Migration is not a new phenomenon. From time immemorial, human beings have left their homes in search for better living conditions or in fear of persecution. Yet, over the past few decades, technological innovation has changed important aspects of migration. Cheaper and more rapid transport and better means of communication (mobile phones, landlines, IT, etc.) have facilitated the creation of transnational networks, by helping maintain ties between migrants and their family back home as well as between migrants in different countries (transnational communities). Recently, researchers, practitioners and policy makers have paid greater attention to the contribution that these networks provide to the development of the home countries. Although land is a crucial asset for rural livelihoods, so far it has not featured very high in this debate.
This chapter outlines a conceptual framework to explore the linkages between international remittances and access to land in the home countries, by placing them within the broader context of the migration and development debate and of household livelihood strategies; and by analysing the different forms that those linkages can take and the key issues that they raise. It draws on a review of the relevant literature from across the world and on conversations with key informants that the author met while travelling to West Africa on other assignments (from migrants to migrants family members, from researchers to NGO officials).
In this study, international migration is broadly defined as the movement of persons from one country to another (home and destination country, respectively), although as stated above this study focuses on South-North migration only. Migration can take very different forms, depending on who is migrating (individuals or entire households), the reasons to migrate (better living conditions, conflict, etc), the duration of the stay abroad (seasonal, short term or long term), and so on. Remittances are the monies that migrants - whether individual migrants or migrants associations - send back to their home country. They can be sent through formal channels, such as banks, post offices and money transfer companies, or through informal mechanisms (by hand, family and friends). In addition to cash payments, transfers of goods (second-hand clothes, vehicles, etc) are also common. Besides remittances strictu senso, this study also considers other types of money transfers relating to employment abroad, particularly pensions.
Land tenure is the system of rights, rules, institutions and processes under which land is held, managed, used and transacted. In many developing countries, several land tenure systems - formal/statutory, informal/customary or combinations of both - coexist in the same territory (legal pluralism). This study primarily focuses on land access issues, and takes into account both formal and informal systems (e.g. informal land purchases which are not recognised by law).
For a long time, migration and development have been two separate policy areas. In destination countries, migration authorities focus on controlling migration flows, while development agencies mainly work in the home countries, with little coordination existing between the two (Van Hear, 2003). In recent years, however, some home country governments have explored ways to harness resources from the diaspora to promote development. New institutions have been created to maintain ties with the diaspora, such as Mexicos National Council for Mexican Communities Abroad, and a variety of policy measures (e.g. tax breaks) have been adopted or discussed.
While migration and development policies have evolved in a largely parallel way, in the literature a longstanding debate on the relationship between migration and development has opposed optimists and pessimists (de Haas, 2003). Optimists emphasise the benefits arising from remittances, which for many countries are a precious source of foreign exchange as well as a major component of household income. Second-hand goods, spare parts, vehicles and other items sent by migrants to their home countries support their relatives businesses. Moreover, in some places, community-based organizations such as hometown associations channel part of the remittance inflows into community development projects, such as schools, health centres and wells. Returned migrants bring back knowledge, ideas, skills and experience (Ammassari, 2003). Furthermore, anecdotal evidence suggests that, where decentralization processes have transferred responsibilities to elected local governments, migrants have played an important role in helping mobilise resources for local development, for instance by facilitating initiatives between local governments in destination and home countries (for examples from Senegal, see below, Chapter 3, Box 7).
On the other hand, pessimists - particularly fashionable in the 1970s - stress the problems created by substantial out-migration. First, migration may entail the loss of scarce skilled labour (brain drain). Moreover, in agriculture-based economies where a significant portion of the adult population is away, labour shortages may make the local economy highly dependent on remittances, raising concerns for its sustainability in the longer term. This is especially the case for long-term and long-distance migration, as migrants moving within a region (for example, between the Sahel and the coastal countries of West Africa) are usually more able to return for limited periods of time linked to the agricultural calendar. Secondly, migrants are not usually the poorest, since moving requires significant financial resources and social networks; moreover, migration may further deepen inequality, as households receiving remittances are able to buy land and other key resources, set up new businesses and improve their childrens education, while households not benefiting from this source of income may see their access to such resources and livelihood strategies reduced.
Beyond economics and its optimists - pessimists divide, migration is a factor for social and cultural change, as movement facilitates exchange of cultural values and social behaviour patterns, and as absence from and remittances to the family back home foster redefinitions of values, roles and social hierarchies (e.g. along gender, age and class lines; see below).
The global volume of remittances is widely acknowledged to be huge, even if the figure cannot be determined with any accuracy because untold numbers of migrants avoid the formal banking services. For this reason, available figures are likely to greatly underestimate the size of remittance flows. The World Bank (2003) estimates that in 2002 remittances amounted to $80 billion, nearly double the level of aid-related flows ($49 billion) and second only to FDI (some $143 billion) as a source of external funding for developing countries.
