sign in sign up
<<
Home , Norfolk Island

Geographical Review of Vol. 77, No. 5, 352-367, 2004

International Migration from , South Pacific: A Behavioral Approach

Raelyn Lolohea 'ESAU Graduate Student, Department of Geography, Kyoto University, Kyoto 606-8501, Japan

Abstract: The purpose of this study is to examine socio-cultural aspects of international migration from Tonga and its impact based on the behavioral approach, which has thus far been neglected in existing literature associated with the country's migration. An interview and questionnaire survey of 150 households from the three island groups in Tonga was conducted. As a result, the following findings were obtained. Household size has recently decreased due to transformation from the extended family to the nuclear family and emigration from Tonga. Consequently, the number of migrants per household is larger than before. The individual or nuclear family rather than the extended family plays a greater role in migration decision making now. With respect to the reason for migration, an increasing motivation to migrate for study abroad since the 1990s is remarkable. Reliance on remittance is not significant partially due to the increase in student migration. Furthermore, there is an obvious tendency for migrants to marry persons with Tongan nationality, and, thus, they are quite likely to settle in their host countries and not return to Tonga.

Key words: student migration, decision-making unit, remittance, dependency model, MIRAB model, Kingdom of Tonga

region, coupled with the scattering of small Introduction islands of approximately 551,059km2 across a vast ocean of 30 million km2, clearly makes the International migration in the South Pacific geographic distances between small island region including Tonga began as a mere trickle states within the region generally very great. in the early 1950s and eventually led to the As a result, these islands encounter 'special dis dramatic outflow of people into the advantages associated with limited domestic countries of , and the markets, remoteness, a narrow resource base, in the 1960s and 1970s (Bedford indivisibilities in investment and proneness to and Lloyd 1982; Connell 1983a; Trlin 1987). natural disasters' (Cook and Kirkpatrick 1998: The South Pacific region contains three 845). The smallness of Pacific societies, how indigenous population groups: ,1 ever, does not imply that world development Micronesia2 and Polynesian The Melanesian has been irrelevant to the region (Haberkorn islands have more than 75% of the population 1997: 220). living in rural areas, but rural-urban migration The historical commencement of interna is now conspicuous (Bedford 2002; Connell tional migration from the Pacific Islands is 1990). is an exception due to the political greatly embedded in the colonial connections to upheavals ensuing from the 1987 and 2000 the metropolitan countries (Appleyard and coup d etats,4 resulting in the dramatic emigra Stahl 1995; Brown 1998a; Connell 1990). This tion of the Indo-Fijian population (Bedford explains the diversification of emigration 1989; Chetty and Prasad 1993; Mohanty 2001). trends from various Pacific Island countries. In contrast, the international migration of Mi For example, the , and Toke cronesian and particularly Polynesian island lau became 'Island Territories' of New Zealand countries is having a great impact on the home in the late 1940s (Hooper 1993), and it has given country population. them the privilege of New Zealand citizenship The remarkable diversity of the South Pacific by birthright (Connell 1990; Hooper 1993). The

352 International Migration from Tonga 131

Western are granted citizenship for an Table 1. New Zealand census counts by Island annual quota of 1,100 by the New Zealand gov population estimate ernment (New Zealand Immigration 2002a). Tonga is a unique case. Since the country was a British protectorate until 1970 (Campbell 2001) and lacked any political dependence with any Pacific Rim countries, it has often con fronted various obstacles in obtaining visas to enter destination countries and international migration has became significant only in the 1970s (Hayes 1992; Connell 1990). Exceptional options associated with migration were Estimate at mid-2001 prepared by the South Pacific through educational institutions and Christian commission Demographic/Population program. churches (Bedford and Gibson 1986; Connell Source: http://www.stats.govt.nz. 1990; Hayes 1992). In addition, a temporary labor scheme to New Zealand was one way of Organization of this paper is as follows. After facilitating the out-migration of unskilled a literature review on international migration workers from , Tonga and Fiji in the late focused on the South Pacific region, the aim of 1970s (Bedford and Lloyd 1982; Bedford and this article is presented. Next, the study area, Gibson 1986; Connell 1983b; Levick and Bed Tonga, is briefly explained and the interview ford 1988). This scheme was terminated in the and questionnaire surveys are outlined. Then, 1980s due to labour market restructuring, the survey results are discussed by pointing to which caused job losses in the manufacturing the significance of the obtained findings. Final sector of New Zealand in the 1990s (Fletcher ly, concluding remarks are presented. 1995: Statistics New Zealand 2002). It is noteworthy that there are more Tongans Literature Review living abroad than the home population. To date, the only available sources have been vari Previous studies on international migration ous population censuses of host countries be of the South Pacific region produced two well cause of the lack of a recording system in known frameworks: the dependency model and Tonga to monitor emigration, implying that the Migration, Remittance, Aid and Bureaucracy specific numbers are undocumented and mainly (MIRAB) model (Hayes 1992). While the de estimates (Connell 1983a). New Zealand is by pendency model is derived from various coun far the most significant destination (Bedford try reports5 by Connell (1983a, b, 1987), the MI and Gibson 1986; Trimn 1987; Brown 2002) for RAB model is from the joint study of Bertram Tongan migrants. As of 1996, there are 40,716 and Watters (1985, 1986). Both models have in New Zealand (Statistics New Zealand 2001), the two constituents of the island (or internal) 36,840 in the United States (United States Cen system and the international system. The for sus Bureau 2001) and 8,430 in Australia (Fisi' mer system consists of the three sub-systems: iahi 2001). The New Zealand government rec socio-cultural, demographic and economic. The ognizes the contributions made by the Pacific latter system represents the connection be migrants to the country's economy and society tween a particular country of the South Pacific (Brown 2002) and started Pacific Access Cate and the destination countries, and has elements gory (PAC) grants to 250 persons from Tonga, such as culture/education, aid, labor markets, 75 persons from and 50 persons from trade and remittances. in 2002 (New Zealand Immigration The dependency model demonstrates the 2002b). Table 1 illustrates the comparison of penetration of the global economy into the is populations between the country of origin and land economies and the transformation of the New Zealand (Secretariat of the Pacific Commu traditional value orientation and the aspiration nity 2001). of the islanders under consideration. According

