The countries and territories of the South Pacific are often considered as not having population problems because of their relatively small size. But they actually face many population issues. For example, they still experience rapid natural population growth rates, which they try to reduce through international migration. Investigation of international migration in the South Pacific has long focused on Cook Islands, Niue, Tokelau, Wallis and Futuna Islands because these small island societies have a higher proportion of their populations living in New Zealand or in New Caledonia than at home.1 Perhaps one of the main reasons for the interest in this form of migration is that it is well documented. By contrast, data on migration from Samoa and Tonga are poor even though the number of Samoans and Tongans overseas is much higher in absolute numbers than that of Cook Islanders, Niueans and Tokelauans.

This article attempts to estimate the size of the largest flows of migrants in the South Pacific, mainly from the Cook Islands, Fiji, Samoa and Tonga, since 1980. Because the 1980s and early 1990s witnessed important changes in the economic situation and migration policies of developed countries in the Pacific, the article also attempts to determine the consequences of those changes on migration trends, characteristics of migrants, age-structure of island populations and island economies. Finally, it considers whether or not migration will continue to play a role in the future demographic and economic development of Pacific island countries.

_____________

* The author of this article may be contacted at INED (Institute National D'Etudes Demographiques), 27 rue de Commandeur, 75675 Paris Cedex 14, France. E-Mail: rallu@ined.fr. The research on which this article is based was conducted with the support of ORSTOM.

Owing to the poor quality of South Pacific migration data, it is necessary to use simultaneously migration statistics (arrivals and departures recorded by immigration services) and census data for both island countries and host countries.

Migration data

The migration data of Pacific island countries are particularly poor. For example, Tonga does not record departures of Tongan nationals, which means that migration data are simply non-existent for that country. A comparison of Samoan data on departures to Australia and American Samoa with Australian and American Samoan data on arrivals from Samoa shows that the former are highly unreliable. Fiji provides rather more reliable data; time series of departures from Fiji to Australia are consistent with arrivals in Australia. As for countries of the so-called "Pacific Rim", New Zealand data provide information only on citizenship since 1987 and the data are difficult to use for populations that include many persons holding dual nationality. Such persons are likely to go through immigration channels with one passport when they arrive and a different one when they leave the country. Australian data are available both by birthplace and citizenship. Data from the United States of America refer to immigrant permits by year, rather than to immigrants by year of entry (which is not the same for legalized migrants). Further, there are no data on departures.

Census data

Underenumeration of Pacific islanders in New Zealand censuses has been well assessed since the 1970s (Bedford, 1983); 1991 New Zealand census data show that the number of island-born Samoans and Tongans enumerated in 1991 who declared that they had already resided in New Zealand in 1986 is higher than the number enumerated in 1986 by 20 per cent for Samoans and by over 70 per cent for Tongans, at ages 20-24 years. It is not probable that the migrants exaggerated how long they had been in-country as a mean of justifying their status. In New Zealand, most illegal immigrants are overstayers whose passports were stamped on arrival; therefore, the date of their entry is known. The only way for illegals to become legal migrants is to regularize their situation, as had occurred in 1988 and 1989. Most of the overstayers who entered the country prior to 1988 should therefore have had their status regularized; those whose status was not regularized most probably escaped enumeration in the 1991 census.

A comparison of net migration with intercensal changes in the population of Pacific islanders in Australia also shows underenumeration, mainly for Samoans, with 5,800 persons enumerated against 8,600 expected based on the figure in 1986 and net migration during the period 1986-1991. Underenumeration of Pacific islanders affects United States censuses as well. According to the 1990 census, there was a decline of 1,000 persons since 1980 in the number of foreign-born Samoans living in the United States compared with an inflow of 2,743 immigrants during the period 1980-1989. Data are apparently more consistent for Tongans for whom there was an intercensal increase of 5,500 against 5,442 immigrants recorded in 1990. But these figures do not include the IRCA (Immigration Reform and Control Act of 1986) adjustment of 827 Samoans and 3,522 Tongans (Ahlburg, 1991), and illegals not yet regularized. There are no data on outmigration from the United States, but it probably is not significant.

Underenumeration of Samoans and Tongans in the 1990 census certainly occurred, but it might have been slightly smaller than suggested by the aforementioned figures. The censuses of island countries are not always reliable. For example, it has been acknowledged that Samoa's 1986 census was not complete, even though matais (chiefs of extended families) acted as enumerators (after teachers refused to do so because of the wages offered for the work). Further, Fijian and Tongan censuses provide information only every tenth year, an intercensal period which is too long for making a useful study of migration.

Migration in the Pacific subregion does not include large numbers of illegal arrivals such as would be the case in Europe and the United States. However, although the number of stow-aways is normally small, illegal entry still occurs; for example, illegal entry to Australia from Papua New Guinea through the Torres Strait might be more frequent that elsewhere in the subregion. Australian and New Zealand data show that since 1980 most of the net migration consists of short-term movements and this has been an increasing trend since 1986. An exception was Fijian migrants of ethnic Indian origin who from 1988 migrated to Australia and New Zealand mainly with long-term visas. In 1987, a coup d'etat in Fiji resulted not only in a sudden change of government administration but also of policy towards this ethnic group. Thereafter, most Fijians of Indian origin entered the aforementioned countries with short-term visas (Rallu, 1994). Thus, illegal migrants in Australia and New Zealand are mainly overstayers whose arrivals had been recorded. The main pattern of illegal migration of islanders appears to be: arrive with a short-term visa, overstay, avoid enumeration by not answering the census questionnaire, later regularize one's status, and answer the census questionnaire.