Regionally, these amounts are not evenly distributed, and in 2002 remittances were estimated to $25 billion in Latin America and the Caribbean, $16 billion in South Asia, $14 billion in Middle East and North Africa and $4 billion in sub-Saharan Africa (World Bank, 2003; see below, Figure 2). In 2001, Mexico received $9.9 billion in remittances, India $10 billion, Philippines $6.4 billion, Morocco $3.3 billion, Turkey $2.8 billion and Senegal $0.2 billion (World Bank, 2003; see below, Figure 3). These figures would need to be adjusted for GDP and population sizes in order to reflect the relative importance of remittances. Indeed, as a share of GDP, remittances are significantly higher in lower-income countries than in other developing countries (Ratha, 2003; see below, Figure 4). Top destination areas from which remittances are sent include North America (United States), Western Europe (Germany, Belgium, Switzerland, France and Italy) and the Arabian Gulf (Saudi Arabia) (World Bank, 2003; see below, Figure 5).
Besides these static figures, it is important to note the growing importance and relative stability of remittances over time. During the 1990s, while official aid decreased (from $49.5 billion in 1991 to $41.6 billion in 2000), remittances almost doubled (from $33 billion in 1991 to $65 billion in 1999; Gammeltoft, 2002, using data from the IMF). Moreover, while capital flows are highly volatile depending on economic cycles, remittance flows are remarkably stable over time (Ratha, 2003).
For many migrants, sending remittances to support their family is a moral and social obligation. Remittances also enhance migrants social status, and enable them to maintain a foothold in their home area, so that they will be welcome upon their return (Tacoli, 2002). Factors affecting the amount and frequency of remittances include the level of migrants earnings, migrants legal position in the destination country (regular/irregular), the length of the stay abroad, migrants marital/family status (e.g. whether he/she has children in the destination country), migrants desire to return, exchange rates, political risk, access to facilities for transferring funds, and the nature of the relationship between the migrant and family back home (van Doorn, n.d.; Kabki et al, 2003).
Figure 1. Workers remittances received by all developing countries (1995-2002)
Source: World Bank 2003
Figure 2. Regional shares of workers remittances received by developing countries (2002 estimates)
Source: World Bank 2003
Figure 3. Top 10 developing-country recipients of workers remittances in billions of dollars (2001)
Source: IMF, Balance of Payments Yearbook; World Bank, 2003
Figure 4. Top 10 developing-country recipients of workers remittances as percentage of GDP (2001)
Source: IMF, Balance of Payments Yearbook; World Bank 2003
Figure 5. Top ten sources of remittances, 2001
Source: IMF, Balance of Payments Yearbook; World Bank 2003
Remittances and development linkages may be explored at three levels. At the macro level, remittances are a precious source of foreign exchange, and as such inject capital in the economy and may help reduce balance of payments deficits.
At the meso level, many communities benefit from development projects initiated and funded by associations of migrants overseas. Typical examples include the construction of schools, health centres, religious buildings, wells and irrigation schemes. Migrants associations usually work on the basis of quotas paid by their members, and mobilise additional funding through NGOs and other development agencies. Although these associations are not specific to any particular nationality, some ethnic groups, such as the Soninké of Mali and Senegal, are known to have a stronger tradition than others. Problems experienced in the past and reported in the literature include different perceptions of needs and priorities between migrants and home communities (FASTI, 1991; Yatera, 1997; Lavigne Delville, 1991; Smith and Mazzucato, 2003; Tacoli, 2002; Okali et al, 2001; Sander, 2003).
At the micro level, remittances are a major source of support for households livelihoods. Unlike aid, remittances flow directly to individual households and unlike loans they incur no debt. Besides contributing to consumption in the short term (by enabling to pay for food, clothes, healthcare, etc), remittances can foster longer-term development through investment in education, land and small businesses. Among the different aspects of the remittances-development relationship, this study focuses on these micro, household-level dynamics.
Box 1. Migration and remittances in international legal instruments
Given migrants invaluable contribution to their home countries and their often vulnerable position in destination countries, a range of international instruments aims to protect migrants rights and to promote their development contribution.
The International Convention on the Protection of the Rights of All Migrants Workers and Members of Their Families was adopted in 1990 and entered into force in 2003. It protects a range of rights for legal (documented) migrants as well as some basic rights for illegal migrants. Among other things, the Convention protects the right of legal migrants to send remittances, and requires destination country governments to facilitate such transfers (article 47). Moreover, legal and illegal migrants have the right to transfer earnings and savings upon termination of their stay in the destination country (article 32). However, the Convention has only been ratified by a very small number of states (20 in 2003), and although some major home countries have ratified it (including Senegal and Ghana), no destination country has done so as yet.
Moreover, in 2003, within the context of the WTO/GATS negotiations, several developing countries filed a proposal for liberalization of Mode 4 service provision, aimed at easing entry restrictions on temporary movement of natural persons through, among others: a clearer differentiation between permanent immigration, to remain subject to normal immigration rules and procedures, and temporary movement of service suppliers, to be liberalised under the GATS; and greater recognition of professional qualifications (WTO document TN/S/W/14, 3 July 2003).
Steps have also been taken at the bilateral level. For instance, France has signed a series of bilateral treaties with Francophone African countries, concerning entry requirements and procedures (Conventions rélatives à la circulation et au sejour des personnes), migrants rights (Conventions détablissement) and co-development. The Co-development Convention between Senegal and France (2000) aims to harness skills and resources from Senegalese migrants in France to promote development in Senegal; to support the professional integration in Senegal of Senegalese students in France; and to jointly manage migration flows (Plancade, 2002). An agreement with similar aims was signed in 2000 between Mali and France (Martin et al, 2002).