353 132 R. L. 'ESAU to the model, migration is encouraged by in 1986, 1993). With these contrasting points in creasing aspirations for a satisfactory standard mind, Hayes (1992) advocated an integrated of living, desirable occupation and sufficient model, which incorporates both the 'kinship accessibility to various services including edu system in conditioning migration decisions and cation and health (Connell 1983a,1990). Conse the growing autonomy of individual family quently, there occurs a reduction both in the members from kin authority' (Hayes 1993a: quality and quantity of the labour force in the 304). Much recently the MIRAB model has island country through selective emigration of been subject to various reconsiderations (Ber the educated or skilled and an influx of remit tram 1999; Poirine 1998). tances from host countries. Noteworthy is that I would like to devote attention to existing the dependency model lacks any equilibrating researches that have applied the two models to process, implying that the island population migration in the South Pacific region. continues to decline (Hayes 1992). With respect to the dependency model, Va'a In the meantime, the MIRAB model was made (1992) criticises the model's conclusion that mi from the studies of the five small South Pacific gration should be discouraged and rural devel countries of the Cook Islands, Niue, , opment should be encouraged (Connell 1983b). Tuvalu and Kiribati by using long-run statisti In addition, empirical studies show that severe cal material. According to this model, a major adverse weather in Niue (Barker 2000) and a motivation to migrate is the need to maximise lack of human resources especially in the rural income and to reduce labour surplus within the areas in Samoa (Muliana 2001) are chief push extended family unit (Bertram and Watters factors that encourage emigration. There have 1985, 1986). Ties and networks between the been fewer studies using or referring to the migrant and the remaining kin are strength dependency model than the MIRAB counter ened by remittances and reciprocal visits, part so far. which further contribute to the making of a The MIRAB model is commonly mentioned in new trans-national corporation of kin.6 The many studies that emphasize the economic con MIRAB model assumes equilibrium in popula tribution of migrant remittances. Previous tion change due to the constant exile of surplus studies that examined the applicability of the labour, which leads to the constant flow of re MIRAB model to the Pacific islands include, for mittances into the islands. Population decline example, Fensterseif er (1993) on Western Sa does not occur because out-migration is com moa, Hayes (1993b) on Ware Island in pensated for by high fertility. , Hooper (1993) on Tokelau, Poirine According to an interesting and suggestive (1994) on French and comparative comparison by Hayes (1992), while the MIRAB studies of various countries (Brown 1998a, b; model expects that migration is motivated by Brown and Conne11 2004; Connell and Conway the kinship system, the dependency model 2000). However, McCall (1996, 1997), Hau'ofa lacks such an idea and attaches importance to (1993) and Laplagne (1997) criticize the conti the role of individualism derived from Western nentalists' views of migration focused on the capitalism. A major contrasting difference be South Pacific islands as misinterpretation. tween the two lies in view of the migration Other studies include Cook and Kirkpatrick impact. The dependency model claims that (1998) on , Treadgold (1999) on Nor migration fails to promote development in the folk Island exemplifying the possibility of get islands and remittances foment economic ting out of the MIRAB situation and the hopes inequality and dependence (Connell 1983a, of boosting agricultural yields and fisheries pro 1983b, 1987, 1990). In contrast, the MIRAB duction to attain greater economic autonomy model argues that island economies manage to for Kiribati (Thomas 2002). Unfortunately, be financially sustainable by the continuous with respect to the integrated model developed emigration of surplus labour that could in turn by Hayes (1992), there has thus far been no guarantee the flow of remittance to the islands detailed empirical investigation. (Bertram and Watters 1985,1986; Bertram Next, let me devote attention to previous re-

