Remittances play an increasingly important role, especially in the smaller island states. In many countries remittances form a significant part of disposable income, hence the smaller island states (specifically initially Kiribati, Tokelau, Cook Islands and Tuvalu) have long been conceptualised as MIRAB states, where Migration, Remittances, Aid and the resultant largely urban Bureaucracy are central to the socioeconomic system (Bertram and Watters 1985). The notion of MIRAB is applicable also in rather larger states such as Samoa and Tonga, where remittances constitute some of the highest proportions of GNP of any country in the world. While this acronym is disliked in the Pacific, for cultural reasons and because of its implication of a ‘handout mentality’, it nonetheless suggests the centrality of migration and remittances in the island states, and has been largely unchallenged for two decades (Bertram 1999). It spawned other acronyms such as MURAB, which places extra and appropriate emphasis on the attendant urbanisation within island states such as Tuvalu (Munro 1990), and even MIAB, where migration did not initially stimulate a significant flow of remittances, as in some parts of Micronesia (Ogden 1994; Karakita 1997), though there is now a shift in this direction (Grieco 2003; Connell and Brown 2005).

Because of the continued and increasing significance of remittances, the sustainability of remittance-dependent development is particularly important — and necessarily uncertain — especially if, in the countries of origin, the need for remittances grows faster than its supply (Macpherson 1992), or if the number and flow of migrants dwindle. The rate of growth of migration to major destinations, namely New Zealand, Australia and the USA, has declined in recent years due to the restructuring of migration controls; return migration has also sometimes been considerable (Maron 2001), and migration has become increasingly more selective towards those with skills (Liki 2001; Brown and Connell 2004). Even with continued migration, however, an anticipated imbalance is assumed to occur because of the dynamics of settlement migration. With family reunification and with greater integration of migrants in the host communities, their ability and willingness to remit have been expected to decline over time (e.g., Macpherson 1994). If that were so, without other sources of income, the future of the economies of remittance-dependent Pacific countries would be uncertain.

Only recently have second and third generations of Pacific Islanders grown up outside their island ‘homes’; hence the extent to which they will remit to the island Pacific (and even whether they can be described as ‘Islanders’ or ‘migrants’) is not well known. There is some evidence that the links between second-generation Samoans in New Zealand and Tongans in Australia with Samoans in Samoa and Tongans in Tonga respectively are declining, though, at least in the latter case, links are maintained with other migrant Tongans elsewhere (Muliaina 2001; Lee 2003, 2004a). Similarly, these new generations are more likely to act as individuals rather than perceive themselves as members of wider transnational social groupings. Anecdotal evidence points to the growing individualism of overseas migrants, but especially to the increasing numbers of second-generation Islanders born overseas, and the reduced likelihood of such people sending remittances to their ‘home’ countries, especially if they take up host country citizenship. Thus one migrant Tongan has observed of the Australian context: ‘People who were born here [Tonga] or went to school here send remittances. People born there? No way!’ (quoted by James 1997: 20). Many Tongans in Melbourne have lost interest in continued financial support of their overseas kin, as their sense of kavenga (obligations) has declined over time (Lee 2004). Similarly skilled Tongan migrants in Sydney are increasingly stating that they no longer remit (Fusitu’a 2000). Though data on Tongan and Samoan nurses in Australia indicate that even skilled migrants sustain remittances at high levels and over long periods, those data date from the mid-1990s (Brown and Connell 2004, 2005), though circumstances might not have changed.

Remittances are particularly important in the smaller states of Polynesia and Micronesia, and in the more remote islands in those states. Thus for the tiny coral atoll of Manihiki (Cook Islands), migration and the resultant remittances have been seen as so crucial that they constitute nothing less than a socioeconomic strategy for collective survival (Underhill 1989). In Nanumea, in Tuvalu, remittances grew from being about half of the island income in the 1970s and 1980s to some 75 per cent in the 1990s, in large part because of the collapse of copra marketing as world prices slumped (Chambers and Chambers 2001: 156). In Kiribati, the 2000 census recorded that overseas remittances, primarily from seamen, were the primary source of income for as many as 30 per cent of households in the urban centre of South Tarawa, and, in 2002, some 35 per cent of all households in Tuvalu received remittances. In both countries this was usually the main source of income in the outer islands (Abbott and Pollard 2004; Borovnik 2003). Similar situations occur in other small islands and island states, and there is mounting evidence that the northern Micronesian states are now following this pattern (Connell and Brown 2005: 6).