It is well known that livelihood strategies in developing countries are highly diversified and build on a range of assets and activities. The sustainable livelihoods literature identifies five broad types of capital assets - human, social, natural, physical and financial capital (Chambers and Conway, 1992; Ashley and Carney, 1999; Carney, 2002). In this context, remittances constitute a specific type of financial capital, alongside e.g. wages, pensions and credit, and in many areas account for a substantial share of household income. In Bangladesh, for instance, Siddiqui and Abrar (2001) found that remittances accounted for 51 percent of households total income. Similarly, in the Mexican state of Guanajuato, 45 percent of households depend on remittances from the US as their main source of income (Sander, 2003). These resources contribute to household livelihoods both by increasing the level of income and wealth and by diversifying income sources, thereby helping manage risk (de Haas, 2003).
As cash income, international remittances present strong similarities with other forms of financial capital (wages, pensions and internal remittances). Specificities nevertheless exist in terms of both nature and scale. First, the amount of money involved is usually much higher than for incomes earned in-country. This broadens the portfolio of possible uses to capital-demanding projects such as house building and land purchases. Secondly, international remittances come with a variety of other assets that migrants may gain in the destination country, from skills and entrepreneurial attitudes (human capital) to extensive networks of friendships and contacts in the home, destination and other countries (social capital).
Remittances can be used in a variety of ways, and may increase access to other capital assets. For instance, they may be used to pay for education (human capital), to buy a shop (physical capital) and to improve access to land (natural capital). Greater access to these assets may in turn enable the household to support the emigration of other members, thus raising the level of remittances. At the same time, other capital assets contribute to the effective use of remittances. For instance, migrants usually rely on their networks in the home country (social capital) for help in building houses, buying land and setting up businesses. The following paragraphs review a few recurrent uses of remittances (the order does not entail any ranking; land-related uses are not included here, as they are dealt with separately in section 2.4 below). Before addressing these, however, some methodological challenges and conceptual caveats need to be highlighted.
Studying how households use remittances and the developmental impact of these uses at the micro level is riddled with methodological difficulties. First of all, as cash income, remittances are subject to fungibility problems: once they enter the overall household income pot, remittances become indistinguishable from other income sources. Therefore, it is difficult to associate them with specific income uses and with any particular changes in expenditure behaviour (Adams, 1996). For instance, although remittances per se may not be intended to be invested in land, they may enable the household to free other income for investment. To overcome this problem, some researchers have studied statistical correlations between international migration and specific behaviours or assets, rather than investigating the uses of remittances as such (e.g. de Haas, 2003).
Secondly, even where correlations are shown, it may be difficult to establish causation. For instance, if it emerged that migrant households were more likely to buy land or invest in education than non-migrant ones, it would still need to be ascertained whether this is due to remittances or to the fact that migrant households are likely to be among the wealthier and better-educated ones (de Haas, 2003).
Thirdly, the developmental effects of remittances are not limited to the immediate consumption or investment of the household receiving them. These first-round uses generate multiplier effects which are very difficult to measure (Adams, 1996; Glystos, 2002).
Finally, data are usually available at one point in time, which makes it difficult to assess change over time (Adams, 1996). This is particularly problematic considering that there may be a substantial time gap between migration and investment of remittances, as in the early stages households are often forced to use remittances to improve living conditions (de Haas, 2003).
Remittance uses vary greatly depending on variables concerning both the migrant (level of earnings, seasonal/long-term/permanent migration, etc) and his/her household back home (for instance, poorer households are more likely to use remittances to meet their basic needs). In this sense, what migrants and migrant households do with remittances largely depends on who they are. Moreover, remittances use also varies depending on the national and local context. For instance, using remittances to improve access to social services such as health and education may be the result of policy measures that negatively affect access to those services (e.g. structural adjustment). Generalizations should therefore be avoided.
Furthermore, households should not be treated as monolithic entities in remittance dynamics. Anecdotal evidence suggests that, consciously or unconsciously, migrants may use their role in supporting the family to increase their weight vis-a-vis other siblings, even breaking traditional age hierarchies. Moreover, while migrants are usually eager to help their family back home, they also need to protect themselves against a large number of unsolicited requests from the extended family (favours, loans, etc). Finally, while household members are usually the main receivers of remittances, friends and neighbours may also benefit, and businesses may be set up not only with family members but also through new relationships developed with other migrants while overseas (Kabki et al, 2003; Smith and Mazzucato, 2003). All these factors may affect use decisions (e.g. setting up a business in the capital city rather than in the home village).
In the literature, some distinguish between productive and unproductive uses of remittances. This approach is not followed here, as the border line between the two is blurred. For instance, a car can be used both to meet the households transport needs and to set up a transport business. Similarly, food, clothing and health care, often seen as consumption, also contribute to asset creation by increasing human capital.