354 International Migration from Tonga 133

searches of migration associated with Tonga, proach, it is difficult for them to carefully exam the study country of this paper. Although ine the much smaller micro units, such as Tonga was not included in the studies that household and individual, which are actually resulted in the creation of the MIRAB model, involved in migration. the features of the model have affected Tonga Hence, the purpose of this study is to examine in many ways. Especially notable is the in socio-cultural aspects of international migra creasing number of Tongans abroad (Bertram tion from Tonga and its impact based on the and Watters 1986) and the benefits of remit behavioral approach, which has thus far been tances and improvement of their social neglected in existing literature associated with statuses7 by an emerging middle class of well the country's migration. Specifically, question educated commoners (Marcus 1974, 1981, naire and interview survey on an individual/ 1993). Previous studies on Tonga (for example, household basis was utilized. As is well known, Fuka 1984; Tongamoa 1987; Faeamani 1993, the behavioral approach in migration study is 1995) have emphasized the contribution of re not new, having started in the 1960s (Boyle et mittances to the household economy and to the al. 1998). Nevertheless, such an approach seems development of Tonga. These studies con very useful and promising within the context of cluded that the aspects of the MIRAB economy migration investigation for contemporary are inevitable, since there is high dependence Tonga. on remittances. They argued that migration should be encouraged because it continues to The Study Area be a success with the strong support of the kinship system. Outstanding remittances sent The study area is the Kingdom of Tonga, by migrants and the dominance of the govern located in the South West Pacific within 15 and ment over aid (James 2002) enable us to classify 23 degrees latitude South and 173 and 177 Tonga as a typical MIRAB economy. longitude West. Its population of 101,000 Hence, it can generally be mentioned that, as (Tonga Statistics Department 2001) inhabits far as the migration of Tonga is concerned, the about 36 of the 170 islands with a total land MIRAB model has attracted more attention and area of 744.34km2. These islands are scattered thus has tended to demonstrate higher explana along a sea area of 700,000km2 and clustered tory power than the dependency model. Addi into three main groups (Figure 1). The popula tionally, migration studies by anthropologists tion density is 156 persons per km2, and the were ethnographies (James 1991; Small 1997; main island of Tongatapu houses 69% of the Sudo 1997; Morton 1998; Kishida 2000; Evans entire Tongan population. The country has a 2001). relatively large young population under the age In terms of the structure of the dependency of 15 (39%). The census of Tonga shows a model and the MIRAB model, mentioned ear declining annual growth rate from 0.5% (1976 lier, existing studies of the migration of the 1986) to 0.3% (1986-1996). A most notable South Pacific region including Tonga have change is the drastic decrease of population given much attention to 'remittance' and 'aid' aged 15-34 in 1986-1996, mainly due to inter within the international system, 'migration' in national migration (Tonga Statistics Depart the demographic sub-system and 'income' in ment 1986, 1996). the economic sub-system within the island sys Located just below the Tropic of Capricorn tem. On the other hand, however, the socio and west of the International Dateline, Tonga is cultural sub-system has generally been over 13 hours ahead of Greenwich Mean Time. Its looked in the existing literature. Particularly, west and north neighboring countries are Fiji although the 'motivation to migrate' forms an and Western Samoa. The total land area is important element in the socio-cultural sub slightly larger than Amamiohshima Island (712 system, it has been elucidated only partially. km2) of Kagoshima Prefecture, and the sea area Furthermore, since both the dependency and is three times as wide as the area of Honshu the MIRAB models take a macro-level ap Island. The smallness of Tonga can be under

355 134 R. L. 'ESAU

ment of Tonga Official Website 2002). The general household in Tonga has been said to be the extended family, which operates as a basic unit to fulfill socio-economic obliga tions and therefore plays an important role as the backbone of Tongan society (Vaden 1998). Nowadays, however, Tonga is undergoing fam ily transformation from extended family to nu clear family and it is described in a detailed way in James (1991, 1998), Marcus (1993) and Mor ton (1996). A major threat is the natural disaster of tropi cal cyclones, which hit Tonga annually, in com bination with the country's relative smallness and isolation. Cyclones disadvantage the coun try in competing on an international level for promising trade markets. Hence, international migration is regarded as a safety valve (Connell 1983a,1990; Appleyard and Stahl 1995), which implies safety measures for economies prone to instability and trade deficits. Tonga's economy depends largely on agricul ture products, forestry and fishery. These sec tors constitute 28% of the real Gross Domestic Product (GDP) in 2001/02. Other major eco nomic sectors are government services (18.5%), finance, real estate and services (15.1%), com merce, restaurant and hotel (14.2%) and the official estimates for Tonga's real GDP growth rate is 0.5% in this fiscal year. Tonga suffers a very high inflation of 10.6%. This inflation is Figure 1. The map of the Kingdom of Tonga. due, first, to an increase in shipping freight costs, second, the weakening of the Tongan stood from these facts, currency (pa'anga) against the United States The three main island groups are Tongatapu, dollar, Japanese yen and other major curren Ha'apai and Vava'u (see Figure 1). Tongatapu cies, and, finally, the increase of imported goods (259km2) is the largest island, whose area is prices twice or thrice in a year by overseas half of Awajishima Island, at 40 kilometers long suppliers. and 20 kilometers wide, with its highest point Tonga has a trade deficit. Large import items at 82 meters. The Ha'apai group is comprised of are consumer goods such as foodstuff and vehi small raised coral islands and atolls. Inclusive cles. In consideration of the impressive impor in this group is the volcanic island of Kao, tation of vehicles to Tonga, a continual increase which has the highest point (1,033 meters) in in fuel price is also a chief contributor to the Tonga. The Vava'u group consists of one large deficit. The country fails to improve the trade raised coral island, which is higher on the balance since the level of exports cannot keep southwestern side and rises to a height of 213 up with that of imports. Ironically, however, meters. The major natural resources in Tonga there is little effort made to encourage local are the land, sea and its people; therefore, the production. New Zealand, Australia, the United yields from the land and the sea are the major States and Fiji are Tonga's major import part contributors to the Tongan economy (Govern ners (National Reserve Bank of Tonga 1999).