Conventional wisdom suggests that remittances are used overwhelmingly for consumption objectives and inadequate amounts are directed towards investment. There are, however, alternative perceptions of the use of remittances and issues of fungibility (where the specific use of remittances cannot easily be distinguished from other spending patterns and might distort them) (Connell and Brown 2005). After debt repayment, remittances are used for housing and for community goals (such as water tanks and churches), airfares and education (an investment in social capital). They are also used for various forms of investment, sometimes in the agricultural sector but more frequently in the service sector, and especially in stores and transport businesses. In Samoa and elsewhere, remittance money has constituted the start-up funds for many shopkeepers and other small business entrepreneurs. Half of all market vendors in Apia (Samoa), all of whom received remittances, claimed that some had been used as capital for the purchase of seeds, fertiliser and tools to engage in food production for sale (Muliaina 2001: 28). Even on a small outer island such as Falahola (Tonga), remittances have been used for economic ventures, ranging from agriculture to tourism, though remoteness has limited their success (James 1991: 18–20). Throughout Tonga the increased use of remittances for investment purposes, in fishing, agriculture, stores and transport businesses, attests to the shift from consumption to investment (Faeamani 1995; Walker and Brown 1995), which has occurred in part as consumption goals have been satisfied. This transition has also occurred in Pakistan (Helweg 1983), parts of rural PNG (Boyd 1990) and in similar small-island environments in several parts of the eastern Caribbean (Connell and Conway 2000). While the transition might benefit economic development, at least one anthropologist (Small 1997: 134, 195, 199) has raised concerns that it might also further emphasise intra-village (and country) economic inequalities and hamper social development.

Remittances have also contributed to urban investments of various kinds. The expansion of the Nuku’alofa flea market in Tonga since the mid-1980s demonstrated the manner in which many households used remittances as an investment in their market stalls, and later economic diversification (Brown and Connell 1993). Moreover, as James has noted, ‘It has been argued that remittances take away the motivation of locals to produce, but the facts of local entrepreneurship seem to contradict this since large consignments of local products … are sent to relatives overseas for sale among the Polynesian population’ (1991: 2). Even in the most difficult circumstances, remittance recipients make efforts to invest where they can. In the outer islands of Kiribati, where most remittances went into providing basic needs, and the custom of bubuti (a request that cannot be refused) makes savings let alone business almost impossible, all recipients nevertheless sought to retain some income to invest in land purchase, doughnut bakeries, stores or even in sewing material for blouses that might later be sold (Borovnik 2004). Here, as elsewhere, there is no evidence that any part of the economy is abandoned or neglected but that remittances enable some limited diversification. Where there are opportunities, and where consumption goals have been satisfied, remittances are used for investment, to stimulate entrepreneurial and trading activity, increase the extent of formal sector employment and produce multiplier effects.

Even so, it has often been argued that remittances (and aid) are not conducive to private sector development, broadly what Ahlburg (1995: 42) has called the ‘disincentive effect’. Indeed, the MIRAB model implies that there is a strong bureaucratic outcome, which might stifle, rather than enable, increased productivity. This perspective was stressed by a series of authors in the 1990s, in overt or subtle attacks on the perceived stultifying role of aid in the region and the consequent necessity to stimulate the private sector (e.g., MacMaster 1993; Browne 1995; Duncan 1994; Pollard 1995; La Plagne et al. 2001). In the case of Tonga, Sturton argued that ‘[t]he Tongan economy displays all the characteristic markings of the “Dutch disease” where a dominant export activity attracts a disproportionate command over resources, pushes up domestic production costs, and reduces international competitiveness. In the Tongan case the “booming” sector has become development assistance and migrants’ remittances’ (1992: 3). Similarly, Faeamani has argued that, through the combination of the loss of young adults and an inflow of cash in the form of remittances and goods, ‘there is a consequent reduction in garden size and production’ (1995: 140; see Fairbairn 1993). More generally, several authors have stressed the wide-ranging notions of dependency that remittances appear to create.

It is implausible that remittances have no disincentive effects, but there is remarkably little direct evidence of this. MacMaster has suggested that in the Cook Islands, Tonga and Samoa remittances are ‘a mixed blessing as they undermine the incentive to work and are rarely spent on productive investment. They are normally used for unproductive ceremonial purposes or on imported luxury consumption items’ (1993: 279). These and other similar statements and conclusions (e.g., Ahlburg 1991: 39; Finau 1994: 308; World Bank 1990: 4) have suggested overall that ‘it is not clear that the net effect of remittances and aid is conducive to long-term economic viability and prosperity’ (Cuthbertson and Cole 1995: xiv). Few of these studies present data that justify these assumptions and conclusions.