Size of flows in the 1980s

Total migration from island countries and territories has been estimated using migration data or census data of countries of the Pacific Rim (table 1). Almost 18,000 Tongans would have left Tonga in the 1980s, which represents an annual net loss rate of 1.9 per cent. As departures were concentrated in the second half of the 1980s, net losses were higher than the natural growth rate of 2.2 per cent, thus, population declined in Tonga during this period.2 Fiji shows the highest number of migrants, i.e. 50,000 or more according to available sources of data, but for the decade the annual net loss rate to Fiji was only 0.7 per cent. However, the rate exceeded 1 per cent during the second half of the 1980s; it was even higher for the Indian population of Fiji which declined from 349,000 at the end of 1986 to 338,000 in mid-1989, stabilizing thereafter.

Results for Samoa give two rather different estimates, i.e. a net loss of 16,000, or 23,500 for the 1986-1991 period, depending on whether census or migration data, respectively, are used. Given the level of natural growth in Samoa, the increase in the "Samoan" population from 156,000 in 1986 to 158,100 in 1991 would allow for emigration of only 18,400 persons during the period. Thus, the figure of a net loss of 23,500 means that the population growth rate would have to have declined, or it implies that the 1986 census of Samoa would have to have underenumerated the population by about 5,000 persons, with no increase in the population occurring between 1986 and 1991. If underenumeration at the 1991 census was actually fewer than 5,000 persons, then the population of Samoa would have declined too. Cook Islands experienced a small level of emigration owing to the economic situation in New Zealand, and noted a return migration of 700 persons. The population in the islands increased by 1 per cent yearly, according to census data of 1986 and 1991 which are consistent with the rate of natural increase and net migration.

Destination	Places of origin
	Cook Islands	Fiji	Samoa	Tonga
American Samoa
1986-1991 Census	-	-	5,030	510
Australia
1981-1986 Census	820	5,394	2,200	1,856
1981-1986 Immigration	-127	8,078	2,280	2,300
1986-1991 Census	850	15,350	2,800	1,730
1986-1991 Immigration	875	18,220	5,380	2,110
New Zealand
1981-1986 Census	1,830	870	9,950	1,990
1981-1986 Immigration1	-	1,695	9,290	1,520
1986-1991 Census	-270	9,470	9,250	5,850
1986-1991 Immigration	-	11,375	11,400	5,620
United States (USA) and Canada
1980-1990 Census (USA)	-	8,400	-1,000	5,500
1980-1990 Immigration (USA)	-	9,209	2,743	5,442
1980-1990 Census (Canada)	-	10,350	-	-
Totals for places of origin
1980s Census	3,230	49,834	28,230	17,436
1980s Immigration	-	58,118	32,380	17,560
1986-1991 Census	580	-	16,0802	-
1986-1991 Immigration	-	-	23,4713	-

Notes: 1. New Zealand migration data are by citizenship.

2. Assumes that the decline in the United States was between 1986 and 1990.

3. With 1,661 immigrants in the United States during the period 1986-1990 (fiscal years).

Recent trends in migration

Despite the economic recession in New Zealand and Australia during the second half of the 1980s, it appears that migration from Polynesian countries did not decline significantly for the period as a whole, but changes in migration policies had an important effect on the timing of migration. The visa waiver experiment from December 1986 to March 1987 in New Zealand caused mass arrivals, primarily of Samoans and secondarily of Tongans and Fijians. Then restrictive migration policies resulted in a decline in Samoan migration to New Zealand, followed by peaks in their migration to Australia in 1988 and American Samoa during the period 1989-1990. Since 1990, there has been a net loss of Samoan citizens in New Zealand, but this situation probably does not reflect migration trends only, as the number of migrants with dual nationality was increasing and the way they use either of their passports may change. Except for a drop in 1991, net gains in the number of Samoans migrating to Australia were rather steady during the period 1989-1993, but they comprised mainly New Zealand citizens. Tongan migration to New Zealand and the United States increased significantly during the second half of the 1980s, but net losses have appeared in New Zealand since 1990 and a slight decline has occurred in their migration to Australia since 1989. The second half of the 1980s also witnessed important levels of emigration of Fijians of Indian descent following the political uncertainties in that country during 1987; however, emigration of ethnic Fijians also increased. The trends reversed since 1989; according to Fijian data, emigration of Fijians of Indian descent as well as ethnic Fijians has been declining, as has been the case with other islanders.

From the viewpoint of Pacific Rim countries, migration from the South Pacific to Australia increased during the period 1986-1991 mainly owing to migration from Fiji and secondarily from Samoa (table 1). The Trans-Tasman Agreement between New Zealand and Australia allows free movement of residents between the two countries. Data by citizenship show that 62 per cent of Samoa-born migrants arriving in Australia are New Zealand citizens whereas the figures are only 20 per cent for Tonga-born migrants and 7 per cent for the Fiji-born migrants. In 1991, the net migration of island citizens was negative for Samoans and Tongans.