356 International Migration from Tonga 135

Japan, which occupies about half of the total to the "household" definition in Tongan Census, exports, is the largest export partner. The namely, a group of persons living in the same squash export to the Japanese market is a ma village allotment (ideally 0.4 acres) regardless jor export item, which nets around 4 to 13 of the number of generations therein. Such million pa'anga per year (National Reserve group can constitute not only nuclear family Bank of Tonga 2000). The net trade balance for but also extended family; actually, about two 2000/01 was -96.7 million pa'anga, which rep thirds of the sampled households in my survey resents a deterioration of 38.1% from -70.0 are nuclear families and the rest are extended million pa'anga in 1993/94. families. Based on such definition of "house Tonga receives support in aid from the Euro hold," transformation from extended family to pean Union, New Zealand, Australia, , nuclear family in the country has been de Japan and other countries. The total grant scribed in previous studies (James 1991, 1998; from abroad, however, accounts for only 3.02% Marcus 1993; Morton 1996). of the Tonga government total revenue for A certain part of the sample households in 2000/01 (Government of Tonga Budget State this survey has plural migrants per household. ment 2002). However, remittances constituted This is because there is an increasing tendency about 50% of the GDP in 2002 (Asian Develop of emigration from Tonga in recent years and it ment Bank 2003) and increased by 128.1% from is not unusual for a particular household to 1994 to 2001. They are the major contributors have two or more emigrants, although each to Tonga's balance of payments (National Re household had only one emigrant previously. serve Bank of Tonga 2000). Moreover, it is well Regarding the household which has two or known that these official figures of the transfer more emigrants, the first person who emigrated have underestimated the real amount of remit from it was selected. In the data analyses in tances that enter Tonga, because a large cluding figure and tables, 150 households were amount is considered to be through various used for investigation associated with house unofficial channels8 (Brown 1994, 1995, 1997; hold, and 200 migrants were used for investiga Brown and Foster 1995). tion associated with migrants themselves. The samples taken from Tongatapu were se Interview and Questionnaire Survey lected from the four main districts (Greater Nu ku'alofa, Central, Eastern and Western District). The interview and questionnaire survey was Each district has about 15 to 20 villages. Fortu conducted in Tonga during July and August nately, the district officers assisted me in identi 2002 for 150 households of the three groups of fying households with migrants abroad.9 Two islands of Tongatapu (80 households), Ha'apai or three households were selected from each (20 households) and Vava'u (50 households). village on average. The two outer islands of Sample households with migrants abroad were Ha'apai and Vava'u were also included in the selected and information of 200 migrants was survey to examine differences among the major obtained from these households. Please note islands. In these two islands, the survey was that information about emigrants from Tonga attempted only in the principal districts, and was collected not from the emigrants them five households per village on average were selves, but through their family members stay selected from an identified list of families with ing in Tonga. It is not the best survey method; migrants abroad, based on the author's prelimi however, household members who were inter nary investigation. viewed knew the emigrants very well. This I gave informal interviews using a prepared method was inevitable in limited survey time. questionnaire sheet to obtain information Worthy-of-note here is that a similar method about the migrants' decision-making and mi was also used in James (1991), Faeamani (1993) gration impact. With reference to the house and Evans (2001). holds recorded in the 1996 population census of The "household" in this survey is defined here Tonga, the sample households covered 0.74%, as people living in the same house. It is similar 1.36% and 1.83% of Tongatapu, Ha'apai and

357 136 R. L. 'ESAU

Vava'u, respectively (Tonga Statistics Depart are predominantly smaller in size, '4-6 persons' ment 1996). About 25% of the migrants in this or '1-3 person' in terms of "household size," '1 survey visited Tonga during the period of my person' and '2 persons' in terms of "household survey, due to various church conferences and members currently abroad," and '1 migrant family events, which enabled the author to in only abroad' and '2 or more migrants in the terview them personally. Migrants with email same destination' in terms of "pattern of per addresses were later contacted for confirmation household migrants abroad." Table 2 indicates of information obtained from their families in a smaller household size in comparison with six Tonga. The data obtained from this survey persons per household on average according to were analyzed using SPSS (version 10.0). the 1996 population census (Tonga Statistics Department 1999) and suggests an increasing Results and Discussions tendency for households to have two or more migrants currently abroad. These findings Major findings obtained from the interview seem to imply an intensification of emigration and questionnaire survey are explained below from Tonga under way. based on figures and tables. Then, implications Figure 2 illustrates the trend of emigration for existing literature are discussed. from Tonga during the past four decades based Table 2 is a summary of selected characteris on the questionnaire survey. The return mi tics of the households and the migrants under grants from overseas are not included and the analyses. While the items of "gender of mi comprehensive data for emigration from Tonga grants abroad" and "kindred of household mem is not available due to the absence of a depar bers currently abroad" concern the migrants, ture card recording system by Immigration De those of "household size," "household members partment of Tonga. It shows a gradual increase currently abroad" and "pattern of per house of departures since the 1970s. In spite of annual hold migrants abroad" concern the households. fluctuation, more than ten emigrations are ob The migrants are predominantly 'male' and served annually in recent years. Whereas the ' son' or 'daughter.' Meanwhile, the households emigration in the 1980s generally arose from

