Volume 78 1969 > Volume 78, No. 2 > Some comments on the 'Report on the Results of the 1966 Census,' Kingdom of Tonga, 1968, by Garth Rogers, p 212 - 222
                                                                                             Previous | Next   

- 212

The 1966 census of Tonga was modelled in planning, method, analysis and presentation on its highly successful predecessor of 1956. The 1966 material is intended to be directly comparable with that presented a decade earlier, with the addition of data on land-holding. The census of 1956 was the first in Tonga to benefit from the methods and principles of scientific demography; the census officer M. U. Tupouniua, a graduate of Auckland University, acknowledged his gratitude to Dr Norma McArthur, experienced demographer of South Pacific societies and at that time Census Commissioner for Fiji. 2 The 1966 Report was prepared by Sione Fiefia, a graduate of Hawai'i University; M.U. Tupouniua and Dr Zwart, Census Commissioner for the Fiji Government, were among the advisers and consultants. Much of the processing of census material was done in Fiji on an IBM contract; the total cost of the census was reduced by judicious accounting from the 1956 figure of sixpence per head of population to fourpence per head. The review which follows presents the statistical facts and a commentary on their social significance. 3

The total enumerated population of the Kingdom on 30th November 1966, was 77,429 persons, 98.31% of whom described themselves as Tongans. This represents an increase of just over 36% (20,591) on the 1956 population (56,838), an average annual intercensal growth of over - 213 3% (possibly 3.46% in 1966). With one half of the population under sixteen years of age and with 33.99% of both sexes in the reproductive age-group (15-44), there is every possibility that present rates of increase will be maintained if not exceeded in the next intercensal period. On this evidence the total population would be well over 100,000 by 1976, and would double between 1966 and 1986.

The overall density of population on inhabited islands in 1956 was almost 277 persons per square mile or 2.31 acres per person. By 1966 the comparable figure was 376.6 persons per square mile or 1.69 acres per person, an increase in overall density per square mile of nearly 100 persons. 4

The largest increases took place on the principal island Tongatapu where the intercensal growth amounted to 54%; the smallest increases in overall density occurred in the Ha'apai group, already the most heavily populated islands in Tonga, where the density rose from 480 persons per square mile in 1956 to 523 in 1966. In 1966 the most densely populated of the Ha'apai islands was Matuku where 138 people in 29 “single family units” in 19 households were living on 82 acres (1,062 per square mile).

The 1966 Report gives warning that overcrowding and shortage of housing are becoming acute throughout Tonga (p. 35) but the figures themselves suggest that this rise is a corollary of population movement to Tongatapu. The figures do suggest that the overall average size of Tongan households has risen from 6.53 persons to 6.7 persons in the intercensal period. 5 But average household size in each of Tonga's three main towns has fallen; Nuku'alofa from 8.04 to 7.96; Pangai, Ha'apai from 8.39 to 7.55; Neiafu, Vava'u from 7.6 to 7.42. On the other hand, in 1956 there were six towns and villages with an average of more than eight, and only one with over nine persons per dwelling whereas in 1966 there were nine villages with eight, and four with an average of over nine persons per dwelling. Most of these were in rural areas of Tongatapu which suggests that rising pressure on housing was more acute in rural Tongatapu than in Tonga overall.

Tongan households are defined in the census first as consumption units, later as sleeping units. Within an api ‘compound’ a sleeping unit and a consumption unit rarely correspond. Family units are defined as spouse(s) and child(ren). Of the 10,942 Tongan households recorded for the entire Kingdom in 1966, 16% or 1,729 contained more than one family unit; 1,454 or 13.2% contained two family units; 244 or 2.2% had three family units; 31 or 0.28% had four or more. Tongatapu Island with 61.8% of the - 214 Kingdom's total population, had 58.9% of the total households with two family units and 63.2% of the households with three or more family units.