For New Zealand, immigration from Fiji became as important as that from Samoa during the period 1986-1991, a time when Tongan migration also increased. Migration from Fiji and Tonga to the United States has increased. A noticeable point is the increase in Samoan gains in American Samoa. This could be related partly to the longer procedures required to enter the United States, but arrivals in American Samoa actually increased as a result of the difficulties Samoans faced in entering Australia and New Zealand.

Migration in the South Pacific during the 1980s has retained one of its main characteristics of very low migration from western Melanesia and Kiribati (although Micronesian island territories affiliated with the United States experienced high rates of emigration) and significant migration from the Central and Eastern Pacific. With more and more islanders taking advantage of the Trans-Tasman Agreement by migrating from New Zealand to Australia or leaving directly for Australia and the United States, an extension of migration networks, beyond specific links to former colonial powers, appears to have developed towards the "general circulation" of people (Chapman, 1985). Although Melanesian migration has so far remained insignificant, a big change for the future of migration in the South Pacific will be the beginning of significant migration flows from Melanesia, prompted by the pressures of unemployed youths and urbanization. The number of such migrants could be considerable owing to the size of the populations in Papua New Guinea and Solomon Islands.

Participation and unemployment rates

Censuses provide detailed information on the employment status of migrants. However, under-enumeration probably affects migrants differently according to employment status and occupation. Those who escape the census are mainly overstayers who were not yet able to regularize their status, i.e. mainly persons not in the labour force, unemployed or in lower level occupational groups.

Nevertheless, changes in labour-force participation have been extremely rapid, reaching very high rates of non-participation and unemployment in Australia and New Zealand in 1991. In New Zealand, the proportion of persons aged 20-49 not in the labour force increased from 7 per cent to 25 per cent for island-born males who already resided in New Zealand in 1986. Increases were relatively smaller for females, but the levels were higher, i.e. increasing from about 35 per cent in 1986 to between 45 per cent and 50 per cent in 1991. Unemployment rates also increased steeply from about 7 per cent to 20 per cent or more for males during the period 1986-1991 (table 2); the increase for females was from 13 per cent to 17 per cent. It may be observed that levels of non-participation in the labour force and unemployment rates are about 3-5 percentage points higher for migrants who were not residing in New Zealand in 1986 (recent migrants). Thus, the proportion of males aged 20-49 years currently employed declined from 85 per cent in 1986 to 59 per cent in 1991 for migrants who had already resided in New Zealand in 1986; for females in the same age group, the proportion declined from 56 per cent to 43 per cent. The proportions employed were only 48 per cent and 34 per cent, respectively, for recent male and female migrants. New Zealand-born Pacific islanders adjusted much better than the island-born to the new labour market situation, with only 35 per cent of males and 49 per cent of females not being employed.

	Island-born				New Zealand-born
	Cook Islanders	Fijians**	Samoans	Tongans	Cook Islanders	Samoans	Tongans
Australia*
1986 Males	16.7	10.5	17.9	17.2
1986 Females	28.9	13.0	21.9	21.5
1991 Males	26.2	14.7	32.0	23.8
1991 Females	29.8	18.2	33.1	26.0
New Zealand
1986 (ages 20-49) Males	5.9		7.0	9.1	8.8	8.2	7.8
1986 (ages 20-49) Females	12.9		11.9	14.3	15.6	9.5	8.3
1991 (ages 20-49) Males	20.0	12.7	20.0	22.0	20.4	19.0	16.7
1991 (ages 20-49) Females	17.4	16.5	16.7	18.5	20.3	18.5	17.3
	Foreign-born						Native Samoans
	Melanesians		Samoans		Tongans
United States
1980 Males	5.2		9.6		7.9		9.0
1980 Females	2.3		9.8		4.3		10.7
1990 Males	7.4		8.8		9.5		9.6
1990 Females	6.1		9.8		10.8		11.2

Notes: * For Australia, data are by place of birth; and

** Data for Fiji include Fijians of Indian descent.

In Australia,3 unemployment rates among Pacific islanders reached similar or even higher levels in 1991 than in New Zealand, with around 25 per cent each of the male island-born Tongans and Cook Islanders being unemployed, and 32 per cent of the male island-born Samoans; the respective data for females are 26 per cent, 30 per cent and 33 per cent unemployed. However, it should be mentioned that unemployment rates of Pacific islanders were already high in 1986, i.e. around 17 per cent for males and 22 per cent for females.

In the United States, increases in the unemployment rates in 1990 for Pacific islanders had been moderate, generally remaining under 10 per cent, with only small differences observed between native and foreign-born. However, native Samoans (including American Samoans born in American Samoa) have not performed as well as foreign-born Samoans (all of these being Western Samoans) who were selected by immigration.

In the second half of the 1980s, a significant decline occurred in the Samoan-born labour force employed in New Zealand, which was not compensated for by increases in Australia and the United States, given the probable levels of underenumeration. The Samoan-born labour force overseas increased only as a result of increases in American Samoa. The Tongan-born labour force overseas increased in all countries of the Pacific Rim, owing to significant emigration during the period 1986-1991. Also, there was a decline in the overseas labour force of Cook Islanders.

