form of income, from employment, business transactions or other assets. The other half lives in households that appear to rely entirely on remittances. The majority of remittance recipients reside in permanent housing made of brick or stone and do not pay any money for where they live.

How do remittance recipients compare with other residents of Hargeisa? The average estimated remittance receipt of survey respondents, around USD 200 per month, is substantially higher than average per capita income—estimated at USD 491 per annum in urban Somaliland (Academy for Peace and Development, 2004). Thus, even before counting any other sources of income, remittance recipients are much better off than the average town dweller. In Hargeisa specifically—where cash incomes tend to be higher than elsewhere in Somaliland and where there are relatively fewer non-cash sources of food—USD 200 per month is commonly seen as enough for a household to manage on, depending on the number of people. Compared with the general urban population of Somaliland, a higher proportion of remittance recipients live in homes made of brick or stone, indicating a higher socio-economic status (UNDP Somalia and the World Bank, 2003).

Hargeisa’s urban poor have tended to end up in nine particularly deprived areas in the city and on its outskirts: Sheikh Nur, Mohamed Mooge, Aw Aden, Ayaha, Sinia, Stadium, Dhami, State House and Beerta Xorriyada. There are marked differences between the fortunes of those in these settlements and elsewhere in the city. Fifty per cent of people in these settlements are classified as destitute compared with only seven per cent of urban residents (Medani, 2000). It appears that people in these areas less commonly have relatives abroad or resources to finance emigration. Remittances were a main source of income for 23 per cent of urban households, compared with five per cent of households in these settlements (Medani, 2000). However, it is important to note that there is an element of circular logic in saying that remittance recipients are better off. Even if a migrant can only spare USD 50, this is still a considerable amount by Hargeisa standards and the average amount received in the survey— around USD 200—is enough to lift a family out of poverty.

As already explained above, large numbers of people of the Isaq and Darod clan families were displaced in 1988 and 1991. These clans are associated with strong chain or networked migration patterns to the West. Receipt of remittances among all of these groups is not uncommon. Other groups in Hargeisa were affected differently by the war, and migration to the West has been less common, for example among the Dir and minority groups. The urban poor particularly include Oromo and Somali refugees originating in Ethiopia, Isaq returnees after the civil strife of the mid-1990s, people belonging to minority groups displaced from southern Somali areas, and the Gaboye (previously known as Midgan), Tumal and Yibir families. Most of these groups are less likely to have relatives in the West. For instance, the Gaboye historically had low social and economic status in Hargeisa, was unable to marry majority clan members, and was restricted to working in certain occupations. The Gaboye in the north were mobilised by Barre against the Isaq-based insurgency and were not able to return to Hargeisa immediately after the SNM victory.

When they did return, they settled in Dhami, on the outskirts of Hargeisa. Considerable social discrimination continues. A small portion of those who fled to neighbouring countries ended up in richer nations. Some others were already working in the Gulf when the war broke out. In general, though, poverty and more limited social networks prevented chain emigration to richer states on the scale of that by the Isaq and Darod clans. It is reasonable to assume that fewer Gaboye households receive remittances than among the general Hargeisa population.16

Figure 6: Main paid work and other activities by gender

Do not need to work       200
Too elderly or ill        150
Student                   100
Looking after family/home 50
Looking for work          0
Self-employed             0
Casual work               0
Regular employment        0
Work type or reason not working 16 Source of data on urban Somaliland: Academy for Peace and Development, 2004.

3.5. Roles and uses of remittances among people in Hargeisa

This section explores how remittances affect roles within households; how remittance income interacts with other livelihood activities; and how remittances are used. Remittances can affect roles and relationships in many families. In urban households, men would traditionally have the dominant role of breadwinner and resource owner (Ibrahim Cadbi, 2005). Conventionally, a husband will give his wife a set amount of money each month for regular household expenses, often paying it into their account at the local shop. For extra expenses, the wife will ask her husband for more money. The wife will have little oversight over the family finances and may not know how much the husbands earns and saves. While these traditional patterns remain very common in Hargeisa, a portion of households today increasingly depart from convention, frequently as a result of unemployment, displacement or divorce, which has led to more female-headed households (UNIFEM, 1998; Ibrahim Cadbi, 2005). An example is given in Box 1. Remittance inflows are part of this picture. Migrants today send remittances to a variety of female and junior relatives, as shown in the profile of recipients. This puts money in the hands of people who were traditionally entirely dependent on the resources of a male breadwinner, usually a husband or father. Where a woman has her own income from employment or remittances, she has a stronger say in family affairs. She may also have more responsibility: some people suggest that when the wife has her own income, the husband may take it for granted that she will feed the family. Research in the UK suggests that migrants are often very aware of these dynamics when choosing to whom to send money, regularly opting to remit to female relatives to ensure that the children benefit from the money sent (Lindley, forthcoming).

Much research on remittances has focused on whether remittances create and entrench dependence on migration or generate greater economic independence through investments in local livelihoods. Concentrating on the issue of dependency, the first question to ask is to what extent do people have other income sources? While income figures are not available, as indicated in the profile of recipients above, the survey suggested that for around one-half of the households, remittances represent the only source of income (remember that the survey sample is a little biased towards people who received remittances regularly).