Among the different types of remittance use, improving the households living conditions is a key priority. This is especially the case for lower income households, who can use remittances to supplement other sources of income. In Bangladesh, for instance, Rahman (2000) found that more than 30 percent of the remittances were used to meet households basic needs. This includes food, clothing and health care (e.g., on Ghana, Smith and Mazzucato, 2003; on Mexico, Basok, 2003; on Nigeria, Mali and Tanzania, Tacoli, 2002). Anecdotal evidence suggests that remittances are also used to buy consumer goods, such as domestic appliances, radios, mobile phones and satellite TV. Migrant households may use remittances to repay debts, including those contracted to support migration.
Remittances are frequently invested in education. In Ghana, for instance, Kabki et al (2003) found that remittances paid for school and university fees of children in the extended family. In Morocco, children in international migrant households were found to be significantly better educated than children in non-migrant or internal migrant households (de Haas, 2003).
Building a new house or repairing the existing one is consistently reported in the literature as the most frequent use of remittances. In Turkey, the 1970 Abadan Survey found that 49 percent of remittances were spent on housing (quoted in Russell, 1986). Similar results were reported for Pakistan (Gilani et al, 1981). In Ghana, Kabki et al (2003) found that the majority of the migrants interviewed were investing in a house for their family, and that most of those who had not done so yet hoped to do it in the future. In the Todgha oasis valley, Southern Morocco, remittances enable families to move out of the traditional village and build new houses (de Haas, 2003). A study from eleven Mexican villages found that between 71 percent and 89 percent of migrant households had expanded or repaired their house or built a new one (Basok, 2003).
House building may take place in the home area, in a different area within the home country (usually the capital city), or in both (Kabki et al, 2003; Smith and Mazzucato, 2003). These houses are often relatively unused, as the migrants are the sole occupants and only use them when they visit home (Smith and Mazzucato, 2003).
There is a widespread negative perception among researchers and practitioners of the migrants construction of luxury houses, which is held to be driven by prestige and status considerations and to subtract resources from productive investment. While social and cultural factors play an important role in house building projects, it must be noted that:
A spacious, functional and hygienic house is a legitimate aspiration, including because it may enable households to become more independent from the extended family (de Haas, 2003), and is at the very heart of what development is all about;
House building generates positive multiplier effects, as bricklayers and other workers use their wages for consumption or investment;
In contexts characterised by few investment opportunities and high inflation, housing is a sound way of protecting savings; and
Housing may also generate income, particularly by renting out the house or part of it.
Box 2. A life of hard work for a house back in Cape Verde
Ivette left her home in Cape Verde and moved to Italy when she was only 16, in search of an independent life. She has been a domestic worker in Rome for more than 25 years, and her daughter was born from an Italian man. Her parents and her ten brothers and sisters are spread across Cape Verde, Portugal and Spain, and she keeps in touch with them by telephone and through periodic visits. Years ago, she used to send football T-shirts to Cape Verde for a small business venture with her brother, until he moved to Spain.
Ivette has always sent money back home to support her family. Moreover, in 1987 she started to build her house, which is now nearly completed. Rather than starting from scratch, she has added a floor to her parents house in the town of Mindelo, and her brothers have helped her with the construction works. A few years ago, she found out that her parents did not have a land title and paid an annual fee to the municipality. In order to feel more secure, she sought a title for the land, and obtained it with the help of her sister. Although renting out the flat while she is away would earn her good money, she is reluctant to do so as she fears that tenants may damage the flat (after all the hard work she put in it!). Ivettes sisters in Spain have followed a different route, as they have bought flats in Spain through mortgages.
Although she used to cultivate her fathers fields when she was a child, Ivette is not interested in agriculture, as land in Cape Verde is very arid. She would like to return to her country on retirement, and finally enjoy her house. However, she first needs to work until she fulfils the requirements of the pension scheme in Italy. Source: interview with Ivette.
Many migrants use part of their earnings to set up businesses in their home country. This may involve sending goods to a family member, who then sells them (see e.g. Ivette in Cape Verde, box 2 above), or transferring money. Examples of businesses include communication centres (telephone kiosks, Internet points, etc), small hotels and restaurants, trade in second-hand goods, and commercial transportation (Kabki et al, 2003; Smith and Mazzucato, 2003; de Haas, 2003; Black et al, 2003). Frederic, a taxi driver in Burkina Faso, told us that his minibus was sent to him by a cousin in Germany, who bought it second-hand and shipped it via the port of Lomé. He now uses it as a taxi in Ouagadougou and to take tourists for short trips to the countryside. Taxi drivers across West Africa told us similar stories.
In setting up new businesses, migrants rely not only on kin-based networks, but also on relationships developed while abroad with other migrants or, in some cases, with Europeans or Americans (Smith and Mazzucato, 2003).
Remittances are also used to erect new religious buildings (churches, mosques, etc.), to support religious institutions (e.g. church donations) and to pay for religious ceremonies, particularly funerals and weddings in the extended family. Besides faith, these decisions are influenced by social status considerations. In Ghana, Kabki et al (2003) report remarkable differences between funerals funded by migrants and local funerals, especially in the type of coffin, the number of invited guests, food and drinks, and entertainment. In some cases, decisions regarding religious buildings may be controversial within the community and give rise to disputes (see box 3 below).