Occupational distribution

The steep declines in the labour force employed in New Zealand were a result of economic restructuring, with automation replacing unskilled labour. Pacific islanders were particularly affected by these changes as most of them (around 80 per cent) had been employed as labourers at the beginning of the 1980s. The restructuring of the New Zealand labour force witnessed an increasing proportion of persons employed in higher occupational categories. The Pacific islanders who were still employed in 1991, followed this trend. There was an even slightly higher relative increase in the number of managerial, administrative, professional and technical employees among island-born than New Zealand-born Pacific islanders. Nevertheless, the proportion of island-born employees in these occupational categories at around 9 per cent in 1991 remained much lower than that of New Zealand-born Pacific islanders at almost 30 per cent for Tongan and Samoan males (table 3). The proportion of clerks was stable for males but increased for island-born females, whereas the proportion of production workers and sales and service workers declined steeply. Some of the sales and service workers might have taken higher level jobs, but most of the production workers became unemployed. Nevertheless, it is worth noting that migrants took action in line with the restructuring of the New Zealand economy with some of them attaining high-level occupations. However, as late as 1991, 75-80 per cent of the island-born population still comprised production workers. With regard to occupational distribution, islander migrants remain well behind all European New Zealanders, New Zealand Maoris (although New Zealand-born islanders are doing better than the latter group) and South-East Asian migrants (26 per cent of ethnic South-East Asian migrants works as managers, administrators, professionals or technicians compared with 10 per cent of ethnic Pacific islanders). It will be more and more difficult for Pacific islander migrants to compete for jobs with the better educated Asian migrants.

In Australia, the proportions of islanders who are working in both higher level occupations and production are not much different from those in New Zealand. But in contrast to New Zealand, no important change towards higher level occupations occurred in Australia. On the contrary, for most islanders, the proportions of managers, professionals, technicians and administrative workers were lower in 1991 than in 1986; as for production workers, there has been no clear trend. Analysis of migration data to determine the occupation of migrants shows that long-term migrant islanders who hold New Zealand citizenship and have free access to Australia had lower level occupations than island citizens who were selected by immigration (Rallu, 1994). The decline in the proportion of islanders in higher occupations in Australia also reflects an increasing level of short-term migration which escapes selection with the migrants taking lower level occupations.

Occupation	Island-born								New Zealand-born
	Fijians**		Cook Islanders		Samoans		Tongans		Cook Islanders		Samoans		Tongans
	M	F	M	F	M	F	M	F	M	F	M	F	M	F
Australia*
1986 MPPTA	8	12	29	23	24	25	9	17
1986 Clerks	5	15	8	30	6	23	2	13
1986 Sales/services	4	13	7	16	5	12	2	8
1986 Production workers	79	59	53	28	62	38	82	58
1991 MPPTA	5	11	27	20	9	16	9	16
1991 Clerks	4	16	8	26	4	17	3	15
1991 Sales/services	3	12	7	16	3	13	2	9
1991 Production workers	77	51	50	30	72	42	74	46
New Zealand
1986 MPPTA	3	6			4	7	4	8	7	10	13	15	18	16
1986 Clerks	4	13			6	15	3	10	7	35	13	46	6	38
1986 Sales/services	6	31			7	29	7	28	12	21	17	20	16	28
1986 Production	83	47			79	45	82	50	73	32	56	18	58	17
1991 MPPTA	10	13	26	22	9	13	8	12	20	24	28	30	31	31
1991 Clerks	4	17	6	31	6	19	3	13	7	35	13	43	8	36
1991 Sales/services	5	19	14	26	6	18	4	19	7	12	13	12	11	14
1991 Production workers	79	48	48	17	74	45	82	52	63	27	45	12	46	16

	Foreign-born						Native
	Melanesians		Samoans		Tongans		Melanesians		Samoans		Tongans
	M	F	M	F	M	F	M	F	M	F	M	F
United States
1980 Managers and professionals	12	11	12	13	11	11			12	14
1980 Tech/sales	10	5	6	10	6	9			5	15
1980 Adm. supply clerks	10	16	9	26	5	23			13	30
1980 Service workers	20	41	17	24	22	33			18	22
1980 Farm/fisheries orkers	4		3		10				3
1980 Production workers	43	24	53	26	45	24			50	19
1990 Managers and professionals	12	10	15	13	8	8	35	11	13	14	16	22
1990 Tech/sales	12	16	7	18	3	10	0	13	8	16	8	17
1990 Adm. supply clerks	6	27	11	29	5	22	9	22	11	36	13	26
1990 Service workers	25	31	17	28	18	37	45	49	19	19	23	25
1990 Farm/fisheries workers	3		3		17		0		2		4
1990 Production workers	42	16	48	13	49	22	12	5	48	15	37	10

Notes: * For Australia, data are by birth place; and

** Data for Fiji includes Fijians of Indian descent. Totals do not add to 100 owing to number "not stated".

Abbreviations: MPPTA = managers, professionals, para-professionals, technicians and administrative staff.

