Migration decisions are usually shaped within a family context, as migrants leave to meet certain family expectations, the key one of which is usually financial support for kin. Migration has rarely been an individual decision to meet individual goals, nor has it been dictated by national interests (except perhaps in the case of Kiribati and Tuvalu). Migration is directed at improving the living standards of those who remain at home and the lifestyle and income of the migrants. In Tonga, ‘there are few opportunities for socioeconomic advancement … and migration is perceived as the only solution’ (Lee 2004: 135). In Samoa, the reason for migration was simply ‘to seek wealth for all’ (Muliaina 2001: 25). Consequently, ‘families deliberate carefully about which members would be most likely to do well overseas and be reliable in sending remittances’ (Gailey 1992b: 465; Cowling 1990). Through this process, extended households, as in Tonga, have transformed themselves into ‘transnational corporations of kin’, which strategically allocate family labour to local and overseas destinations to maximise income opportunities, minimise risk and benefit from resultant remittance flows (Marcus 1981). To an even greater extent, therefore, than for internal migration (where health, education and social reasons explain some part of migration), international migration is more evidently an economic phenomenon, though other factors are necessarily involved.

Developing this notion, Bertram (1986: 820) has suggested that migrant extended households are characterised ‘by remittance transfers among various component parts of the “transnational corporations of kin” which direct the allocation of each island’s family labour around the regional economy’, and in so doing not only help to maintain these family and communal networks but even enlarge their social fields of interaction, incorporating them into multi-local networks of support and empowerment. Similarly, for households in Samoa, ‘having young wage earners abroad diversified families’ earnings streams and reduced their dependence on high-risk activities. Having family members in several locations abroad diversified earning sources and reduced risk levels still further’ (Macpherson 2004: 168). Moreover, Macpherson went on to argue that families ‘using intelligence from migrants abroad, periodically surveyed risks and returns in various enclaves and encouraged others abroad to relocate in places in which returns were found to be higher and risks lower’.

In this way Samoans were, for example, encouraged to join the US military because jobs were assured, wages were higher and education could be obtained without loss of earnings. ‘If this analysis depicts Samoans as calculative and instrumental, it is because in relation to risk and return they are necessarily so … [as] risks and returns available in various places were formally canvassed and modeled by families’ (ibid.). While this form of household consensus certainly occurs, and demonstrates the significance of access to the migration-remittances nexus, it has been argued that applying the same kind of model in Tonga tends to portray families as being in agreement about their economic aims and functions, whereas there are often conflicts and tensions within them (Lee 2004: 136). In Kiribati and Tuvalu, where seafarers are away from their wives and parents and send remittances to both, there are frequent disagreements about their allocation and use (Dennis 2003: 35). There are household uncertainties about outcomes in various places. Moreover, more than a decade ago, James argued that in many Tongan villages remittances were becoming individualised and the idea of a transnational community of kin was becoming increasingly invalid (James 1993b: 361; Lee 2003: 31). The extent of greater individualisation is impossible to determine, but such conflicts over use emphasise, rather than downplay, the role of remittances.

Political factors have also been significant influences on migration in various contexts, notably in the migration from Fiji that followed the 1987 and 2000 coups, and in the skilled migration from Bougainville and the Solomon Islands during the recent crises. In Fiji, migration was not only of ethnic Indians, arguably most affected by the coups, but of Fijians, emphasising how migration could be seen as ‘a barometer of fear’ of further conflict (Narayan and Smyth 2003). Environmental factors have similarly influenced migration. There has been recent movement from such places as the Carteret Islands in PNG, where localised sea-level rise posed particular problems, and even more recently from Manam (PNG) and the northern islands of the Marianas, after volcanic eruptions. Cyclone Heta, which devastated Niue in January 2004, destroying almost one-quarter of the houses on the island, even prompted some thoughts about the permanent abandonment of the island (with the remaining 1,500 people following so many of their kin to New Zealand). Global warming poses a future threat to the islands, and especially to more than 100 populated coral atolls, should sea-level rise occur. In Tuvalu, fears are such that emigration has already occurred in anticipation of new difficulties (Connell 2003d). Island states are in no position financially, geographically or politically to defend themselves against such potential threats, but if — perhaps when — the worst does occur, Islanders might become a new stream of environmental refugees to metropolitan states.