Besides contributing to household livelihood strategies, remittances may also support conflict. In Somalia, for instance, remittances provided funds for the militias that eventually overthrew the Siyad Barre regime, and are now a source of funding for the powerful warlords (Gundel, 2002). Similar reports have been made for the Tamil Tigers in Sri Lanka (Sriskandarajah, 2002). In other cases, dynamics generated by remittances may fuel violence in an unintended way (see box 3 below).
Box 3. The construction of a mosque sparks violence in Mali
Yerere is a small village in the cercle of Nioro, in Mali, an area characterised by substantial out-migration. In 2003, the construction of a village mosque funded by a migrant in Gabon sparked violent clashes. In pursuing his project, the migrant followed the legal route and secured the approval of local authorities, but by-passed the customary chief and village community, who were opposed to the new mosque. The reasons for this opposition are complex: social (the migrants family had come to the village from another district), personal (the migrant had divorced the daughter of the local chief), religious (the mosque serves a minority cult) and political (the migrants family and the local chief support opposing political parties).
When construction works started, local opposition turned violent, resulting in several deaths (including the migrant who had come to the village for the occasion), many wounded and a large number of arrests. After the clashes, government authorities strongly backed the continuation of the project, and the mosque is now completed.
Sources: Dama, pers.comm; Maiga, pers.comm.
Within the highly diversified livelihood strategies of rural households, land constitutes a key asset. The linkages between remittances and access to land are extremely complex and likely to vary considerably from place to place, depending on local socio-cultural factors, on the local economic and ecological context, and so on. They can take many different forms, from land purchases funded through remittances to a variety of other land access mechanisms (rentals, administrative allocations, loans, etc). They raise a range of issues, such as effects on land use patterns, on land competition and disputes, and on land access for non-migrant households. The land-related effects of remittances may take decades to materialise; for instance, de Haas (2003) found that many migrants only started to invest in land after more than two decades of absence.
Studies from across the world show that purchases of agricultural land constitute a frequent form of remittance use. Land purchases may be formal transfers of land ownership, or informal transactions where the seller is not the legitimate owner (e.g. as land may be owned by the state) or has no land title to prove his land rights. In Bangladesh, Mahamood (1992) found that some 15 percent of the remittances sent by migrants in the Middle East went in land purchases, second only to housing (19 percent). Moreover, a survey of over 700 households in rural Pakistan revealed that international remittances had a positive and statistically significant effect on the accumulation of both irrigated and rain-fed land (Adams, 1996). In the Indian state of Kerala, Zachariah et al (2001) found that a larger proportion of migrant households had bought land than non-migrant households during the five years preceding their survey. In rural Egypt, Adams (1991, quoted in Adams 1996) found that 73 percent of total per capita expenditures on investment by external migrants went into the purchase of agricultural and building land.
In the Todgha oasis valley, Southern Morocco, international migrant households are found to have a higher propensity to invest in land than other households: more than one quarter of international migrant households have purchased agricultural land, compared to less than 10 percent of non-migrant households. Factors accounting for this difference include not only higher income levels associated with remittances, but also the greater income stability and future security ensured by remittances and migrants greater entrepreneurial attitude. Indeed, the remittances-land correlation exists even after controlling for income levels (de Haas, 2003).
Fewer studies are available which document the location, quality and size of the land purchased. Basok (2003) compares land purchases in eleven Mexican villages by migrants participating in the Canadian Mexican Seasonal Workers Programme. Her research found that land purchases were more likely in the worst endowed areas (in terms of proximity to the municipal capital, paved roads, etc), where 23 percent of the migrant households had purchased land for subsistence agriculture (compared to nine percent in the best endowed areas). This is explained by the fact that land prices in best endowed areas were too high for migrants, although these results may be influenced by the specificities of the Canadian Mexican Seasonal Workers Programme (most migrants participating in this scheme are poor; Basok, 2003).
In his study in Southern Morocco, de Haas (2003) found that most land was bought outside the oasis area, in previously barren lands reclaimed through the creation of new water points (see below). This is because migrants prefer to avoid the fragmented land tenure system, the inflexible water regulations and other social constraints associated with oasis lands. On the other hand, aged returned migrants tended to buy relatively small plots in the oasis, as they are more oriented towards traditional oasis agriculture than younger migrants (de Haas, 2003).
More research is needed on the nature of remittance-supported land purchases, particularly on the extent to which they are formalized or informal transactions, and on the process through which they take place. Anecdotal evidence from areas around Bamako, Mali, suggests that purchases are usually done informally, as sellers rarely have land titles; it is then up to the buyer to formalize his land rights by acquiring formal title from the authorities. Moreover, we were told that negotiations, purchases and subsequent regularizations are usually done by an intermediary on behalf of the migrant, and that the intermediary makes a profit if he manages to buy land for less than what was agreed with the migrant.