The Report does not distinguish whether households with several families have come about as a result of grandparents, son and wife, and grandchildren living in the same household, or as a result of brothers and their wives sharing one household. It is also unclear how the local census officers demarcated dwelling units clustered together and cooperating as the one single household. In many parts of Tonga there is a precedent in custom for three or even four generations of the same kin-group to live within the same 'api, the family unit of the oldest son occupying the main dwelling with other units (parents and unmarried siblings; families of married siblings) having separate dwellings around. Because of the traditional brother-sister tapu there may be an additional hut for unmarried sons. There is less precedent in Tongan custom for two or more family units of the same generation (e.g. siblings, their spouses and children) to live within the one dwelling: this could be a criterion of over-crowding.

It seems from the census figure of 16% (“households” with more than one family unit) that overcrowding is not yet severe in Tonga except in Nuku'alofa. Here families of siblings are sharing dwellings, and huts are being built on swamp land. Elsewhere, the occurrence of households containing more than one family unit may be accounted for as the persistence of the traditional co-resident extended family, comprising several generations where each “extra” family unit may have its own separate dwelling house which may or may not have been recorded in the census schedule as a household. The increase in average household size has been accompanied by a fall in average overall acreage of land per household from 15.08 acres per household in 1956, to 10.98 acres in 1966.

The population densities already discussed have little meaning except in relation to the availability and utilization of natural, agricultural and human resources. The natural resources of Tonga comprise almost exclusively the soils, reefs and surrounding seas with, in late 1968, an elusive possibility of oil; the agricultural resources comprise almost entirely trees—particularly coconut, banana, breadfruit—and ground crops grown mainly without mechanical aids by slash-and-burn methods. Human resources are discussed at the end of this paper.

Dependence on agricultural produce is enormous; in 1961 and 1964 these products amounted to 98.19% and 96.39% respectively of Tonga's total export earnings with copra (the dried kernel of the coconut) forming the basis of the economy and constituting over 80% of the visible exports throughout the 1960s. 6 Agricultural products are the surpluses and small cash-crops of Tonga's 12,000 or so peasant farmers who are scattered throughout the Kingdom and rent their allotments either from titled nobles or directly from the Government.

- 215

Systematic development of agricultural resources was the prime concern of Tonga's first development plan (1965-1970), the main effort and spending being allocated to a comprehensive Coconut Replanting Scheme. Banana exports to New Zealand and Japan have fluctuated although they have gradually increased throughout the 1960s. But other agricultural exports—such as watermelons, fungus, groundnuts, pineapples, vegetables and ginger—comprise only a very small fraction of the total exports. Other crops are either completely absorbed by home consumption or are, as yet, only in an experimental stage.

A fishing industry has been established in Nuku'alofa with the aid of Japanese experts but in 1968 it was barely supplying the home market. Local private fishermen would sell a string of small fresh fish, in all about three to four pounds, for about fifty seniti (NZ50c). Fish was sold in the government market at intervals of about two weeks but the supply rarely lasted more than one day. Tuna and shark caught by the government vessel were refrigerated in the government market and sold for ten and four seniti per pound respectively. Labourers' wages at that time in Nuku'alofa were one pa'anga (100 seniti) per day. By early 1969 small shipments of crayfish tails were being transhipped to the United States via Fiji by an Australian fisherman who owned his own craft but employed local Ha'apai labour.

A very high proportion (75.57% or 57,526) of the total Tongan population was classified in the 1966 census as “economically inactive”, that is, as persons not engaged in any form of “industry” ranging from subsistence agriculture to public service. Included in this category were 16,308 or 81.3% of all adult Tongan women who were engaged in “home duties”, all children under fifteen years of age, and all who were students over fifteen years or retired. It should be understood, however, that women perform some important economic tasks such as assistance with food-gathering and copra-making. Moreover, according to the traditional division of labour in Tonga, women make mats, baskets and bark-cloth as part of their “home duties”. Nevertheless, only 1,013 or 5.0% of adult Tongan women were engaged in activities other than these. 7 No Tongan women were listed as employed in “government and administration” (excluding government clerks); 607 or 3.02% of all adult Tongan women (compared with 1,715 Tongan males) were listed as employed in “services” (religious, educational, medical, personal; not transport); 173 or 0.86% (compared with 180 Tongan males) in “commerce”; and only 18 or 0.089% (compared with 579 Tongan males) in “part-manufacturing and processing” industries. 8 In Tonga, the traditional idea and practice that man is the provider and head of the household still persists.