It is all too often assumed that there is no desire to maximise (or even improve) incomes, hence Islanders become perceived as somewhat irrational or even lazy. Indeed, as Muliaina has observed, even second-generation Samoans in New Zealand assume that their relatives in Samoa are failing to take proper economic advantage of obvious land and marine resources (2001: 33). A similar situation appears to be more broadly true in the Pacific:

The growing flows of remittances into some countries (Kiribati, Samoa, Tonga, Tuvalu and increasingly Fiji Islands) are giving rise to what many [Islander] respondents termed laziness or over-dependence on others. This ‘easy money’ was perceived by many respondents to be a disincentive for young people to actively look for work. (Abbott and Pollard 2004: 61)

It is at least as likely that actively looking for work might well have been a waste of time. Moreover, subsistence activity is only exceptionally abandoned or reduced, despite remittances and attitudes in opposition to it. Where conditions are appropriate to adequate income generation, even where remittances have reached high levels, the private sector might flourish and be stimulated by remittances.

The ‘crowding out’ argument was prominent in the initial formulation of the MIRAB model by Bertram and Watters, who argued that, with additional sources of income from remittances (and indirectly from aid), people ‘can be expected to evaluate the return on [agricultural] investments relative to the alternatives. On this basis it would be expected that as the alternatives to commercially oriented agriculture would improve, so a reallocation of household effort away from agriculture would take place’ (1985: 511). Despite Faeamani’s (1995) observations on declining production in Tongan villages, this might be exceptional, and indeed might be more likely to be a function of labour shortages (Evans 1996, 2001). Even so, as agricultural decline does occur, as has partly happened in the Cook Islands and Niue (Connell and Brown 2005), it is ‘what is required for efficient economic behaviour: that the family allocates its resources to the highest productive use, even if it happens that this particular use is not “productive” in the “domestic” economy, but rather in a “foreign” economy’ (Poirine 1998: 77).

Most studies of remittances have observed that a significant proportion of remittances support ‘traditional’ customs and obligations. This is at least in part because economic opportunities are few, so investing in custom avoids what would amount to ‘intensive self-exploitation in agricultural activity’ and gives villagers respect and autonomy (Evans 2001: 17–18). Similarly, sellers in the Nuku’alofa flea market, most of whose goods arrived as remittances — and who might be seen as involved in trade and investment, the fetishisation of cash, sales rather than gifts, thus epitomising the rise of market capitalism — preferred to see themselves as located within complex, reciprocal exchange systems that ‘maintained the social integrity of Tongan society despite diasporic fragmentation’ (Besnier 2004: 19). It is simply more appropriate to engage in exchange and gift-giving rather than sale and purchase, hence commercial practices are downplayed in favour of social obligations. The social and the economic cannot be disentangled.

Successful development of small businesses provides some incentive for those who have sent remittances to return home and manage those businesses. In overseas Polynesian households that include nurses, the greatest propensity to return comes from those with business investments at home (Brown and Connell 2004). International migration has provided new opportunities for women, including those in remote areas and outer islands, as in Tonga, where the production of traditional textiles (koloa) for the ceremonial economy is a prime example (Horan 2002). Overseas Tongan women lack access to the required raw materials and often also to time, instead importing these textiles from Tonga. Their reciprocal contribution of remittances further exemplifies how remittances are constituted through a process that might be seen simultaneously as social exchange and trade.

The MIRAB model has been largely unchallenged for two decades (Bertram 1999), and continues to be recognised in new contexts. It has, however, been appropriately observed that the model appears to largely deny any real semblance of agency to Pacific Islanders, other than as migrants, though this has been evident, for example, in a range of autonomous agricultural and other activities, such as the 1990s’ boom in squash cultivation in Tonga (van der Grijp 1999; cf. Hooper 1993), and in the above example of koloa production. Moreover, it is evident that

At the very least, remittances have complex and important social and economic dimensions.

The use and structure of remittances have changed over time, with significant intergenerational shifts in their structure. Initially remittances are sent to parents — as is so clearly happens in the majority of Pacific cases — and in an economic sense can be seen as repayment for their past investment in the human capital of the migrant; in a social sense this is usually expressed as duty, loyalty and maintenance of family ties. A second wave of remittances is subsequently more likely to be dominated by brothers and sisters and by children; this phase might also correspond with a decline in the volume of remittances. Decline after the death of parents seems ubiquitous (e.g., Muliaina 2001: 25). That phase can be seen as an investment in the human capital of the next generation (Brown and Poirine 2005). The third and final phase represents payments to spouses — and indirectly (via investment) to the remitters themselves, as return migration and/or retirement are approached. Some shifts in the destination of remittances are universal, sometimes paralleled by a decline in volumes, but both trends point to a structure that increasingly favours the interests of the senders.