Occupational classification in the United States is different from that of Australia and New Zealand. In the United States, the proportions in managerial and professional occupations in 1990 reached 15 per cent for foreign-born Samoan males and 12 per cent for Melanesians but only 8 per cent for Tongans. Except for Melanesians, these levels were higher than in Australia and New Zealand, despite the fact that these categories are more restrictive in the United States. Craft workers, precision workers, operators, fabricators and labourers, categories comparable to production workers in Australia and New Zealand, account for less than 50 per cent of Pacific island employees in the United States, a much smaller proportion than in Australia and New Zealand. Seventeen per cent of Tongans in the United States are classified as agricultural and fishery workers.4 In contrast to New Zealand, a change in the occupational distribution of Pacific islanders in the United States between 1980 and 1990 was towards more production workers, especially for Tongans. The proportion of technicians, sales staff and administrative support workers in the United States declined steeply for Tongans and Samoans in 1990 as did the proportion of service workers among Tongans. For Melanesians in the United States there were only minor changes in occupational distribution. Data on native Samoans in the United States include American Samoans born in American Samoa, which might be the reason why native Samoans do not perform much differently from foreign-born Samoans, whereas in New Zealand they do.5

In general, skilled Pacific islanders prefer to migrate to the United States, which partly explains the high proportions of islander production workers in New Zealand and Australia. However, the proportion of islander production workers in the United States increased for Tongans who in recent years have stepped up immigration to that country.

Increasing difficulties for Pacific islanders to find jobs in host countries of the South Pacific subregion should result in an increasing level of return migration.

Return migration

No data on return migration are available for the 1970s or early 1980s; however, data for the mid-1980s and early 1990s do not show significant return migration.

Return migrants have been defined as island-born persons enumerated in island countries who reported that they resided overseas when questioned about their previous residence. Return migration rates have been calculated as the ratio of return migrants to the island-born population enumerated (or estimated) in Pacific Rim countries at the date referenced in the question on previous residence (table 4). In 1986, rates of return migration were about 10 per cent in Fiji and Tonga. However, after 1986 emigration increased and net losses became even more important in these countries (see table 1) yielding further potential for return migration. In the early 1990s, the proportion of return migrants was around 5 per cent in the Cook Islands and American Samoa.6 Both societies experienced mass migration to countries to which their citizens have free access, and this is probably the reason why rates of return migration are lower in the latter than in Fiji and Tonga.

Country/territory	Census year	Date of previous residence	Return migrants	Island-born population in Pacific Rim countries and territories1	Return migration rate (%)	Proportion of return migrants2
American Samoa	1990	1985	720	18,000	<4.04	1.5
Cook Islands	1991	1986	858	17,800	4.8	5.2
Fiji	1986	1981	2,369	23,500	10.1	0.3
Tonga3	1986	1982	1,722	18,800	9.2	1.8

Notes: 1. Island-born (or ethnic Tongans) enumerated or estimated in Australia, New Zealand and the United States at the date of previous residence.

2. In island population.

3. Ethnic Tongans.

4. Estimated.

Despite the low rates, return migration to the Cook Islands -- simultaneously with lower emigration -- has had a significant impact on population trends and age-structure. It should be noted that estimates of departures from New Zealand are twice as high as return migrants enumerated in the islands, which means that one out of two migrants who left New Zealand went to Australia or the United States. This situation is worth further study as it may affect Samoa and Tonga in the near future.

The shape of age-pyramids in island countries and territories has been affected by emigration as can be seen by the marked bottle-neck from ages 20-25 (figure 1). This aspect was very typical in the Cook Islands in 1986. Reduced net losses during the five years preceding the 1991 census caused increases in the population aged 20-24 and 25-29, and return migration also appears after age 30. Actually, 67 per cent of the return migrants were aged 20-49, with a rather even distribution between these ages. As a consequence, the proportion of adults increased from 54.6 per cent in 1986 to 57 per cent in 1991. Despite higher numbers of births in 1991 owing to the increasing number of women entering the reproductive age groups, the proportion of children under 15 years of age declined from 38.4 per cent to 35.5 per cent. Age-dependency ratios also declined from 0.833 to 0.754.

Employment

The re-shaping of the age-pyramid and decline in age dependency should have positive effects on island economies and overall development, provided the increasing number of adults can find jobs. According to the 1991 census of the Cook Islands, most return migrants in Rarotonga, the main island where the formal sector is more developed than in outer islands, were employed, with the rates of employment being 80 per cent for males and 63 per cent for females. However, these figures were 10 percentage points below the employment rates for the total population. Moreover, despite the Cook Islands having had a rather dynamic economy during the second half of the 1980s compared with other South Pacific island countries, the ratio of the population employed (by age) to the total population (figure 2) declined from 1986 to 1991, mainly at ages under 35 for both males and females.7 A similar trend was seen in Samoa throughout the 1980s with regard to the population employed in the formal sector. Under such conditions, the favourable impact that a reduced level of migration is supposed to have on an economy becomes highly questionable.

The relationship between migration and development needs to be re-conceptualized somewhat, at least for South Pacific island countries. Migration increased steeply after Samoa attained independence in 1962 and it started after Tonga attained independence in 1970. This phenomenon appeared to be a consequence of poor employment opportunities in the islands following independence, with no new jobs being created. The historic cause of mass migration from island countries and territories and the recent difficulties island economies have had in providing jobs for a slowly growing working-age population would seem to support the view of those saying that migrants can find in Pacific Rim countries the jobs they cannot find in their islands and that migration is well compensated for by remittances. Therefore, a lower level of migration would have rather negative effects on island economies and their way of life. How much is this still true?