- 216

Of the 20,766 Tongan males over fifteen, 17,686 or 85.17% were listed as being employed in some “industry”. 13,219 or 74.74% of these were recorded as being self-employed in “agricultural production for household consumption and export”. This represents a rise over ten years of 3,483 persons or 35.77% in this category, but unfortunately the category includes all those without regular or specific occupation, that is, no attempt was made in the 1966 census to differentiate self-employed from unemployed, nor those with title to land from those who have no title but are growing subsistence crops on someone else's land; it is difficult therefore to quantify pressures on land and the 'api system.

The Tonga land tenure system is unique in Polynesia and is the result of laws enacted over a hundred years ago by the founder of modern Tonga, King George Tupou I. Every male Tongan is entitled, on reaching the age of sixteen years, to receive a grant of eight and a quarter acres as an 'api tukuhau ‘tax allotment’, and a grant of up to one rood twenty-four perches in a village or town as an 'api kolo ‘town allotment’. No one may legally hold more than one tax and one village allotment, but if a person forgoes the latter he may obtain under the law, a bush lot of up to twelve and three-eighths acres, that is, one and a half times the usual size of an 'api tukuhau. The annual tax on tax and village allotments together is eighty seniti (NZ80c); tenure is for life and is usually reallocated to the eldest son. 9 These laws have created a nation of small but secure tenant farmers alongside a landed aristocracy. The head of every family or household in Tonga, irrespective of his rank or wealth, has thus until very recently been assured of the means of production both of basic necessities such as food and housing, and of cash crops both for purchasing clothing, tools, household goods, and for education. With shortage of land, and the strain traditional housing puts on coconut trees, housing is passing from subsistence to the monetary economy.

According to the Report of the Lands and Survey Department for 1966 (p. 2), the total number of tax allotment holders in December of that year was 12,517. The 1966 census records that only 8,305 or 41.57% of all Tongan males over sixteen said that they were holding tax allotments, whilst 11,553 or 57.84% reported that they held no land (p. 96). Some 116 men over sixteen did not state their position regarding land but even if they all had held tax allotments, 4,096 of the allotment-holders recorded in the Report of the Lands and Survey Department would remain unaccounted for. However, some of the Lands and Survey allotment-holders would undoubtedly be widows holding their husband's allotments as a life interest, 10 or the sole surviving daughters of the deceased land holder. 11 The number of widows recorded in the 1966 census was 1,552 but it is not stated how many of them were holding the lands of their - 217 deceased husbands. As the law prohibits any person from holding two or more tax allotments, either a very high proportion (32.78%) of tax allotment holders must be women, or the census returns have not revealed the way land is allotted.

Whatever interpretation is cast on allotment holdings and holders the fact remains that at least one half of all Tongan men over sixteen stated that they held no tax allotment. According to Tongan ideology these men are landless, “ta'ema'u 'api tukuhau”, even though they may have access to land for food production. As expected the greatest number of “landless” men fell within the lower age-groups but it is remarkable that so many men in the middle age-groups were without tax allotments. 83.5% of all men between fifteen and twenty-nine years of age and 51.7% of those between thirty and forty-four stated that they held no land. One fifth of these “landless” men were living in Nuku'alofa but the greatest majority were in rural areas of Tongatapu where there is public transport to Nuku'alofa. For example, in nearly all of the rural villages of Tongatapu (44 out of a total of 53), more than 50% of the males eligible for land stated that they had no land. In the Ha'apai islands, on the other hand, in only eight out of twenty-two rural villages were there more than 50% of all adult males without tax allotments.