These processes may be characterised by high transaction costs, speculation and even fraud. Again, evidence from Mali provides an example. A recent study from Banko, a village close to Bamako, documents the case of a man who pocketed seven million CFA (some 4500 euros) by selling ten hectares of land over which he had no rights (the plot was part of a land area used communally by the village). When the buyer - an intermediary acting on behalf of a Malian residing in France - tried to occupy the plot, he met with the resistance of the local community. The migrants attempts to seize administrative authorities and local notables of the matter were vain, while the fraudulent seller vanished with the money and the intermediary was secretly satisfied - quite understandably so, as he had himself pocketed twice the amount given to the seller by deceiving the migrant about the price of the plot (Djire, 2004).
The relatively small body of literature on the linkages between remittances and land seems to focus on land purchases. However, evidence suggests that remittances may directly or indirectly improve access to land in many other ways. More research is needed better to understand these different aspects of the remittances-land linkages. Here are just a few examples:
Increasing the security of existing land access. Examples include paying for the procedure to obtain title over family land; releasing land by paying out mortgages (e.g., for Bangladesh, Siddiqui and Abrar, 2001); and, more generally, making more productive use of land through hired labour and agricultural inputs (in many Francophone West African countries, land rights are conditional upon productive land use - mise en valeur).
Leases and rentals. Migrant households keen to invest remittances in agriculture (hired labour, agricultural inputs, etc) but without access to sufficient land may improve their land access through rentals and long-term leases. These are particularly useful where land sales are prohibited, or where individuals or groups are not prepared to sell customary land (e.g., on Ghana, see chapter 4 below).
Land allocations. Migrant households may apply for land to institutions responsible for land allocation, whether customary or statutory, by showing their increased capacity to cultivate more land through remittance-supported hired labour and agricultural inputs. This links up with decentralization processes underway in many developing countries, including across West Africa, as local governments may be granted land management responsibilities (e.g., in Senegal, the rural councils; see chapter 3 below). In this scenario, relevant issues include mechanisms to ensure the representation of migrants before land management institutions, to protect the interests of non-migrant households and to ensure the transparency of the land allocation process.
Land loan dynamics. Remittance inflows may also affect land loan relations. For instance, in Banamba, Mali, migrant households may temporarily increase their land for cultivation by borrowing plots; conversely, migrant land-owning households who had lent to others the lands that they were not able to use, and who are now able to cultivate more land through hired labour and agricultural inputs, tend to claim back their lands (Diarra, pers.comm.). In some cases, these dynamics may give rise to land disputes (e.g., on Senegal, see below, chapter 3).
Effects on land inheritance. Within the household, remittances may affect relations between migrant and non-migrant siblings and between children and parents. Because of their important contribution to family income, migrants may be rewarded with preferential treatment in land inheritance. For instance, in Tanzania, Diyamett et al (2001) found that as the traditional inheritance system becomes less frequent, sending remittances has often become a way to gain parental favour and inherit the family land (page iv). Whether this case is part of a more generalised phenomenon is a question worth further research.
The linkages between remittances and land may be mediated by other natural resources, particularly water. In the Todgha oasis valley, Southern Morocco, de Haas (2003) found that remittances supported a major expansion in use of motor water-pumps, with migrant households using remittances to fund the creation of new water points. Greater access to water has been key for economic development in the valley since the 1970s, as it enabled all-year round cultivation and increases in agricultural production. Pump-owning households also sold water to other households. Moreover, new wells increased access to land for oasis dwellers by permitting the reclamation of lands in previously barren areas outside the oasis. These newly reclaimed lands were first allocated to villages and then to individual households (de Haas, 2003). Water-point creation is also reported to be a key activity of migrant associations in Senegal and Mali.
However, the creation of water points may also bring about negative side effects. In the Todgha oasis valley, the absence or lack of enforcement of any form of regulation has resulted in an anarchic rise of water pumping, raising concerns for environmental sustainability. Many wells have dried up, presumably due to excessive pumping (de Haas, 2003).
More research is needed to ascertain whether similar processes are also happening elsewhere, particularly in the Sahel. In many Sahelian pastoral societies, access to dry season grazing land is determined by rights over the water points located there, and water rights are therefore crucial to manage pastures sustainably. In the past, failure to recognise these complex relations between water and land, and government provision of open-access water points have undermined customary land tenure systems and had negative environmental and socio-economic consequences. Whether similar processes are now taking place in relation to water points created with funds from the diaspora is a question worth exploring.
The existing literature seems to pay relatively little attention to changes in land use associated with remittance-supported land purchases and more generally with remittance inflows. Questions to be investigated include for instance whether migrant households use the newly acquired land for subsistence farming or for commercial agriculture; in the latter case, whether they cultivate traditional crops or new types of crop; whether they mainly rely on family labour or hire farm workers; whether these processes affect the relations between farming and herding; and so on.
A claim that is sometimes made is that out-migration and the ensuing labour loss may result in a reduction of land under cultivation (Rubenstein, 1992, quoted in de Haas, 2003). However, in some cases remittances may help tackle this issue, by enabling households to recruit hired labour to cultivate the fields. For instance, in Southern Morocco, de Haas (2003) found that the incidence of fallow land was highest among non-migrant households, and that migrants usually entrusted land cultivation to other household members (women in particular), to sharecroppers or to hired labourers. This improved wage levels and sharecropping conditions: while traditionally sharecroppers used to retain 1/5 of the yield, at the time of the study they retained on average 41 percent of the harvest.