Table 2. Summary of respondents and the migrants abroad

Note: The figures in parentheses imply percentage. Source: The author's questionnaire survey.

358 International Migration from Tonga 137

Figure 3. Initial motivation to migrate. Figure 2. Emigration from Tonga based on the Source: The author's questionnaire survey. questionnaire survey. Note: The return migrants from overseas are not in "means of financing cluded and the data for general emigration from ," the family's role, shown in Tonga is not available due to the absence of a de ' family overseas' (32%) and 'family in Tonga' parture card recording system by the Immigra (25%), and the economic independence of mi tion department of Tonga. Source: The author's questionnaire survey. grants shown in 'own money' (18%) are con spicuous. With respect to the decision-making the flexible immigration regulations of destina unit associated with emigration, 'nuclear fam tion countries, emigration in the 1990s was ily' of emigrants forms the largest share (65%), largely characterized by student migration and followed by 'individual' (31%); meanwhile, the family reunions. share of 'extended family' (4%) is negligible. It Figure 3 illustrates the initial motivation to could be implied that there is a declining migrate obtained from the data the author col authority of the 'extended family' in the lected. 'Study abroad' and 'job' constitute the migration decision making process. It has also two largest categories, which together form ap been mentioned elsewhere (James 1991; Morton proximately two-thirds of all migrants. The 1996) that the decision making of an extended highest rate of 'study abroad' is mainly due to family tends to be limited to ceremonial occa the shortage of post-secondary education insti sions, while the decision to migrate is becoming tutions within Tonga. atomized. Table 3 reveals the relationship between Figure 4 denotes the relationship between "means of financing" for migration and the "mi "household income" and "return visit fre gration decision-making unit." Regarding quency." Generally, the frequency depends

Table 3. Means of financing and migration decision-making unit

Note: The figures in parentheses imply percentage. Source: The author's questionnaire survey.

359 138 R. L. 'ESAU

the motivation to migrate. A great majority of Case A consists of migra tion that aimed at a 'job' overseas and occurred in the recent period, since the 1990s. The domi nance of 'annually' and 'frequently' is related to the attendance of church activities and family obligations in Tonga. With regard to other notable migrants contained in this case, the difficulty of obtaining a permanent visa, espe cially for those in the United States, needs to be mentioned. These emigrants have to continu ally renew their short-term visas in order to

Figure 4. Relationship between household in maintain their jobs and continue remitting to come and the return visit frequency. their families.10 Note: 1) Case A: low income and high remittance, Case B: Typical migrants of Case B are those who left high income and low remittance, Case C: high Tonga to 'study abroad' in the 1990s and find a income and no remittance. There are no cases of the three categories 'low income and no 'job' in the decades of the 1970s and 1980s. remittance,' 'low income and low remittance' and Although this case forms the largest share of all ' high income and high remittance.' The terms households, there is no significant difference in ' high' and 'low' here denote annual income of shares of 'annually,' 'frequently' and 'hardly.' more than, and less than 10,000 pa'anga, Thus, a convincing explanation for the relation respectively. The total income was only cash income quoted by the respondents. ship between household income and return 2) 'Frequently' denotes return once in every 2-5 visit frequency is difficult. years and 'hardly' denotes return once every 6 Regarding Case C with no remittance, there is -10 years. Source: The author's questionnaire survey. a tendency for return visits to be 'never' or ' hardly,' although the number of these cases is upon factors such as the visa type, economic relatively small. Case C is mostly formed by the affordability and social status of migrants motivations 'study abroad' in the past 5 years abroad. Please note that the term 'household and 'job' in the 1970s. The low frequency of the income' does not contain remittances. Al latter migrants seems to arise from a develop though I was able to obtain income information ing social network and an increasing will to for each of the households under consideration settle for a longer duration in the host coun from the interview, the exact amount of the tries. remittances for each of them was unobtainable. Table 4 explores remittance uses. To com Thus, to anticipate a relative weight of remit pare the findings here with those in previous tance and to discuss its relation with the return studies, investigations by Tongamoa (1987) and to-visit frequency, I have to use the criterion, Faeamani (1993) are referred to as well. Ac based on the answer in the survey, as to cording to my survey, 'daily consumption' whether the remittance is larger than the in (35%) and 'household bills' (21 %) constitute ma come, and, finally, classify all households into jor uses, and 'church' and 'education fees' fol three groups (Cases A, B and C). The weight of low. Meanwhile, Tongamoa (1987) indicates remittance in the household economy decreases that 'daily consumption,' 'church' and 'educa from Case A, to Case B to Case C in this order. tional fees' are the three most important uses on Figure 4 suggests that the contribution of a multi-answer basis; Faeamani (1993) clarifies remittance to the household economy is not that the remittance uses of 'church,' 'daily con significant, because Case B and Case C consti sumption' and 'household bills' are important. tute approximately two-thirds of all house The results of the three studies including my holds. Moreover, a great difference among survey, shown in Table 4, are greatly different these three cases can, to a certain extent, be from each other. Since such difference may explained in terms of the migration time and arise from differences of year, place and respon