The Census Officer for 1966 wrote that “. . . even if the larger tax-'apis are further subdivided into four-acre allotments, there would still be not enough land for all male taxpayers. This simply means that the number of landless males will continue to increase yearly”(p. 31). There are, however, some unhabited high islands (see footnote 3) and an (unpublished) amount of land held by Tongan nobles which is either not made available as tax allotments or is farmed by commoners resident on noble estates who have not been granted registered title to the land.

What then are the prospects for these “landless” males? Firstly, according to the custom of the land they should all have access to land belonging either to a kinsman or to a noble. These rights of usufruct are in most cases for the growing of subsistence crops only and not for cash crops, and the traditional right is becoming a privilege which is paid for. Since it is becoming more difficult for even rural dwellers to do without money, as modern manners create new “necessities”, it is likely that “landless” men and their families will not share the same standards of living as those with registered land. It is noteworthy that if the eight-acre allotments were divided in half (as has already happened in parts of the Ha'apai and Tongatapu islands), there would be a corresponding lowering of cash income and consequently of standard of living of the new occupants, as well as the possibility, under current methods of cultivation, of soil exhaustion and erosion.

Secondly, there may be a case in Tonga for the development of cooperative land-holding groups centred on the village, with household units working land within the cooperative. Specialized crop production, mechanization, fertilization, and fencing could all be developed within the cooperative, but each household unit could retain its own 'api lands for subsistence cropping and household requirements. Such a scheme - 218 would be more likely to succeed on government than on noble estates, and only, of course, in areas where land is available for lease. There is evidence in Tongatapu that some banana-producing cooperatives have achieved success by using vacant land and employing tractors.

A third alternative for landless men is migration, and the 1966 census shows in a number of ways that “. . . the Tongan population has become mobile, both internally and externally . . . the drift of young people from the atolls of Ha'apai especially and rural areas of Tongatapu to Nuku'alofa is probably indicative of this trend” (p. 11).

Tonga now has thirteen towns and villages each with a population of more than one thousand people; nine of these are situated on Tongatapu island whereas the Ha'apai and Vava'u groups contain only two each. These thirteen towns and villages showed a combined percentage increase of population of 48.3% during the intercensal period, although the two Ha'apai towns combined actually lost population over this period. Population growth in the nine Tongatapu towns and villages ranged from 34.0% to nearly 357.0% (Haveluloto) for the decade from 1956; Nuku'alofa and suburbs (historically three villages) increased over the same period by 6,483 persons (by 70.0%) to a total of 15,685.

Nearly 32.0% of all persons over sixteen living in Tongatapu on the night of the census were born in other parts of the Kingdom, with 14.4% deriving from Ha'apai and 11.4% from the Vava'u group. Some of these people living away from the place where they were born were undoubtedly students and temporary residents from the northern islands, but differential rates of increase for Tongatapu (53.2% during the intercensal period) and the Ha'apai group (6.78%) suggest that there is a higher rate of internal migration than could be explained simply by students and visitors. Walsh has described the process whereby temporary immigrants become permanent residents. 12

Some imbalance between the sexes is apparent in Tongatapu's excess of males “born elsewhere” over females “born elsewhere”. Supporting the argument that many of Tongatapu's immigrants derive from Ha'apai, is the evidence that there is a shortage of males in that group; there are only 491 males per 1,000 persons in the Ha'apai district—the only deviation from the national pattern of male predominance.

The average age for marriage has increased since 1956 when over 3.0% of all men and over 9.0% of all women were married in the fifteen to nineteen age-group. The 1966 returns indicate that only about 1.0% of the total male population and just over 7.0% of the total female population in this cohort were married. It is possible that some of the reasons for this delay in marriage are the shortage of land in rural areas, the excess of men in urban areas, the increasing cost of traditional weddings, the shortage of and desire for wage employment and money, and the enormous advantage post-primary education gives to obtaining wage employment.