The second question to ask is whether recipients are more or less economically active than the general population? While quantitative data on economic activity in Hargeisa are scarce, it is possible to make an (imperfect) comparison between the remittance recipient sample with information on people in towns in Somaliland in general (Academy for Peace and Development, 2004), and urban Somalia in general (UNDP Somalia and the World Bank, 2003), as illustrated in Figure 7 below. The sample of remittance recipients does appear to have lower than average rates of economic activity (that is, working for money or looking for work). Remittance recipients who are economically active have higher employment rates. Overall, the proportion of all adults who are working is considerably higher among the remittance recipient sample than the general population, and the proportion of people looking for work is considerably lower. These results should be interpreted with caution, not least because the indicators used in all three surveys of working status are fairly crude and do not take account of seasonal employment and underemployment, and because the data are not directly comparable. Remittance recipients may be able to afford to be economically inactive because of remittances, or conversely, they may be receiving remittances because they have been unable to find work. One might speculate that the greater success of remittance recipients in finding employment compared with the rest of the economically active population is due in part to educational and business investments, in many cases paid for by remittances. The third question, touched on above, to ask about dependency is whether there is evidence that receiving remittances diminishes efforts to pursue other livelihood activities? It is not possible to give a conclusive answer here. The survey evidence does not allow us to assess causality in the relationship between receiving remittances and economic inactivity. The qualitative research provided indirect evidence that remittances may diminish efforts to generate other sources of income.

Certainly, if one is able to rely on hard currency from abroad, and if it is enough to meet one’s needs, the returns with regard to local labour in Somaliland may be relatively low. As Mohamed, a young man I interviewed, put it, ‘Many relatives who receive remittances from abroad do not work and do not try to work effectively and may rely on it, and stay at home without doing anything. It depends on the nature of the person. I want to upgrade and enhance myself and help those in my family who need it.... It depends how the person responds to the situation. Some people, [in] the absence of relatives abroad may do a lot to subsidise themselves…. Others, in the case of no constant assistance, would have many difficulties’. Several interviewees pointed up their frustrations with the local labour market in terms of finding and holding down a job, and to the high levels of unemployment and underemployment in Hargeisa, and suggested that many who depend on relatives abroad do not have many reasonable alternatives. This viewpoint was typical: ‘When you have choice you can talk about dependency. But when there is such high unemployment, you can’t blame people for not finding a job’. Moreover, it is also important to underline that many of those receiving remittances are people who would ‘normally’ depend on the family members who send them remittances. Boxes 2 and 3 illustrate the simplest examples of remittances transfer: a nuclear family depending on the migrant husband/father; and an elderly woman depending on her adult children. In these cases, as with many of the survey respondents, the recipients are entirely dependent on remittances for income.


Box 1: Case study: Ubax17

Ubax 18 is in her late forties. She and her husband used to live in Mogadishu but left in 1991 after the state collapsed. One of her daughters was shot dead. She lives now with her husband and six of her children in Hargeisa. In the city, she has six married sisters, two married brothers and two brothers who are studying, and in the rural areas, she has two other siblings. When they arrived times were hard. She earned a living by making sweets that she sold outside the Madrassa, to children leaving school. She would take the money she made and go to the market to buy food for the family. Later she trained as a nurse. Her wage was fixed several years ago at 240,000 SlSh per month, then worth around USD 70, but now worth USD 36. Her husband is jobless. The rent of their house is USD 100 per month. To supplement her earnings, she also runs a small pharmacy business and performs operations privately, such as Sunna circumcision and the cutting of circumcised genitalia in preparation for marriage assists at home deliveries. She also founded a local women’s group. The family is doing fine. Her father had six wives and she has numerous half-siblings in the rural areas. When they are sick, they call her or they come and ask for treatment or help.

Her two brothers in the UK send her USD 100 each month specifically to pay for the education of the children: USD 46 for the two boys to attend private secondary school and USD 30 for her eldest son’s university fees. Sometimes they send more, for example at Ciid, for the children and for clothes.


Another key issue is how remittances are used. Where the household has other sources of income, in some cases, remittances are clearly allocated to specific expenses, often schooling, as illustrated in Box 1. However, in those households with no other sources of income, remittances may be treated as a fungible resource and not allocated to a single specific purpose. Respondents were prompted by a checklist of potential spending categories to specify whether or not they used remittance money for a series of particular purposes during the past 12 months. The results are shown in Figure 8. Recipients use remittances for a variety of purposes ranging from subsistence to business investments. It should be noted that the fact that some expenses, including important social events such as weddings, register rather low, does not necessarily mean that remittances are not used towards these ends: such expenses may be large but infrequent. Figure 8 provides a rough guide to regular spending items, rather than to volumes of spending.

When asked how the remittances were used, many respondents replied ‘it’s for daily living’ or ‘it’s barely enough’. The majority of respondents (96 per cent) were using remittances for biil (regular basic expenses that may include food, education, health, rent, qaad and household items). This fits in with evidence on other contexts (Durand et al., 1996; Black et al., 2003; Goldring, 2004; Maphosa, 2004; Citizens International and Bruks Associates, 2005; World Bank, 2006). Some of this went into education and health, which not only improves the social welfare of the household, but can also help its long-term economic prospects. Specifically, around two-thirds of respondents spent remittance money on education. The school attendance rate for 6–15 year olds among recipient households was substantially higher than estimates of general primary school attendance in Somaliland (37 per cent) (UNICEF, 2004). In higher cost forms of education (private primary and secondary schools and universities), remittances appear to play a particularly important role (Lindley, 2005). Around 60 per cent of respondents said that remittances had been spent on health expenses—regular medicine, doctor’s bills or emergency treatment. In-kind remittances of medicines are also important for some sick people: one young woman explained that her uncle in Finland regularly sends a drug to help with her father’s gastric problems, although he did not usually send money. This is particularly the case when the medicine required is not easily obtainable in Hargeisa or only a cheap version is available containing little active ingredient. Many people commented on the visibility of people with mental disorders and disabilities, linking this to war injuries and trauma, post-conflict educational and economic frustrations, and the lack of adequate medical attention. Remittances help support families burdened by the economic inactivity of a member through illness, as illustrated in Box 3 above.