Monetary flows from external sources

In the 1960s and 1970s, Polynesian countries developed economies based on external sources of money, remittances and aid, often referred to by the acronym MIRAB (migration, remittances, aid and bureaucracy) (Bertram and Watters, 1985). Remittances and ODA (official development assistance) represent a high proportion of GDP (gross domestic product) (table 5). Remittances in kind (either new or used goods) and unrecorded remittances represented up to 57 per cent of total remittances in 1993 (Brown, 1994) and this proportion might have increased recently owing to high unemployment rates among migrants and their difficulty in sending money home. Under these conditions, using published data, it is not easy to measure the impact of overseas unemployment on remittances. As a consequence of a still growing labour force employed overseas,8 remittances had been increasing in Samoa until 1991 (latest available data). Remittances increased in Tonga from T$30 million (1 pa'anga (T$) = US$0.69 in 1987 and US$0.80 from 1988 to 1991) during the period 1987/88-1988/89 to T$38 million for the period 1989/90-1991/92; however, they declined slightly during the latter period, i.e. from T$39.6 million in 1989/90 to T$38.5 million in 1991/92 (SPC, 1993).

	GDP (per capita)	Percentage of annual GDP growth (1986-1990)	Remittances (% of GDP in 1989)	ODA (per capita in 1989)	Balance of trade (millions of US$)	Budgetary deficit (millions of US$)
Cook Islands	3,416			730	-46
Fiji	1,767	2.3	*	64	-254	35
Samoa	750	0.6	36	210	-81	22
Tonga	1,038	0.5	33	287	-51	28

Source: South Pacific Commission (1993), ESCAP (1993).

* Note: In Fiji, net private remittances are negative.

In the near future, the labour market in countries of the Pacific Rim will not become more favourable for Pacific islanders and negative net migration of Samoans and Tongans in New Zealand will not always be compensated for by migration to American Samoa and the United States. Eventually remittances will level off, or even decline. International aid is also likely to decline as a result of the end of the Cold War and "donor fatigue", among other factors (Australian Minister of External Aid and Pacific Islands Affairs in a Conference at the University of New South Wales, September 1994). Also, Pacific island countries have already received very high levels of aid per capita (table 5). Considered together, both remittances and external aid are likely to stabilize or decline.

External solutions

For historical reasons dating back to the arrival of Europeans in the subregion, island people -- not to mention Governments -- would seem to be under the impression that development comes from abroad, so they still tend to depend on external sources of aid. In the face of high unemployment rates overseas, some countries have embarked on steps that will increase the "brain drain"; the thinking is that professionals who emigrate are less likely to be unemployed in host countries and persons of such calibre should be able to send home significant amounts of money. Samoa's Seventh Development Plan contains a section dealing with "human capital exportation". Recognizing that labourers have increasing difficulties in finding jobs overseas, the Plan states that it is possible "to embark on a conscious policy of exporting skilled labour in order to ensure a continuing flow of remittances".9 This attitude is not sustainable as it does not take into account the negative aspects of the brain drain at home, i.e. the necessity either to employ costly expatriates to replace professionals who migrate, or leave positions in the public sector vacant or occupied by underqualified staff. Nor is it consistent with the bonds placed on government-supported students to repay grants, unless there are provisions that such students educated overseas would have to work in the islands for a few years before migrating. As for producing highly qualified professionals or technicians in small island countries, this would not seem possible because of their severe resource constraints and limited opportunities for such high-level training. Actually several Governments of countries in the subregion do not seem really concerned about the brain drain, since it is seen as part of a more basic economic process based on monetary inflows from overseas. There are other dimensions to the problem as well. For example, Fiji would seem to place ethno-political considerations before the negative effects of the brain drain. Witness the extremely high brain drain that occurred during the period 1987-1991, with about 50 per cent of managers and 15 per cent of professionals, technical, clerical and related workers, mainly Indians, emigrating (Chetty and Prasad, 1993). To replace those lost in that brain drain, Fiji has been attempting to attract Asian migrants in the expectation that they would bring in valuable skills, make investments and generate new jobs for local people.

Another way for Pacific island countries to increase the level of remittances would be to send out more migrants in the future, which would imply fostering higher fertility. Since the mid-1980s, fertility in most island countries has been either relatively stable or on the increase, with total fertility rates being around 4.5 children per woman, thus delaying the completion of the fertility transition. Such fertility might also be a compensation for emigration. However, it is noticeable that societies such as American Samoa and the French Overseas Territories, which have different economies, still show declining fertility.

Such solutions are based on mass migration (or transmigration) and presuppose that migration would be free between island countries and countries of the Pacific Rim. There is a strong misunderstanding of the situation by Governments of island countries. It is true that migration might not have the same negative effect on development in situations where economies are profoundly inter-related as is already the case in the Pacific, with important flows of goods and remittances. However, a recent problem, which would occur whenever migration becomes free, is that of integration of migrants on the labour markets of Pacific Rim countries. The reasons why islanders cannot compete with other migrants from South Asia or South-East Asia have their roots in the island societies themselves.