- 219

Further evidence that Tongan attitudes to marriage are changing is suggested by the proportion of husbands to wives. The 1956 census asked people if they were married, never married, widowed, or divorced. De facto spouses living together were treated as married, “the aim of the census being to record things as they are and not as they should be”; 13 the total number of recorded husbands exceeded the number of wives by 317. The same questions concerning marital status put in 1966 returned a remarkable reversal; there was, despite a steady ratio of males to females, a surplus of 307 wives over husbands.

These figures may be explained in several ways. The census figures may be in error because of the question being misunderstood or because of the informant concealing his or her de jure as against de facto marital status from the often strange census enumerator (teachers): this would require some widowed or unmarried mothers to have been included in the 1966 but not in the 1956 returns as “married” persons and/or some widowers or de facto husbands to have been included as “married” in the 1956 but not in the 1966 returns. A second possibility is that the number of married men overseas (in Australia, Hawai'i, United States, New Zealand, New Hebrides, United Kingdom) has greatly increased. Lastly, there is a case here for serial monogamy, that is, contraction of second marriages without legally divorcing the first spouse, especially in a country where legal divorce is uncommon 14 because of its expense 15 and the fact that it is not traditional custom.

As already mentioned it is not possible to quantify unemployment or underemployment in Tonga from the data of the 1966 census; what is apparent, however, is that Tonga is industrialising very slowly—considerably slower than is necessary to absorb projected increases in the size of the work-force. It has been estimated in the Report (p. 30) that over 11,000 young men will pass the age of fourteen during the decade beginning 1966, but only 1,444 men will pass the age of sixty-four over the same period. It is likely, therefore, that apart from deaths and migration, the size of Tonga's male work-force will increase by about 9,700 in ten years. Taking into account the land shortage discussed above, where are these young men going to find employment?

There was no industry of any importance in Tonga before 1940, but the Tonga Government took control of copra production and marketing in 1942 after Lever Brothers made the mistake of selling out on a rock-bottom market. This state enterprise has produced economic independence for Tonga through the policy of gradual industrial development by government agencies without outside financial assistance. 16 Since 1956, there have been state-owned developments in shipping, fishing and manufacturing (desiccated coconut, coir-processing, soap, concrete blocks, - 220 hard-board, cabinet-making, ship's gear and housing), as well as developments in private industries such as cigarette and tobacco processing, handicrafts, tailoring, boat-building, bread-making, and in government service industries, communications, electricity and water supplies and mechanical engineering. Until 1956, Tonga maintained a favourable overseas trade balance without incurring a national debt. Since 1965, moneys have been borrowed from Britain for development plans. But these developments are not sufficient to employ the increasing numbers of young men migrating to Nuku'alofa seeking work. For example, in 1956, 2.41% of the Tongan male population over fifteen years was engaged in “part-manufacturing and processing”; by 1966 this percentage had rised to 2.78% of the adult Tongan males, an increase of some 192 persons, but Tongan females employed in this category decreased over the same period by some 77 persons, that is, in the decade following 1956 the total number of Tongans employed in “part-manufacturing and processing” industries rose by only 115 to 597 persons. The “services” branch of the economy absorbed an increase of 568 Tongans, “transport and communications” 65, whereas Tongans employed in “commerce” dropped from 786 to 353, a loss of 214 workers. The number of non-agricultural labourers rose during this period from 192 to 786—a gain of 594 workers. Comparable figures for non-Tongan residents in Tonga in 1956 and 1966, show decreases in the number of persons employed in “part-manufacturing and processing” industries (a decrease of 83 persons), “services” (down 63 persons), “transport and communications” (down 24 persons) and “commerce” (down 103 persons). These figures conform with the Tonga Government policy of gradually replacing non-Tongan workers with Tongans.