Box 2: Case study: Yusuf

Yusuf is 19 years old and attends a public secondary school in Hargeisa. He was born in Djibouti and moved with his family to Hargeisa in the mid-1990s. He lives with his mother and seven siblings. All of the siblings are in public schools, the eldest is at university. The family’s main expenses are rent (USD 60 for three rooms), school fees, transport to school, food, electricity and water. He thinks that his family is doing alright. His father is in Djibouti most of the time, he works in shipping and sends Yusuf’s mother around USD 200 each month. This is the only source of regular income for the family. Sometimes an uncle who lives in Sweden also sends some money.


Box 3: Case study: Amina

Amina, an elderly widow, lives with her sick son in Hargeisa. Her daughters in Europe jointly funded the building of the house in which she and her son live. Her younger sister (in her fifties) often stays to help her look after him and the home in general. The daughters send regular remittances to pay for food and medicine when possible, but have not been able to find appropriate medical attention for their sick brother’s psychological condition. However, without their assistance, he would probably be destitute or dead. Amina is well over 60 years old, and her health is ailing: at this stage of her life, she has to rely on her family for support. The daughters would naturally look after both mother and son if they lived in Hargeisa; as they do not, they remit money instead to help care for her. The daughters pointed out that there is no pension, and no social services in Somaliland: Amina has no other source of income.


Figure 7: Comparison of economic activity of remittance recipients with the general population

70%
60%
50%
40%
30%
20%
10%
0

Remittance Urban Urban
recipient Somaliland Somalia
sample
Working Looking for work Not economically active
Sources: Academy for Peace and Development, 2004; UNDP and the World Bank, 2003

Chewing qaad, the stimulant leaves and stems of catha edulis, a tree or shrub cultivated in East Africa and the Arabian Peninsula, is very popular among Somali men—and also, increasingly, among some women (Carrier, 2005; Ibrahim Cadbi, 2005). Qaad is visible throughout Hargeisa, with trucks racing into Hargeisa in the morning with fresh supplies and numerous stalls lining the streets. According to Somaliland government figures, USD 300,000 a day is spent on qaad in Somaliland, mostly imported from Ethiopia. An average man might chew two bundles a day, perhaps with a drink, which might come to USD 100–150 per month, more than people spend on food. In the survey, only around 10 per cent of respondents said that remittances were used to buy qaad, and only around 25 per cent of male respondents. This was the subject of heated debate when I presented the survey results in a forum at the University of Hargeisa—people felt that in fact, given the prevalence of qaad chewing, most men receiving remittances would use some of the money for qaad. In the words of one participant, ‘Most men will chew on qaad if supported by women overseas’. The consensus was that there is something of a social stigma about men depending on others to sustain their consumption, which might lead respondents to deny that remittances are used for qaad.19

Remittances are also used by some—around one-fifth of survey respondents—to pay off debts. This refers generally to paying off debts owed to a local shop for food items. While shopkeepers no longer routinely allow customers to settle their account at the end of the month, if regular customers run out of money before then, the shopkeeper will often extend credit. Qaad sellers also often give credit to those unable to pay but eager to chew. In contrast to those paying off debts, far fewer people—less than five per cent of survey respondents— manage to save some of the remittances they receive. Many people do not have much money left over after satisfying the expenses outlined above. Moreover, savings facilities are limited—some deposit sums at the Central Bank or at a remittance company; others give money to shopkeepers or traders for safe keeping; and some people participate in hagbaad, a rotating fund utilised by groups of neighbourhood women or co-workers to save or borrow small amounts. However, in harmony with Islamic financial principles, which prohibit riba (interest), those with more substantial amounts of money may be more likely to invest it in business or property rather than accumulating cash in accounts.

This brings us to the investment of remittances in business, land and property, and emigration. While such investments may be infrequent (see Figure 9) they may be large and life changing. In our survey, of individual recipients surveyed, 13 per cent was self-employed, and of households receiving remittances, around 25 per cent derived income from some form of business or other assets that may or may not have been set up with assistance from relatives overseas. Moreover, those who have been able to establish a successful, self-sustaining business with diaspora support may be not appear in the sample as they may no longer be in receipt of regular remittances. Lump sums of money from a relative abroad may be invested in shops or taxi, bus or trucking businesses in Hargeisa. The emphasis on trade and service activities reflects the state of the urban economy as a whole: investment of remittances is constrained by the hardship of recipients and the local investment climate. Among British citizens who remit money, a common aspiration is to set up family members in a sustainable business. There are some success stories; there are also many accounts of businesses failing. However, in some cases, even without being directly invested, regular remittances can support the development of other income streams, by enabling families to cover their daily living expenses when they are starting up a business or to continue that business despite low profit margins (see Box 4 below for an example of this).