360 International Migration from Tonga 139

Table 4. Remittance uses in three migration here. Regarding "destination country," New surveys on Tonga Zealand, the United States and Australia are popular, in this order. In Ha'apai, however, emigration to New Zealand is exceptionally rare (only one person), although the reason is unclear. Rather traditional methods such as ' telephone only' or 'telephone and postal mail' are the dominant means of communication be tween the migrants and their family members in Tonga. With respect to "nationality of mi grant's spouse," 'Tongan' is quite large (69%) and there is an obvious tendency for Tongan Source: Tongamoa 1987: 106; Faeamani 1993: 112; and migrants abroad to marry within the same na * present study . Note: 1) Information only for main uses of remittances is tionality. The nationalities 'American,' 'Austra presented. lian' and 'New Zealander' have only 21%, but 2) 'Household bills' refers to the water, electricity the most popular nationality among these three and gas payments. countries is 'American,' implying a slight diffe 3) Tongamoa's survey permits multiple-answer way, implying 'total' exceeds 100%. rence of order in terms of "destination country" and "nationality of migrant's spouse." dents of the survey, it is difficult to extract a Table 6 illustrates the relationship between general tendency from them. the reason for and duration of the return visit Table 5 shows the results of "destination to Tonga. Significant reasons for migrants to country," "communication method" and "na return are 'family matters' and 'church activi tionality of migrant's spouse." The total num ties,' and the most remarkable duration is '1 ber (109 persons) of "nationality of migrant's month.' Some explanation is needed about 'an spouse" means that the other migrants are sin nual conference' in 'church activities.' It lasts gle. Since the strength of the social network for about a week and the main agenda for dis between particular origin-destination pairs cussion cover the maintenance of church infra through chain migration is well known, the structure, clergymen reshuffling, progress of data for the major islands are also presented church-owned schools, decision of the next con

Table 5. Destination country, communication method and nationality of migrant's spouse

Note: The figures in parentheses imply percentage. Source: The author's questionnaire survey.

361 140 R. L. 'ESAU

Table 6. Reason for return trip and its duration

Note: 1) 'Family matters' refer to birth, weddings, funerals, and reunions. 'Church activities' refer to annual conferences and inauguration of church property. 2) Figures in parentheses imply percentage. Source: The author's questionnaire survey.

Table 7. Transition of marital status of migrants

Note: The figures in parentheses imply percentage. Source: The author's questionnaire survey. ference venue and so on. Table 6 signifies that ' temporary visa' are likely to return to Tonga church-related activities form a very important once their visas expire. However, half of these element in Tonga's contemporary socio migrants are positive as to extending their stay cultural life. 'House building' is also significant abroad, and, ultimately, to continue living there due to the damages of the tropical cyclone permanently, if possible. In addition, there is . This factor is especially considerable for hardly any indication that migrants in the host migrants from Vava'u Island, because cyclones countries will return even when they retire raged in the northern islands of Tonga during the first half of 2002. Table 7 shows the transition of the migrants' marital status before and after migration from Tonga to foreign countries. Two-thirds of the migrants (133 persons), half of whom are formed by students, left Tonga as 'single.' More than half of them (73 persons) married after migration, although the status of the great ma jority (118 migrants) remained unchanged. Worthy-of-note is the change of marital status from 'single' to 'married,' implying the making of new ties in the host countries. Figure 5 depicts the relationship between the Figure 5. Current visa statuses of migrants and the likelihood of return to current visa status of migrants and their likeli Tonga. hood of return to Tonga. It is very obvious that Note: 'Permanent residency' and 'citizenship' imply the the migrants with rights of 'permanent resi visa statuses acquired from the destination dency' or 'citizenship' have a definite will to countries after a period of residency. Source: The author's questionnaire survey. settle in the host country. The migrants with a