The limited available evidence on the remittance patterns of the second generation indicates that, as Lee (2003, 2004a, 2004b) and Muliaina (2001) have emphasised, they respond only indirectly through the urgings of their parents and their sense of family, and they therefore contribute very limited sums. This is particularly significant as migration opportunities decline and the number of overseas-born ‘Islanders’ becomes the majority. Thus overseas-born Tongans (or ‘Tongans’) in New Zealand, as also in Melbourne (Lee 2004b), alongside Cook Islanders, Niueans and others in New Zealand, are now a majority rather than a minority. Not only does this probably mean that for all these groups their remittances are limited, but their social and economic ties are likely to increasingly be with each other rather than with ‘home’, so accentuating this trend. Lee (2004b: 10–11) has thus concluded that

She further warned against the complacency of many people in Tonga, and institutions outside, towards the notion that remittances would continue into the indefinite future (ibid.). As overseas generations lose language and cultural skills, or ‘marry out’, their sense of belonging must decline. Moreover, there is no certainty that migrants’ economic status will always improve.

Overall, remittances have contributed substantially to welfare in most states, especially improved housing, and to raised levels of consumption. Despite widespread concerns that remittances are spent rather than invested and constitute a ‘moral hazard’ by reducing the incentive of recipients to work (Wheatley 2003), there is limited evidence in support of this in the Pacific. Remittances are invested where this is feasible and opportunities exist and, as in other parts of the world, there has been a shift in the use of remittances from consumption to investment (Connell and Brown 2005). Remittances have contributed to employment (especially in the service and construction sectors) and eased balance of payments problems, despite contributing to inflation. Moreover, remittances have been sustained to a higher level and over longer periods than has been predicted, or has occurred in other parts of the world. This has often entailed some sacrifices by senders, to the extent that this might have hampered their own futures; thus, in a recent article lamenting the educational and economic success of Pacific Islanders in Australia, a Tongan lawyer stated, ‘If the kids are not doing too well at school, they could spend $30 on tutoring. Rather than send away $1,000 to the folks back home that $30 investment will pay off in the long term’ (quoted in Tora 2005: 5). That shift towards prioritising the nuclear family is occurring only slowly.

There is no consensus on whether remittances improve or worsen income distribution. This is unsurprising given diverse contexts. Until relatively recently, the dominant view was that remittances tended to reinforce income inequality, by enhancing the capacity of recipient households to invest in additional migration, education and other income-generating assets (Connell 1980; Shankman 1976). Some village-level studies have demonstrated considerable income inequality (Hardaker et al. 1987) and suggested that this is partly a result of remittance flows (Gailey 1992a; Small 1997: 134, 195). It is certainly a widespread perception; Marcus thus suggested, almost a decade ago, that ‘the capacity to call on international resources has become a crucial factor in influencing a family’s local economic conditions. The lowest stratum in contemporary Tonga are those totally dependent on the nation-state framework, and the limited resources it embodies, without any overseas options at all’ (Marcus 1993: 29–30). Indeed, it is increasingly argued, as in Tonga, that ‘every family needs to have someone overseas. Otherwise the family is to be pitied’ (quoted in Small 1997: 152). Hence, in contrast with Western societies, it is often the single-female-headed households that survive most effectively (Gailey 1992a). More recent empirical studies, however, based on survey data have tended to challenge this view, and some macro-economic data suggest that remittances have not led to increased inter-household income inequality, at least within Tonga (Ahlburg 1991). Ahlburg (1991, 1995) and Brown (1995) found that the distribution of household income with remittances was less skewed than the distribution without remittances, while other recent studies indicated that inequality was a function of many factors, of which the migration-remittance nexus might have been an unimportant part (Evans 2001; Muliaina 2001; Halatuituia 2001). The most recent studies of migration and remittances in Tonga and Fiji have shown that those households with migrants were more likely to have a higher income, independent of remittances, but the direction of causality was unclear (Brown et al. 2006). A consensus in very diverse circumstances is improbable.

A number of conclusions on remittances are, however, possible. Firstly, there has been a consistently substantial and growing volume of remittances especially in the Polynesian states (making up a significant part of national income, in excess of the value of exports and aid). Secondly, the use of these remittances has gone through a partial transition from consumption to investment, as many consumption goals have been met, at least in part. Thirdly, remittances have been particularly important in the most remote islands where development needs are less well met (and probably therefore reduce inequality). Fourthly, remittances contribute to valuable objectives such as human resource development, and are a means of maintaining social networks and creating social capital (Grieco 2003). In several contexts, especially in smaller islands, education is highly valued, in a general sense and for the development of specific skills (for example, in health provision), in order to create human capital for potential migration. Overall remittances are positive and satisfying for households but insufficient in and of themselves to influence national development goals. Fifthly, households seek to increase incomes by migration and remittance strategies, even by fostering obligations and ‘implicit contracts’. Even with imperfect knowledge, households are consciously making decisions in favour of the quantity and quality of education of children that boost their chances for migration and thus the supply of remittances (Brown and Connell 2004). Migration and remittances thus stem from and contribute directly and indirectly to human capital formation.