Figure 8: Frequency of uses of remittances

Biil                  0
Qaad                  100
Household items       200
Education expenses    300
Health expenses       400
Rent                  500
Land/house            600
Business investment
Pay debt
Social event
Pay for emigration
Savings
Compensation
Gift

Diaspora investment in land and housing is more clearly visible throughout Hargeisa. There was a lot of reconstruction work as people returned to destroyed homes, helped in some instances by migrant relatives. In more recent years, some migrants have also built their own homes. They often entrust the management of the process to a relative or visiting during their holidays to oversee the process. The home is an investment, in the medium term, as it is often rented out to NGOs or visiting disapora members, and in the long term, if people anticipate one day returning to live in Hargeisa. This demand has created a long boom in construction in Hargeisa, generating casual employment for people in the city and people coming from rural areas. While emigration of family members is often seen as an investment, and is in fact very often financed by relatives in the diaspora. However, this is frequently arranged from abroad rather by sending money to the prospective emigrant. It is also true that emigration opportunities have narrowed as in general people from Hargeisa are no longer eligible for asylum in the West.

Furthermore, several purposes for which remittances are sometimes used have the effect of reinforcing or extending existing social networks or redistributing the benefits of migration. A few recipients had spent remittances on important social and religious events. Several interviewees’ households had held weddings in the past couple of years that were largely funded by brothers or sisters abroad, indeed, in one case, the groom himself lives in the UK. In many ways, qaad is also seen as a social network investment as it is generally chewed by men while relaxing in groups. Other examples are explored in the next section. In sum, remittances can affect household roles and relationships. For many recipients remittances are central to their livelihood. Remittances may both respond to need and in combination with high local unemployment, entrench dependency, as well as in some instances improving local livelihoods. Remittances are used largely for subsistence. Sometimes remittances are invested in business and property. Remittances are also ‘invested’ in human capital and social networks, which also affect the social welfare and economic prospects of the household.

3.6. Remittances in crises: some insights

Theoretical approaches that conceptualise migration as a household livelihood strategy have suggested that anticipated remittances are a key part of the decision that a household member should migrate. In one of several theoretical models developed to describe this relationship, remittances indicate an implicit co-insurance agreement, between migrant and nonmigrant family members (Stark and Bloom, 1985; Schrieder and Knerr, 2000; Gubert, 2002). This model posits that the original household has income that is vulnerable to risks and crises, such as crop failure, livestock diseases, price fluctuations, land tenancy insecurity, and therefore benefits from diversification of income sources by allocating one member to work abroad.

Conversely, the migrant faces risks in entering a new country and a new labour market, which means that he/she benefits from being able to rely on family support in the event of unemployment, illness or deportation (Stark and Bloom, 1985). In this model, one would expect remittances to increase in response to crises experienced by household members back home, smoothing their income and consumption. There are clear problems with applying this functionalist microeconomic approach to understanding the dynamics of labour migration to the very different contexts of forced migration from Hargeisa, in the late 1980s and subsequently. Here, the act of emigration, and the subsequent act of remitting money, may not be so integrally linked (Lindley, forthcoming). However, the research in Hargeisa suggests that the sending of remittances may nevertheless respond to contexts of crisis.

It is possible to identify five main types of crisis affecting the population of Hargeisa in the past two decades. First, there is outright conflict in 1988–91 and again in 1994–96. Second, there is the ongoing crisis of state institutions in the sense that the Republic of Somaliland is not recognised as a state by the international community and hence is unable to take advantage of the privileges of statehood, with implications for the people of Hargeisa, its capital city (WSP International, 2005). Third, as the pastoral and urban economies are closely connected, disruption to the livestock trade affects city life, directly and indirectly: ‘Droughts can trigger food shortages, the forced liquidation of livestock at depressed prices, migration to urban centres or camps, and the loss of human life from starvation or social conflict’ (Xasan Gaani, 2005, p. 203).

The import ban on Somali livestock by Saudi Arabia in 1998–99, and by all the Gulf States in 2000 (because of fears of Rift Valley Fever among Somali livestock), also had consequences for the Hargeisa economy (King, 2003). Fourth, exchange-rate depreciations related to the above events affect Hargeisa residents: unless wages keep pace, imports can become too expensive. The Somaliland shilling, introduced in 1994, depreciated particularly steeply against the US dollar when the government printed large quantities of new currency to finance the war effort in the mid-1990s; again in 2002–03, when the government printed new currency apparently to finance the electoral campaign; and after September 2000, when the Gulf States banned the import of Somali livestock (King, 2003; Xasan Gaani, 2005). Fifth, there are the personal crises of people’s everyday lives— displacement and death, livelihood collapse, illness and family break up—often closely linked to the wider political and economic contexts in the city in which they live (King, 2003).

In the absence of reliable longitudinal macro- or micro-level data on remittance volumes, any study of the impact of these different types of crises on remittances remains limited. Below, I assess this claim in relation to relevant qualitative and limited quantitative evidence available. The survey provided some interesting data on the timing of emigration and remittances. As shown in Figure 9, emigration increased in the 1980s, and peaked at the turn of the decade, when the civil war in the north was at its height, and subsequently in the mid-1990s, a time of civil strife in Hargeisa, and in 1999, at the time of the first Saudi livestock export ban (1998–99). Thus, it is clear that the majority of people who today remit money to Hargeisa left Somalia in particularly large numbers at times of crisis, and throughout the politically uncertain period of the 1990s. The time lag between these migrants arriving in their current country of residence and beginning to send money to Hargeisa varies, as show in Figure 10. On average, they began to send money after around six years, but 40 per cent started within two years of arriving. While current remittance senders began sending money after 1988, there was a significant peak in 2000–03 (see Figure 11). Indeed more than 40 per cent of these remittance senders started sending during that short three-year period.