362 International Migration from Tonga 141

from their present occupations. Fourth, according to my survey, reliance on Next, keeping the existing studies in mind, I remittance is not significant partially due to an would like to refer to the significance of the increasing share of student migration. This findings obtained here and their implications. finding contrasts with previous studies, which First, household size has been reduced re demonstrate a strong dependence on remit cently for two reasons: transformation from ex tance (e.g., Fuka 1984; Tongamoa 1987; Faea tended family to nuclear family and emigration mani 1993). There has been an interesting de from Tonga. As a consequence, the number of bate as to whether remittances were used either migrants per household is now larger than the negatively (for example, daily life consumption ones previously. This can be confirmed from and house building/repair) or positively for fu Tonga's population censuses and similar find ture development (for example, investment and ings are described in Tongamoa (1987), James savings), as shown in Fuka (1984), Tongamoa (1991), Faeamani (1993) and Morton (1996). (1987) and Faeamani (1993, 1995). Although I Moreover, according to my survey, most house cannot answer this question convincingly holds can afford the cost of migration to travel based on my survey, a certain use of remittance abroad. This finding contrasts with Connell for education fees, which I have elucidated here, (1990), who indicated that migration is selective seems to serve as evidence that the remittances and limited to the highly educated and skilled. have partially been directed to positive pur Second, my finding that the individual or poses for future development. An increasing nuclear family rather than the extended family motivation to emigrate for the purpose of study plays a greater role in migration decision abroad is noticeable, but it does not mean that making is notable. A prominent role taken by remittance has been losing its role. However, it the extended family is a major characteristic of is still important particularly for households, the MIRAB model. Given the traditional Ton which have to depend upon it. gan way of thinking associated with migration, Fifth, although a change in the migrants' such a trend is new. It has also been reported marital status in host countries and its implica elsewhere (James 1991; Morton 1996) that the tions have hardly been discussed in existing decision making of an extended family tends to researches associated with migration from be limited to ceremonial occasions, while the Tonga, I have presented the finding that there decision to migrate is becoming atomized. This is an obvious tendency for migrants to marry finding including mine is contrary to the MI persons with Tongan nationality. While this RAB model (Bertram and Watters 1985) and may contribute to the persistence of migrant seems to support the dependency model (Con ties with Tonga, it can lead them to develop a nell 1990), which demonstrates an important new social network within the host countries. role of individualism. Furthermore, based on The latter view seems to explain other findings such a perspective, strong communication net in this paper, or a less remarkable role of remit works currently observed may not necessarily tance and a marked likelihood not to return to lead to longevity of reciprocity between Tonga Tonga among the migrants. It remains un and the migrants in the host countries. known, however, whether the fact that emi Third, an increasing motivation to migrate grants from Tonga get married with persons of for study abroad during the past decade is clari Tongan nationality and develop a new social fied, but it has seldom been mentioned in previ network in the host countries will lead to a ous studies of migration focused on Tonga. possibility that the ties with Tonga and their This is related to a high rate (73 migrants, or families decline or to another possibility that it more than one-third of all migrants) of marital will help to construct a new transnational net status change from 'single' to 'married' after work including Tonga. This topic needs to be migration. Quite noteworthy is that the tradi examined carefully on a long-term basis hereaf tional motivation of 'job,' which has attracted ter. great attention in existing literature on interna tional migration, ranks second.

363 142 R. L. 'ESAU

view or questionnaire of the emigrants them Conclusions selves was not conducted here. As a result, important in-depth information including the The purpose of this study was to examine migrants' actual legal status, their social net socio-cultural aspects of international migra work and their will to return/settle could not tion from Tonga and its impact based on the be collected. In addition, since this article behavioral approach, which has thus far been adopts the behavioral approach mainly on a neglected in existing literature associated with migrant/household basis, macro-level factors, the country's migration. An interview and such as political and economic relationships be questionnaire survey of 150 households from tween Tonga and host countries and the gov the three island groups in Tonga was con ernment immigration policies of the host coun ducted. The major findings obtained are as tries, are mentioned only partially. How the follows. recent changes in the socio-cultural system, The number of migrants per household is clarified in this study, have affected the demo larger than before partially due to emigration graphic and economic sub-systems of the island from Tonga. The individual or nuclear family system is another important topic. However, rather than the extended family plays a greater these issues belong to future research agenda. role in migration decision-making now. With respect to the reason for migration, an increas Acknowledgments ing motivation to migrate for study abroad since the 1990s is remarkable. Reliance on This paper is based on a master's thesis submitted remittance is not significant partially due to the to the Graduate School of Human and Natural Envi ronment Sciences, the University of Tokushima, in increase of student migration. Furthermore, February 2003. I acknowledge the scholarship from there is an obvious tendency for migrants to the Government of Japan, which enabled this study. marry persons with Tongan nationality, and, The author is immensely indebted to Professor Hirai thus, they are quite likely to settle in their host Shogo of the University of Tokushima for his invalu countries and not return to Tonga. able supervision and support during the progress of this study. In addition, I extend my sincere thanks to The specific implications of the findings ob Associate Professor Toyoda Tetsuya, Associate Pro tained here for existing studies are already dis fessor Takahashi Shin'ichi and Lecturer Higuchi cussed carefully in the preceding section. Thus, Naoto of the University of Tokushima. Moreover, I I would like to mention a general comment on would like to show my appreciation for the kindness the two well-known models, the dependency shown to me by all of the 150 households, related organizations, individuals and the family of the model and the MIRAB model. They were born author in Tonga. Finally, the author is indebted to within the context of survey reports for several Professor Ishikawa Yoshitaka of Kyoto University countries in the South Pacific region. Although for his professional advice and comments on the ear the two models have exerted great influence on lier draft of the article. migration studies in this region and I have (Received 15 July 2003) learned a lot from them, the results of this paper (Accepted 20 February 2004) do not support one of the two models exclu Notes sively. Because the models were made by incor porating circumstances and conditions before 1. , , , the 1980s, we should now attach importance to Fiji and . more recent factors such as an increasing 2. Kiribati, Federate States of Micronesia, , amount of student migration. Mariana, , and . Finally, a few problems of this paper need to 3. Cook Islands, Niue, , Western be mentioned. Because a survey of emigrants Samoa, , Tokelau, Tuvalu, , and Tonga. dispersed in several host countries was quite 4. Fiji experienced a coup d'etat twice due to po difficult, a survey of households that have emi litical and constitutional crisis resulting in the grants from Tonga and live in the country was overthrow of a democratically elected govern inevitably attempted. In other words, an inter ment by military force in 1987 and by civilian