Rapid increases in the Tongan component of the population are occurring: about 38% in ten years; this is producing a very youthful population in which over half of the population is under sixteen years of age. Population expansion is due not to a rising birth rate (which is fairly constant at 3.24%) but to a rapidly decreasing death rate (in 1966 the death rate reached an all time low of 0.29%).

Pressures on land are beginning to be felt but are not yet acute because it is comparatively easy for any Tongan to obtain access to land for subsistence purposes. Pressure on the 'api system as it now stands is very great with possibly well over half of the men eligible for land without land or at least without security of tenure. Possibilities for intensification of land use have scarcely been sought; diversification of crops and scientific methods are, with the exception of the Coconut Replanting Scheme, as yet confined to research stations; large tracts of land belonging to the nobility remain officially unallotted; several high islands are unsettled; larger allotments could be subdivided with the introduction of intensive methods. Pressure on housing is not yet acute except in Nuku'alofa, but the Tongan administration seems to have adopted a European norm of one “family unit” to one household in spite of the traditional custom of - 221 building small dwellings near the main dwelling for siblings of the opposite sex, parents, and married members of the same family.

The need to step up the rate of industrialisation is shown by the beginning of land shortages, the increase of adult “landless” men, the indirect evidence of under-employment throughout the Kingdom, and the movement of increasing numbers of rural dwellers to urban centres, especially to Nuku'alofa.

The most spectacular development in Tonga since 1956 has been the acquisition of a national shipping fleet. Before that date, two motorized sailing ketches were the principal means of inter-island transport; all fuel was imported from Fiji in 44-gallon drums. The progressive, farsighted administration ordered the first vessel, Aoniu (514 tons), in 1956; it was delivered in 1958 and by 1960 it had made a gross earning of £T68,120 with net profits of £T13,970 after providing for depreciation and other running expenses. The Tongan fleet now comprises seven vessels, with the largest, Niuvakai (1,800 tons), plying between Sydney Australia, Japan, Fiji, and Tonga. Although great savings are now made each year on freights to and from the Kingdom, the shipping line has, however, done little to relieve the need for more jobs in Tonga. Raw materials could be transported to the Kingdom by these ships to provide employment in selected (possibly labour intensive) industries. If European and Australasian governments do not take the initiative by offering experts for this purpose it is likely that trade and aid may be forthcoming from Asia, especially from Japan.

Tonga has also shown how a state-owned industry can provide a large degree of independence for a small nation. The Tonga Copra Board and allied government agencies started with improving the quality of export copra, and now, with the aid of British finance, have launched a highly successful Coconut Replanting Scheme throughout the land. Government industries have already been discussed. Philatelic and numismatic sales are still being fully exploited by the Minister of Finance.

Nevertheless, economic growth, although spectacular for a country using mainly its own limited resources, is not fast enough to cater for the demand for wage labour. While the population continues to increase at the present rate, the need for the expansion of wage labour, either on the land or in industries, will persist.

Finally, economic expansion and industrialization is not by itself a panacea for the pressures and problems besetting this tiny island Kingdom; techniques will need to be changed, social privileges equalised, and human resources marshalled and developed, especially in the fields of education and vocational training. Education for white collar jobs carries greater prestige, but efficient production of export crops may be more lucrative for the individual.

At present, compulsory primary schooling ends at standard IV, a class reached by many children at the age of ten. Intermediate and secondary schooling is not compulsory, so if primary school pupils do not gain a place in one of the few secondary schools they repeat the last year in the primary school once or even twice. Because of this an enormous percentage - 222 of Tongan children are leaving school with no prospects of either secondary schooling or vocational training. State schools have no likelihood of introducing vocational training under the present budget allocations and places in secondary schools are few. In 1968, a total of 2,400 pupils sat the Tonga High School (Government) examination with only 55 places available. 17 Even in secondary schools there is little education for specific careers; the 1967 Report of the Department of Education estimated that only 8% of a secondary school pupils were receiving vocational training.