There are several likely or possible explanations for this marked ‘triggering’ of remittance activity in 2000–03, which might be explored in subsequent research. The first is based on emigration patterns: the large numbers of people who left Somalia in the early 1990s are becoming more socioeconomically settled in their current countries of residence and hence more able to send remittances. Moreover, migrants who left and began remitting earlier may no longer do so and therefore not appear in the sample. The second explanation is that migrants abroad began to remit in this period in response to events in Hargeisa. The most obvious potential trigger is the second Gulf livestock import ban, which deprived Somali pastoralists of the most important markets for their livestock (since then the ban has been lifted by a number of Gulf States, but not yet by Saudi Arabia, previously the largest importer).

Remittance companies report an increase in the volume of remittances received during the livestock ban (consultations; also see Xasan Gaani, 2005). Those most directly affected include merchants and all those involved in the commodity chain relating to the trade in livestock. However, the economic slow down also affected general business in Hargeisa and government bodies, which lost fiscal revenue. The qualitative research suggests that those who suffered because of this who had relatives abroad tried to obtain assistance from them, often successfully. A third explanation relates to inward migration to Hargeisa, with the resettlement of people displaced during the conflict of the mid-1990s, and increasing urbanisation encouraged by the difficulties of pastoral life (exacerbated by environmental degradation and drought) and attracted by casual labour opportunities (Ibrahim Cadbi, 2005; World Bank, 2005). It is possible that some of these people then managed to tap family support. Lastly, the consolidation of democracy in Somaliland since 2001 may have encouraged migrants to begin sending money for other reasons, reinvesting in family networks and livelihoods in anticipation of returning, signalling reengagement with their place of origin. As stated above, these are all possible explanations that might be explored in further research, given more detailed data.


Figure 10: Time lag between migrant arriving in current
country of residence and beginning to send money

0
2
4
6
8
10
12
14
16
18
20
22
25
28
37
0 20 40 60 80 100
Years between arrival and beginningto remit

Figure 11: Year remittance senders begin to remit
1970
1976
1982
1985
1988
1991
1994
1997
2000
2003
0 20 40 60 80
Years between arrival and beginningto remit

Figure 9: Year of emigration of remittance senders
1950
1972
1977
1982
1987
1992
1997
2002
0 10 20 30 40 50 60
Year remittance sender left Somalia
At the micro level, asking migrant relatives for assistance emerges as one of several mechanisms for coping with income shocks (whether a loss of income or the emergence of a large necessary expense) in Hargeisa. If successful, there may be three types of responses: an increase in the amount remitted, for a period or permanently; the initiation of regular remittance transfers for subsistence, for a period or permanently; or the transfer of a one-off amount to meet a particular expense or to aid coping during a particularly difficult period.

Remittances often do smooth the income and expenditure of recipient households during income shocks, as illustrated in Box 4 below. In this scenario, remittances were not directly invested in the livelihood/business, but by providing family with subsistence expenses, they smoothed consumption during a difficult period and relieved pressure on business profit margins during earlier stages or tricky periods. For other interviewees, remittances facilitated livelihoods in the refugee camps, as well as the return home, and met the needs of otherwise vulnerable families until dependent children are able to earn a living. While remittances may help to deal with income shocks, lack of remittances would be a serious income shock for many households. Moreover, regular remittances very effectively protect households against local currency depreciations as they are received in hard currency (then converted into local currency), meaning that remittance recipients are relatively better able to afford imported goods (which include basic goods, such as cooking oil, rice, sugar, tea and wheat) during times of depreciation. This may somewhat mitigate downturns in trade in Hargeisa.

Somali society is renowned for the strength of the family and clan-based mutual support mechanisms, particularly among the dominant and traditionally pastoralist clans. Many migrants and recipients are still tied to these relationships. Of this sample of people receiving remittances, around 15 per cent said that some was spent on gifts to family, neighbours or friends in Somaliland. This may under-represent the degree to which remittances are redistributed to rural relatives and neighbouring households by receiving households, as it often occurs in informal, very commonplace ways. For example, sharing food with poorer neighbours—indeed, with anyone who turns up around lunchtime—is extremely common. Within urban neighbourhoods, better off households quite often will assist poorer relatives by providing money, food and somewhere to stay. Moreover, one family of parents and siblings may have members in the rural area, Hargeisa, and abroad, with different layers of transfer activity. One survey found that around 50 per cent of urban households in Hargeisa (excluding the poor ‘resettlement’ areas) support relations in pastoralist areas through monthly contributions of USD 10–100 (Medani, 2000). While it appears that rural people receive remittances more rarely, during hard times, such as drought, they solicit assistance in the form of money, hospitality and essential items, from urban households and migrants alike, as illustrated in Box 5. Thus, not only are transfers made from the diaspora to Hargeisa, but also money and resources are transferred from the better-off to poorer households, and from urban to rural households. City dwellers that do receive remittances often are in a better position to help other relatives during periods of drought or an economic downturn. Thus, even when remittances are not targeted at the poorest and most needy people in the community, they may nevertheless provide a degree of social protection.

More institutionalised social support mechanisms, as highlighted in the previous section, also help to deal with crises. Sako (Zakaat) redistributes income as a matter of religious duty from the diaspora to needy relatives in Hargeisa, and from better-off people to others within and beyond Hargeisa. Qaraan (the clan-based collection) is often used to assist needy individuals or those undergoing a particular problem (for instance, payment for an urgent, expensive operation) and may be collected in the diaspora as well as within Somali society. Jaaliyaad is a broader-based community fundraising mechanism, to which migrants contribute, for example, donations to health and educational institutions and projects (Kent et al., 2004; Lindley, 2005). In addition, around one-fifth of respondents said that remittances had been spent on magta or diya, clan-based compensation for injury or death (Lewis, 1961). Such compensation is a traditional mechanism for resolving crises, and the amount due is divided by the elders among the male relatives in a diya-paying group (a small kinship group).