An example of land use change that may be induced by remittance-supported land purchases is the transition from agricultural to residential use. Evidence from West Africa shows that in peri-urban areas, particularly along paved roads, agricultural lands are being converted to residential plots, land prices are soaring and buildings are mushrooming very rapidly without adequate service provision. While many different factors are causing these changes, because of their greater-than-average purchasing power migrants are often among the main initiators of building projects. In some cases, legislation protects the land rights of urban developers that have built beyond a specified height (e.g. in Ghana, the Land Development Protection of Purchases Act 1962). This creates an incentive for developers, whether migrants or others, quickly to build concrete structures beyond that height, and then wait to earn more money to complete the building. This may partly explain the building skeletons that can be seen in the peri-urban areas of many West African countries. More research is needed to assess the scale of the phenomenon, the extent to which international remittances are contributing to it, and its impact for the farmers in peri-urban areas that lose their lands as a result.
Land competition, multiple sales and land disputes
In many parts of West Africa, valuable land, whether high quality farm land or residential plots, has become scarce as a result of a variety of socio-economic changes, particularly demographic pressure. Competition between users has grown strongly. Lively land sale markets have also developed, both formal and informal. These changes have brought some negative side effects. In many cases, conflicting land claimants, even within the same extended family, sell the same plot to different buyers. As a result, land disputes are growing, and tenure security is undermined. Central and local government officials manipulate this increased land competition for corruption and rent-seeking purposes. More research is needed to assess the extent to which remittances, by increasing demand for valuable land, are among the factors driving these changes. The case reported in the box below suggests that they may well be.
Box 4. Multiple sales of residential plots in peri-urban Bamako, Mali
Dialakologi is a commune close to Bamako, in Mali. In 2000, the mayor launched a scheme to sell some 300 peri-urban plots for residential purposes, claiming that the land area was owned by the municipality. Several sales were made to private individuals under the scheme. A buyer from Banamba, a town some distance to the north, paid the municipality a large amount of money for the purchase of 20 plots. Although his motives are not known, it is highly likely that he was acting not only for himself but also on behalf of migrants overseas (Banamba is home to a large number of migrant households, and the amount of money involved would be very hard to raise without access to international remittances).
However, a local landholding family claimed customary rights over the same area of land. They argued that the land title shown by the municipality was forged, and that the land really belonged to the state. Therefore, under Malian law, customary rights over that land should be protected. The family also sold some plots within the land area to private individuals. Again, although precise information on the profile of the buyers was not available, circumstances suggest that international migrants were involved.
Buyers who had bought land from both sellers immediately started to build on the plots, while the dispute between the municipality and the landholding family continued. In August 2003, the Tribunal of Kati upheld the land title produced by the mayor, decided in favour of the municipality, and ordered the demolition of the houses built by those who had bought from the landholding family. However, in November 2003, the Court of Appeal in Bamako reversed the decision, decided in favour of the landholding family, and ordered the demolition of the buildings of those who had bought from the municipality. The situation is currently stalled, as all parties are waiting for the outcome of the next local elections, scheduled for late May 2004.
Source: Diarra, pers.comm.
The effects of international migration on the distribution of income and wealth in the home areas have long been debated in the literature. Because long-distance migration requires a minimum level of information, contacts and resources, those who migrate are likely to be among the better-off within the community, at least in a first stage (Adams, 1996; de Haas, 2003). These better-off households are therefore more likely to receive remittances, which they can use to buy land and other key resources, set up new businesses and improve their childrens education, thereby further widening the gap with non-migrant households.
In Pakistan, Adams (1996) found that international remittances increased land concentration by upper-income groups, as they enabled them to buy irrigated and rain-fed rural lands. On the other hand, internal remittances tended to go to lower-income households, who used them for the purchase of agricultural machinery, and partly compensated the negative distributive effects of international migration. Similarly, in the Philippines, Go (2002) found that international remittances mainly benefited richer regions, classes and households. Moreover, anecdotal evidence across West Africa suggests that where land is a tradable commodity, price rises caused by remittance inflows may make it more difficult for non-migrant households to gain access to land, particularly in peri-urban areas.
However, this distributive impact may change over time. Information may become more readily available, and migrants networks abroad may motivate and help others to follow. Therefore, poorer households may also be able to seize the opportunities offered by migration (Adams, 1996; de Haas, 2003). At this stage, international migration may help reduce income and wealth inequality at home. First, poorer households receiving remittances can increase their income and diversify their livelihoods. Secondly, cultural change facilitated by migration may make lower classes and castes more confident of challenging traditional elites. For instance, in Southern Morocco, as a result of these two factors lower classes (smallholders, sharecroppers and agricultural labourers) have started to challenge the traditional land allocation system, which is controlled by local elites (de Haas, 2003).
Therefore, for households, horizontal mobility (i.e. migration) and vertical (i.e. social) mobility are strictly intertwined in a mutually reinforcing cycle. Ultimately, the distributive effects of remittances may change considerably from place to place, depending also on the distribution of income and wealth before migration. In highly stratified societies, remittances may change the nature of inequality, by superimposing new elites on the old ones; whether they increase inequality or not is more difficult to say (see de Haas, 2003).