The Annual Report of the Police Department for 1967 shows that the number of reported crimes had increased by 21.0% over the previous year and that 52.0% of those convicted on criminal charges were between the ages of nineteen and twenty-five. The proposed scheme to reinforce the police force and build a new jail and police headquarters will probably do little to alleviate this problem.

The United States Government has allocated an annual quota of one hundred immigrants to Tonga; this scheme is fully booked for the next three years. Small Tongan communities exist in Samoa, New Hebrides, Australia and New Zealand, but to New Zealand at least no regular emigration is permitted.

  • TONGA, 1948. The Law of Tonga. Revised Edition. Dunedin, Coulls Somerville and Wilkie.
  • — 1958. Report on the Results of the 1956 Census by M. U. Tupouniua. Nuku'alofa, Government Printer.
  • — 1960-1967. Statement of Trade and Navigation for the years 1960 to 1967. Nuku'alofa, Government Printer.
  • — 1966. Development Plan 1965-1970. Nuku'alofa, Government Printer.
  • — 1968. Report on the Results of the 1966 Census by S. N. Fiefia. Nuku'alofa, Government Printer.
  • TUPOUNIUA, Mahe 'uli'uli, 1960. “A note on Economic Development in Tonga”. Journal of the Polynesian Society, 69:405-408.
  • WALSH, A. C. 1967. “Urbanization in Nuku'alofa, Tonga”, in Urban Problems in the South Pacific. Technical Paper No. 152. Noumea, South Pacific Commission.
1   Because similar articles have since been published in other Journals, the writer has asked for it to be made clear that this article was accepted for publication in January 1969 (Ed.).
2   TONGA 1958:iv, 2, 5.
3   I am indebted to Antony Hooper, Karl Tu'inukuafe, and Edgar Tu'inukuafe for valuable comments and criticisms on an earlier draft of this paper, and to Peter Chignell for a personal communication.
4   Thirty-three out of c. 150 islands are inhabited; the total habited land area in 1966 was less than 15 miles square. Niuafo'ou (13.5 sq. mls.) and Tofua (18.05 sq. mls.), both partially resettled since 1956, have not been included in the assessment of population density, nor have the unhabited islands of Late, Kao, Hunga Ha'apai, Hunga Tonga, 'Ata and Fonualei, all high and fertile but without harbours or anchorages.
5   Comparable figures of mean household size in other island groups are, Cook Island 1966, 6.5; Western Samoa 1961, 5.7; Fakaofo, Tokelau Islands 1968, 10.4 (Personal Communication, Antony Hooper); Niuafo'ou, northernmost island of Tonga 1967, 6.5.
6   TONGA, 1960-1967 passim. Total exports for 1961 amounted to £T1,374,658; agricultural exports £T1,349,461; non-agricultural exports £T25,197.
7   TONGA, 1968: Table 12b, 72-74.
8   In 1961, 9.5% of all adult Western Samoan women were classified as “economically active” (excluding “village agriculture”); 7.9% of all adult women, mainly the 15-29 age-group, were engaged in “industry” of one sort or another. The 1966 Cook Island census (pp. 5,55) lists 1,620 adult Cook Island women or 35.4% of all adult females as “economically active” of whom 1,032 resided in Rarotonga. Cook Island women engaged in manufacturing numbered 301 (compared with 301 males), in services 539 (compared with 737 males), professional, technical and related workers 321 (compared with 412 males).
9   TONGA, 1948:405-459. “The Land Act”. Annual poll tax, $T3.20, is paid by all adult Tongan males, whether landholders or not.
10   TONGA 1948:427-8.
11   TONGA 1948:428.
12   Walsh 1967:14.
13   TONGA 1958:23.
14   There were only 405 divorces recorded on census day, representing only 1.77% of the 23,317 people who declared themselves married.
15   Legal costs for divorce proceedings may range from twenty to several hundred pa'anga, and the cost of maintenance has also recently risen.
16   Tupouniua 1960:405-508.
17   Mr Peter L. Chignell, Director of Education in Tonga; personal communication.