Box 4: Case study: Nimco

Nimco was born in Hargeisa, but used to live in Mogadishu with her husband and eight children. They moved to Hargeisa for safety in 1991. Her husband later left her, but she remains in the house that he owns. She lives with nine others: three sons, a daughter-in-law, four grandchildren and a nephew.

Two sons are jobless: one is a mason and can earn USD 10 per day on a building site, but only get a few days’ work each month. Her eldest son was a merchant, trading sheep and goats. He used to provide for the entire family, but was forced out of business in 2000 by the Saudi livestock import ban. He then started a general shop selling food and non-food items, but this was going badly and the family was struggling. One of Nimco’s daughters had been in the UK at that point for five years and sent USD 100 or 200 now and again, but said that she could not afford to send more often. In response to the family’s predicament, Nimco telephoned her daughter in the UK and said that she needed help.

Now, the shop generates about USD 6 per day for the household, which it uses to buy food for the family. The daughter in the UK is the other main source of income for the household, sending Nimco USD 300 every other month. This pays for the children’s schooling (about USD 40 per month), water and electricity (approximately USD 30 per month), clothes and other items.

Survey data and interviews suggest that, in a marked departure from traditional practice, men’s are sometimes paid using income from local and diaspora women.


Box 5: Rural relatives

One interviewee lives with his mother and brothers and the household receives regular remittances from his sister in the UK. When there is drought and times are hard in the rural area, his uncle often comes to the city. One of the first things the uncle does is call the interviewee’s sister, to tell her there is no rain and to ask for help. She usually sends some money, but the interviewee thinks that it is a source of worry to her. The interviewee also pointed out that the rural relatives also rely on the family in the town: ‘I get asked frequently for help. They take my clothes. They say can you give us something. I say that I don’t have anything to give. They say, at least give me a T-shirt, trousers, boots.… There is a culture that the people who come from the bush do not pay. Also, if you go there, you get good hospitality’.


Migration and remittance transfers have long been a feature of Somali society. Before the war, there were considerable numbers of migrant workers living in the Gulf States and contributing in important ways to the economies of Somalia’s cities. However, the civil war and the subsequent crisis of state institutions generated much greater emigration and created new pathways to the West. However, data on the scale and effects of these remittances are still limited.

The research conducted in Hargeisa paints a broadly indicative picture of the remittance inflow and allocation in the city. Remittances most commonly originated in the Middle East, North America and the UK. Numerous relatives support residents of Hargeisa, with brothers and sisters appearing to play a particularly important role. The most common benefactors of remittances in Hargeisa are women with children, often living in extended family households. Households receiving regular remittances tend to be in the average or better-off segments of Hargeisa society and from large clans rather than minority groups. Several effects on recipient households were observed.

First, remittances can change household relationships and roles depending on who receives the money. Remittances may be seen as an aspect of a broader transformation in gender and household relations in Somali society since the conflict began, which has seen women play an increasingly important economic role in the welfare of their families. Second, around one-half of the households surveyed were entirely reliant on remittances. In different cases, remittances respond to need, generate longerterm dependency or encourage the development of local livelihoods. However, the extent of these effects is difficult to determine. People use remittances for a range of purposes, from subsistence to investment in social networks and human capital, to investment in property and business.

The process of emigration and remittance transfer responds in certain ways to the contexts of crisis in Somali society. The survey data suggest that people have emigrated in the greatest numbers during times of political strife and uncertainty in Hargeisa and that a decline in fortunes of residents can trigger remittances. Asking migrants for assistance is one of several coping mechanisms utilised during difficult times. Remittances may sustain a household after the collapse of their livelihood, prop up an existing livelihood, pay for essential medical treatment, et cetera.

People receiving regular remittances are protected from exchange-rate fluctuations and are often better able to assist relatives in rural areas in times of crisis. More institutionalised social support mechanisms also operate within Hargeisa, that go some way towards helping the poor and the needy, and migrants and remittance recipients sometimes participate in these mechanisms. All of this underscores the importance, in the context of increased global financial regulation, of keeping Somali remittance channels open.

Finally, a neglected dimension to the discussion of remittances is the burden it can place on migrants struggling to make their way in a new country. In some cases, income shocks to households in Hargeisa are converted into substantial financial stress for migrants. While the migrants in the West may receive wages or welfare payments beyond the wildest dreams of most Hargeisa residents, disposable income after paying for accommodation, food and other items may be very small. What begins as an income shock for a household in the Horn of Africa can translate into a financial stress for a factory worker in Minneapolis, a local council officer or bus driver in London or a care worker in Italy, eating into overdrafts and leading to long hours of overtime, and ultimately meaning that some people struggle to get by. The impact of remittances on senders in the diaspora, and the sustainability of remittance transfers at their current levels, remain unexplored.