Migration is a gendered phenomenon. While both men and women migrate, the extent to which they do so and the nature of their migration (duration, type of occupation, etc) tend to vary along gender lines. Where migrants leave women and their households behind, womens responsibilities for household livelihoods and for agriculture increase. Some authors suggest that this increased burden is accompanied by greater decision-making power within the household. Others note that this effect is very limited, as women tend to remain under the protection of the extended family (David, 1995; de Haas, 2003).
More research is needed better to understand the intra-household aspects of remittances-land linkages, particularly whether and to what extent women retain control over remittances and use them to improve their access to land, or whether major remittance use decisions, such as those concerning land-related uses, are taken by male relatives within the extended family. The answers to these questions are likely to vary greatly from place to place, depending on cultural factors, on the nature of migration, on urban-rural differences, and so on.
A final aspect of the linkages between remittances and land concerns the effects of remittances on the land tenure system as a whole. This is an extremely complex issue, and at the stage only some hypotheses may tentatively be put forward.
As for land rights, in areas where land is becoming scarcer customary systems have become increasingly individualized as a result of the breakdown of the extended family, and of agricultural intensification and commercialization. In these areas, land rights are often sold. Demand for land purchases from urban elites and others is contributing to these tenure individualization processes. Remittances may increase the demand for valuable land; the extent to which this is happening and is fostering individualization of land rights deserves further research.
As for land rules and institutions, migration may accelerate the breakdown of customary systems. In Southern Morocco, for instance, migration has accelerated the erosion of customary land and water authorities through: the partial emancipation of formerly inferior groups, as they challenge rules and institutions dominated by traditional elites (see above); land reclamation through water-pumps and establishment of new farms outside the oasis area, where customary systems are less strong; and, more generally, cultural change (de Haas, 2003).
Remittance decisions, and their implications for access to land in home countries, are affected by many factors, some of which were mentioned in previous paragraphs. A good understanding of these factors is key for policy makers and development practitioners to be able to harness the potential of remittances to promote local development. Relevant factors may be broadly grouped in two categories: those affecting the amount and frequency of remittances, and those affecting remittance use and, more specifically, the propensity to invest in land.
Decisions on whether to invest remittances in land are likely to be affected by a range of factors, relating to the migrant (desire to return; amount of remittances sent; age - younger migrants may be keener on non-agricultural businesses; etc), to the family back home (e.g. income level: poorer households may find it more difficult to invest remittances), to the economic and ecological context in the home area (land quality, profitability of agriculture, infrastructure and market access, etc) and to land prices relative to migrants purchasing power (see Basok, 2003). Evidence suggests that while migrants may be willing to invest their savings in their home country, they are often reluctant to do so because they do not trust enough the government or local intermediaries.
Policy frameworks in both home and destination countries are also likely to have a profound influence on remittance decisions. As for land-related uses, here are two examples of relevant policy areas:
Land legislation and the broader legal system, particularly the extent to which they provide adequate tenure security, ensure the transparency of land allocation procedures, and enable land purchases, including by migrants. In countries where land ownership by non-nationals is restricted and where dual nationality is not permitted, migrants may find it difficult to buy land if they acquire the nationality of their new country of residence.
Trade policy as it affects the economic profitability of agriculture. For example, cheap food imports may induce using remittances to buy food rather than to buy land to grow food (David, 1995). Equally, sanitary and phytosanitary barriers may frustrate the opportunities that transnational networks create to export agricultural products to Europe and the United States, thereby reducing the propensity to invest in land and agriculture.
This chapter has shown the complexity of the linkages between international remittances and access to land, and of the issues that such linkages raise. The next two chapters will address these issues with specific regard to the cases of Senegal and Ghana.
 For an excellent analytical
review of this literature, see de Haas, 2003.|
 Note, however, that comparisons between these types of external funding are difficult given their extremely different nature. For instance, FDI involves both financial inflows (capital) and outflows (profits, investment returns, etc).
 Although transport has become cheaper, migration can still be quite costly, especially for undocumented migrants, whose movement is increasingly arranged by highly organised and unscrupulous groups.
 For instance, in Cape Verde, Juana - Ivettes godmother - lives in the upper floor of the house she built and rents out the lower floor, so as to complement the pension she receives from Italy.
 Note however that another study on migration from Bangladesh to Singapore reached very different conclusions: most of the remittances were used to repay the loans contracted to finance migration (52 percent) and to support household consumption (roughly 30 percent) (Rahman, 2000). These different findings might be explained by differences in the nature of Bangladeshi migration to Singapore and to the Middle East.
 Note that this data lumps together agricultural and residential land.
 See e.g. Ivette in Cape Verde, box 2 above. Note that the titling process may be extremely costly for locals, both formally (fees) and informally (bribes). It would be interesting to study whether, while locals may feel relatively secure under customary land tenure, migrants tend to seek land titles.
 Land allocation may complement informal purchases. For instance, reports from state-owned irrigated schemes in Mali and Senegal suggest that tenants informally sell their rights to outsiders, who then occupy the land and on that basis apply for formal land allocation by the authorities.