Socio-political causes of economic difficulties

The main problem appears to be the educational system in island countries, which at face value would seem to be a controversial statement since the Pacific islands enjoy high literacy dating back at least to the end of the Second World War if not to the beginning of this century owing to the efforts of mission schools. However, in the 1970s and 1980s, the educational systems in the Pacific islands did not follow the trends in Asia. As a result, students in the Pacific complete secondary-level courses with lower achievement in mathematics and the sciences, sometimes also in English; moreover, because of the quality of such preparatory courses, they have difficulties in coping with the advanced level of studies found at overseas universities and at the University of the South Pacific in Suva, Fiji. Further, the public school system is not very strong in Fiji; only one school is considered to be of an international standard. Private schools, many of them religious, are often weaker still, with a student's success often strongly related to various non-academic considerations. By contrast, in Fiji, Indian schools are considered better than public and private schools.10 In Samoa and Tonga, more needs to be done to upgrade teaching materials and attract qualified staff for both public and private schools. In Samoa, only 2 per cent of children who entered form 2 in 1979 passed the school certificate level in 1982 and only 1.5 per cent of the same cohort subsequently passed the university entrance examination (Gannicott, 1993). In Tonga, between 20 per cent and 40 per cent of school children are taught to the required standard of primary education (Gannicott, 1993).

Except for the University of the South Pacific, post-secondary education institutions in island countries are weak: "Many of these institutions barely reach international secondary (school) standards" (Hughes and others, 1990). It remains to be seen whether or not island countries can afford local tertiary education, the per unit costs of which are already very high owing to the small size of the national populations in the Pacific. Rather than spend money on setting up local tertiary educational facilities that would face extreme difficulties in attaining a high standard, it would be wiser for Pacific island countries to upgrade their primary and secondary schools to enable more graduates to pursue courses overseas. Actually, most of the Cook Islanders who pursue tertiary education do so in Australia, Japan and New Zealand, where selection is competitive. As a result, the Cook Islands is considered to have better qualified civil servants than those of many other island countries and territories.

In general, it appears that most of the difficulties with educational systems can be traced to familial, social and political structures. Another factor that is acknowledged as hindering social and economic development in Pacific island countries is their political and administrative systems. Some are dominated by noble families and chiefs or aristocrats and religious groups on one hand, or by certain ethnic groups on the other.

Maintaining the island way of life

Economic growth in Pacific island countries during the 1980s has been limited, with the overall growth in GDP per capita being about 0.1 per cent annually (UNDP, 1994). In Fiji, after virtually no growth until 1986, GDP declined following the events of 1987; a substantial recovery occurred in 1989 (11 per cent) but the overall result for the 1980s was an annual growth rate of 0.8 per cent in GDP per capita. During the period 1991-1993, yearly growth was around 0.3 per cent. GDP per capita based on published population figures for Samoa and Tonga should be corrected for population decline (see above) and should be around 0.5 per cent during the period 1985-1990, which is extremely low compared with island countries in the Indian Ocean (2.4 per cent) or Caribbean (3.7 per cent) (UNDP, 1994) where external aid per capita is much lower. However, it must be recognized that the remoteness factor and transport costs are higher in the Pacific.

Thus, recent economic performance is threatening the island way of life. Urgent restructuring of island economies is necessary. In view of the limited size and fragile environment of island countries, economic development should be oriented towards human resources, by bringing up to a high standard the qualifications of national populations. With the availability of new technologies, developing countries have a new opportunity to progress. Industries using high technology, producing high-value-added products and those having almost no adverse impact on the environment are the types suited for promotion in island countries. For example, production of electronic components and, preferably, of software would be considered as ideal "industries" by some island countries. But such enterprises require a highly skilled work force, which as previously discussed is lacking in Pacific island countries. Therefore, educational systems should be improved rapidly, beginning at the primary and secondary levels. This implies that the profession of teaching must be made attractive to qualified persons. Education is also the means to change social structures, if it is granted that every child should be given an equal chance to progress. Public service should also be restructured, with selection being based on merit only. Privatization should also be implemented rapidly and more incentives should be given to private enterprise.

Migration of Pacific islanders was high during the second half of the 1980s, but economic difficulties and restrictive migration policies in host countries have resulted in (a) net losses of islanders from New Zealand since 1990, (b) the extension of migration towards Australia, the United States and American Samoa and (c), to a lesser extent, return migration.

Lower net losses would reduce the depletion of younger adult cohorts in population pyramids and age dependency ratios would decline, although the island economies are not yet able to provide jobs to a steadily growing population. Owing to high remittances from the large population of islanders overseas, South Pacific Governments still favour migration, including an increased brain drain, and ask for free access for migrants to developed countries of the Pacific Rim. However, the problem is not so much restrictive migration policies of host countries as it is integration of migrants in those countries' labour markets. Most islanders are not able to adjust to changes in the labour force and to compete with better qualified Asian migrants. Finally, many of the difficulties they face appear to stem from weak educational systems in the island countries in relation to traditional social and political structures. Further, it would not seem to be possible for a country to rely on the export of unskilled labour for the purpose of ensuring national economic development. The maintenance of the island way of life will be possible only if improvements are rapidly implemented in the educational systems and if the private sector is developed further. Attachment to the "Pacific way" should not mean that traditional institutions dominate education and the economy to the detriment of the Pacific islanders themselves.