NOTES

1 This paper uses the term ‘Somalia’ to specify the Republic of Somalia, as of 1960, and ‘Somaliland’ to denote the self-declared independent Republic of Somaliland as of 1991.
2 This section draws on previous work by the author and her forthcoming DPhil thesis.
3 The main clan families with a pastoral tradition are Darod, Dir, Hawiye and Isaq; other clans with a more agricultural tradition are known as the Digil and Mirifle or the Reewin.
4 That is, less than one US dollar purchasing power parity (PPP) per capita per day.
6 His estimations are based on: official numbers of Somali people in host countries; estimates of numbers of households; public welfare payments in each country; and estimates of proportion of income saved and remitted from interviews. Unfortunately, the estimates of numbers of households and average remittances for each country are not made available to the reader.
7 Note that this may not include aid from Arab countries.
8 Quote from her speech at the Somaliland Recognition Debate, Portcullis House, London, 17 March 2004.
9 Clans of the Isaq family predominate in Somaliland. settlements of internally displaced persons (IDPs) and
11 The piloting process was limited to piloting with other researchers and contacts and in a role-play session involving the interviewers themselves. As a result, there were a couple of questions that did not ‘work’ on the first day, which were discontinued or altered from the second day onwards. This was taken into account in the data analysis.
12 Interviews with remittance company officials, March and August 2005.
13 Between 1988 and 1994, UK citizens could apply for family reunion visas for relatives in Ethiopia, rather than relatives applying through the UK embassy in Addis (Griffiths, 2002).
14 Source of data on urban Somaliland: Academy for Peace and Development, 2004.
15 This section draws on interviews with a women’s literacy worker, with a minority group community leader, with staff at the Ministry of Reintegration, Rehabilitation and Resettlement, and field notes, March and August 2005.
17 This section draws on interviews with a women’s literacy worker, with a minority group community leader, with staff at the Ministry of Reintegration, Rehabilitation and Resettlement, and field notes, March and August 2005.
18 All names have been changed and some details altered to protect the identity of interviewees.
19 I am grateful to Mohamed Ibrahim of the Academy for Peace and Development for raising this issue.

References

- Abby, A. and D. Mahamud (2004) The Position of Minorities in Somalia. Oxford House in Bethnal Green: London. http://www. oxfordhouse.org.uk.
- Academy for Peace and Development (2004) A Socio-economic Survey of Somaliland. Draft paper.
- Adams, L. et al. (2003) A Growing Economy amidst Insecurity. A Study of Urban Livelihoods in Belet Weyne. Hiran Region – Somalia.
- Food Security Analysis Unit (FSAU) and Save the Children (UK) Food Security Project, Nairobi.
- Ahmed, I. I. (2000) ‘Remittances and Their Economic Impact in Postwar Somaliland’. Disasters. 24(4). pp. 380–389.
- Al-Sharmani, M. (2004) ‘Refugee Livelihoods—Livelihoods and diasporic identity constructions of Somali refugees in Cairo’. New Issues in Refugee Research Working Paper. No. 104. Evaluation and Policy Analysis Unit, United Nations High Commissioner for Refugees (UNHCR), Geneva.
- Black, R. et al. (2003) International Migration, remittances and poverty: the case of Ghana and Côte d’Ivoire. Paper prepared for the World Bank.
- Bradbury, M. (2002) Somalia: The Aftermath of September 11th and the War on Terrorism. Unpublished report for Oxfam.
- Carrier, N. (2005) ‘The Need for Speed: Contrasting Timeframes in the Social Life of Kenyan Miraa’. Africa. 75(4). pp. 539–558.
- Citizens International and Bruks Associates (2005) Private Remittance Flows to Ghana: Project Country Report. United States Agency for International Development (USAID), Washington, DC.
- Durand, J. et al. (1996) ‘Migradollars and Development: A Reconsideration of the Mexican Case’. International Migration Review. 30(2). pp 423–444.
- El-Solh, C.F. (1991) ‘Somalis in London’s East End: a community striving for recognition’. New Community. 17(4). pp. 539–552.
- Goldring, L. (2004) ‘Family and Collective Remittances to Mexico: A Multi-dimensional Typology’. Development and Change. 35(4). pp. 799–840.
- Greater London Authority (2005) ‘Country of Birth and Labour Market Outcomes in London: An analysis of Labour Force Survey and Census data’. DMAG Briefing. 2005/1. Greater London Authority, London.
- Griffiths, D. J. (2002) Somali and Kurdish Refugees in London – New Identities in the Diaspora. Ashgate, Aldershot.
- Gubert, F. (2002) ‘Do migrants insure those who stay behind? Evidence from the Kayes Area (Western Mali)’. Oxford Development Studies. 30(3). pp. 267–287.
- Gundel, J. (2003) ‘The Migration–Development Nexus: Somalia Case Study’. In N. Van Hear and N. Nyberg Sørensen (eds) The Migration-Development Nexus. International Organization for Migration (IOM), Geneva. pp. 233–258.
- Horst, C. (2003) Transnational Nomads—How Somalis cope with refugee life in the Dadaab camps of Kenya. PhD thesis, Amsterdam Research Institute for Global Issues and Development Studies, University of Amsterdam.
- Human Rights Watch (2002) Hidden in Plain View: Refugees Living Without Protection in Nairobi and Kampala. Human Rights Watch, New York, NY.
- Hussein, S. (1993) ‘Somalis in London’. In N. Merriman (ed.) The Peopling of London. Museum of London, London. pp. 163–168.
- Ibrahim Cadbi, S. (2005) ‘The Impact of the War on the Family’. In WSP International Rebuilding Somaliland: Issues and possibilities. The Red Sea Press, Asmara.
- Jama, M.A. (1997) ‘The Destruction of the Somali State: Causes, Costs and Lessons’. In H.M. Adan and R. Ford (eds.) Mending Rips in the Sky. The Red Sea Press, Lawrenceville, NJ. pp. 237–254.
- Jamal, V. (1988) ‘Somalia: Understanding an Unconventional Economy’. Development and Change. 19. pp. 203–265.
- Kent, R. et al. (2004) Social Facilitation, Development and the Diaspora: Support for Sustainable Health Services in Somalia. Report funded by USAID, London.
- King, A. (2003) Hargeisa Urban Household Economy Assessment February—March 2003. Famine Early Warning Systems Network (FEWS NET), Hargeisa.
- Lewis, I. (1994) Blood and Bone: The Call of Kinship in Somali Society. The Red Sea Press, Lawrenceville, NJ.
- Lewis, I.M. (1980) A Modern History Of Somalia—Nation And State In The Horn of Africa. Longman, London.
- Lindley, A. (2005) ‘The Influence of Remittances and Diaspora Donations on Education’. Paper presented at a conference on Somali remittances organised by the United Nations Development Programme (UNDP) and the World Bank, Washington, DC, 1–2 December.
- Lindley, A. (forthcoming) The Dynamics, Effects and Politics of Migrant Remittances in Insecure Contexts: The Somali Case. DPhil thesis, Department of Development Studies, University of Oxford (to be submitted in 2006).
- Maphosa, F. (2004) The Impact of Remittances from Zimbabweans Working in South Africa on Rural Livelihoods in the Southern Districts of Zimbabwe. A summary of a report submitted to the Council for the Development of Social Science Research in Africa (CODESRIA).
- Marchal, R. et al. (2000) Globalization and its impact on Somalia. UNDP and the United Nations Development Office for Somalia (UNDOS), Nairobi.
- Medani, K. (2000) Report in internal migration and remittance inflows: Northwest and Northeast Somalia. UN Coordination Unit and the FSAU, Nairobi.
- Menkhaus, K. (2001) Remittance companies and money transfers in Somalia. Unpublished briefing paper. October.
- Minority Rights Group International (1997) ‘Minority Children of Somalia’. War: The Impact on Minority and Indigenous Children. Minority Rights Group International, London.
- Mohammed, A.A. (2000) The Role of International Remittances on the Economy in Somalia. MPhil thesis submitted to the Department of Economics, University of Oslo.
- Omer, A. (2002) Supporting systems and procedures for the effective regulation and monitoring of Somali remittance companies (Hawala). UNDP Somalia, Nairobi.
- Omer, A and G. El Koury (2005). ‘Regulation and Supervision in a Vacuum: The Story of the Somali Remittance Sector’. In S. M. Maimbo and D. Ratha (eds) Remittances: Development Impact and Future Prospects. World Bank, Washington, DC.
- Pérouse de Montclos, M.-A. (2003) ‘A Refugee Diaspora—When the Somalis Go West’. In K. Koser (ed.) New African Diasporas. Routledge, London. pp. 37–55.
- Schrieder, G. and B. Knerr (2000) ‘Labour Migration as a Social Security Mechanism for Smallholder Households in Sub-Saharan Africa: The Case of Cameroon’. Oxford Development Studies. 28(2). pp. 223–236.
- Sheikh Ali, A. (1997) Remittances in Somalia. Unpublished report. UNDOS, Nairobi.
- Shire, S.A. (2004) ‘Transactions with Homeland: Remittance’. Bildhaan. 4. pp. 92–103.
- Simons, A. (1995) Networks of Dissolution: Somalia Undone. Westview Press, Oxford.
- Somalia Aid Co-ordination Body (2003) Donor Report 2003. Somalia Aid Co-ordination Body, Nairobi.
- Sørensen, N.N. (2004) ‘Opportunities and Pitfalls in the Migration–Development Nexus: Somaliland and Beyond’. DIIS Working Paper. 2004/21. Danish Institute for International Studies, Copenhagen.
- Stark, O. and D. E. Bloom (1985) ‘The New Economics of Labor Migration’. The American Economic Review. 75(2). pp. 173–178. UNDP (United Nations Development Programme) (2001) Human Development Report 2001 Somalia. UNDP Somalia, Nairobi.
- UNDP Somalia and the World Bank (2003) Socio-Economic Survey 2002 Somalia. UNDP Somalia and the World Bank, Nairobi.
- UNHCR (United Nations High Commissioner for Human Rights) (2005) 2004 Global Refugee Trends: Overview of Refugee Populations, New Arrivals, Durable Solutions, Asylum-Seekers, Stateless Persons and Other People of Concern to UNHCR. UNHCR, Geneva. http://www.unhcr.org.
- UNICEF (United Nations Children’s Fund) (2004) UNICEF Survey of Somalia’s Primary Schools 2003–04. UNICEF, Nairobi.
- UNIFEM (United Nations Development Fund for Women) (1998) Somalia between peace and war: Somali women on the eve of the 21st century. UNIFEM, Nairobi.
- World Bank (2005) Somalia: From Resilience Towards Recovery and Development. A Country Economic Memorandum for Somalia. Report No. 34356-SO. World Bank, Washington, DC.
- World Bank (2006) Global Economic Prospects. Economic Implications of Remittances and Migration. World Bank, Washington, DC. WSP International (2005) Rebuilding Somaliland: Issues and possibilities. The Red Sea Press, Asmara.
- Xasan Gaani, M. (2005) ‘Regulating the Livestock Economy of Somaliland’. In WSP International Rebuilding Somaliland: Issues and possibilities. The Red Sea Press, Asmara.