Policy makers in the Governments of the small and large island countries should be aware that sustainable development cannot be based mainly on migration and remittances. On one hand, emigration of an unskilled labour force is a "dead-end" in view of the rising qualifications of labour forces world-wide and subsequent competition with better qualified migrants from other countries; the lowest level unskilled jobs will be taken by migrants from poor countries accepting very low wages. On the other hand, emigration of a skilled labour force, i.e. the brain drain, can result in serious shortages of qualified staff in Pacific island countries, which in turn will adversely affect the education of the younger generations. The strongest basis for the socio-economic development of Pacific island countries is raising the qualifications of their human resources through improved universal education and equal opportunity employment, without regard to gender, and social and ethnic origins. Migration without human resources development will result in a loss of advantage for Governments and poor economic prospects for the people and migrants themselves.

1. We should question if this form of "migration" is really international migration at all, as Cook Islanders, Niueans and Tuvaluans hold New Zealand citizenship and Wallisians and Futunians are French citizens.

2. Actually the Statistical Abstract 1989 of Tonga shows declining population estimates from 1986 but, surprisingly, the 1993 issue of the Statistical Abstract assumes 0.6 per cent growth per annum since 1986.

3. Australian data on labour force participation by age group are not available. The Ethnic Communities Package, 1991 Census, Australian Bureau of Statistics, Canberra, Australia.

4. Proportions of agricultural workers among islanders in Australia and New Zealand are below 1 per cent, and have been included with production workers.

5. It should be pointed out that the number of native adult Melanesians and Tongans is very small and the results could therefore be affected by random variations.

6. As no data are available separately for Western and American Samoans in the United States, the number of the latter have been estimated, considering also that all American Samoans are considered to be "native" by the United States.

7. A similar trend in age-structure and youth unemployment was observed in the French Overseas Department, Martinique, Guadeloupe and Reunion between 1982 and 1990 (Domencah-Picouet and INSEE Reunion).

8. This includes American Samoa.

9. These positions contained in paragraphs 4.17-4.20 of the Plan might also be consequences of the fact that it seems difficult and socially unfair "to improve prospects for skilled and professional workers ... while holding unskilled workers' wages down to promote industrial development" (op. cit., 4.10).

10. In Fiji all schools are officially open to all ethnic groups.

Ahlburg D. (1994). "Return migration from the United States to American Samoa: evidence from the 1980 and 1990 censuses", Pacific Studies, June, 17(2).

__________ (1991). Remittances and their Impact, a Study of Tonga and W Samoa (Canberra: NCDS, Australian National University).

Bedford, D. (1991). "Ethnicity, birth place and nationality: dimensions of cultural diversity", New Zealand Population Review, 16(2).

__________ (1983). "Net migration and Polynesian population growth in New Zealand, 1971-1981", New Zealand Population Review, 9(2):18-47.

Bertram G. and R.F. Watters (1985). "The MIRAB economy in the South Pacific microstates", Pacific Viewpoint, 26(3).

Brown, R.P.C. (1994). "Remittances, savings and investments of South Pacific migrants", International Labour Review, (3).

Chapman M. and P.S. Morrison (eds.) (1985). "Mobility and identity in the island Pacific", Pacific Viewpoint, April, 26(1):1-371.

Chetty, N.K. and S. Prasad (1993). Fiji's Emigration. Demographic Report No. 4, SSED (Suva, Fiji: University of the South Pacific).

Domenach H. and M. Picouet (1992). "La dimension migratoire des Antilles", Economica, Paris.

ESCAP (1993). Pacific Island Countries: Economic Performance and Selected Issues in Policy Management and Adjustment, ST/ESCAP/1353 (New York: United Nations) 206 p.

Gannicott, Ken (1993). "Human resource development in the South Pacific". In: The Future of Asia-Pacific Economies, Pacific Islands at the Crossroads? (Canberra: NDCS, Australia National University).

Gubhaju, B. and N. Navunissaravi (1989). "Trends in fertility and mortality in Fiji based on the 1986 census, Asia-Pacific Population Journal, 4(4):45-66.

Hughes H., D. Ahlburg and Sun Hee Lee (1990). Human Resources Development in the Pacific Islands Developing Countries. High-level Seminar on Population and Development for Government Ministers, South Pacific Commission, Noumea, New Caledonia.

McCall, Grant and John Connell (eds.) (1993). A World Perspective on Pacific Islanders Migration: Australia, New Zealand and the USA, Pacific Studies Monograph No. 6 (Sydney: Centre for South Pacific Studies, University of New South Wales).

Rallu, J.L. (1994). "Australia and Pacific islander migration", Asia and Pacific Migration Journal, 3(2-3):431-443.

South Pacific Commission (1993). South Pacific Economies Statistical Summary: SPESS, No. 13, Noumea, New Caledonia.

UNDP (1994). Pacific Human Development Report, Suva, Fiji.

For further information on this material please contact: loftus.unescap@un.org