364 International Migration from Tonga 143

minority in 2000. ment Bank. 5. A major research project undertaken by John Barker, J. C. 2000. Hurricanes and socio-economic Connell producing country reports on migra development on Niue island. Asia Pacific View tion, employment and development in the point 41(2): 191-205. South Pacific on behalf of the South Pacific Bedford, R. 2002. Perspectives on population growth, Commission (SPC) Research, which covered 24 migration and sustainable development in the countries. The project was sponsored by the Pacific region. In 5th International APMRN Con United National Fund for Population Activities ference, Fiji 2002: Selected Papers, ed. K. Lyon and (UNFPA), published by the SPC Noumea, New C. Voigt-Graft, 7-20. Working paper no. 12. Asia Caledonia. Pacific Migration Research Network, CAP 6. It allows the kin members to take advantage of STRANS, University of . the economic opportunities available from the Bedford, R. 1989. Out of Fiji: A perspective on immi international migration of members. gration after the coups. Pacific Viewpoint 30(2): 7. The social stratification in Tonga is based on 142-153. three main groups, with the Royal family at the Bedford, R. D., and Gibson, K.1986. Migration, employ summit, the nobility and the commoners, ment and development in the south Pacific, Country which contains about 98% of the population. report no. 23B New Zealand. Noumea, New Cale The commoners are restricted to this social donia: South Pacific Commission and Interna status and there is no upward mobility since tional Labour Organisation. the upper classes are hereditary. Bedford, R. D., and Lloyd, G. 1982. Migration between 8. Like in Brown (1994: 353), unofficial forms of Polynesia and New Zealand 1971-1981: Who are remittances or correspondences to them in my the migrants? New Zealand Population Review 1 investigation include money sent delivered by (1): 35-43. mail or by hand, goods sent, money carried by Bertram, G. 1986. Sustainable development in Pacific the migrant, goods carried by the migrant, pay micro-economies. World Development 14(7): 809 ments made by the migrant on behalf of a - 822. family member, gifts in the form of air fares Bertram, G. 1993. Sustainability, aid, and material and board and lodging for overseas visits by welfare in small south Pacific island economies, family or others to visit the migrant. 1900-90. World Development 21(2): 247-258. 9. District officers are nominated as representa Bertram, G. 1999. The MIRAB model twelve years on. tives of the government for districts which 11(1): 105-138. consist of 15-20 villages. Usually the `district Bertram, I. G., and Watters, R. F. 1985. The MIRAB officer' post is passed on from father to son; economy in south Pacific microstates. Pacific hence they have a very good knowledge of Viewpoint 26(3): 497-519. their respective districts. The assistance from Bertram, I. G., and Watters, R. F. 1986. The MIRAB the officers was very helpful, but the sampling process: Earlier analyses in context. Pacific View method assisted by them may be biased. This point 27(1): 47-59. was inevitable in limited survey time. Boyle, P., Half acree, K., and Robinson, V. 1998. Explor 10. Noteworthy in this present study is that, while ing contemporary migration. New York: Addison regular remittances are not significant, irregu Wesley Longman Limited. lar remittances are still significant; This differs Brown, R. 2002. In personal communication. Man from previous findings (Tongamoa 1987; Faea ager for immigration services, Tonga Branch. 6 mani 1993) that show remittances were highly August, Nuku'alofa. regular and seldom requested by the recipients. Brown, R. P. C. 1994. Migrants' remittances, savings The occasional remittances for such events ap and investment in the south Pacific. International pear to confirm the tie of extended families, but Labour Review 133(3): 347-367. its importance seems to be declining. Brown, R. P. C. 1995. Hidden foreign exchange flows: Estimating unofficial remittances to Tonga and References Western Samoa. Asian and Pacific Migration Jour nal 4(1): 35-54. Appleyard, R. T., and Stahl, C. W.1995. South Pacific Brown, R. P. C. 1997. Estimating remittance func migration: New Zealand experience and implica tions for Pacific island migrants. World Develop tions for Australia. International Development ment 25(4): 613-626. Issues no. 42, Australian Agency for Interna Brown, R. P. C. 1998a. Do migrant's remittances de tional Development (AusAid). : CPN cline over time? Evidence from Tongans and Publications. Western Samoans in Australia. The Contempo Asian Development Bank. 2003. Economic trends and rary Pacific 10(1): 107-151. prospects in developing Asia. Asian Development Brown, R. P. C. 1998b. Comparative labor market Outlook 2003: Tonga. 194-196. Asian Develop performance of visaed and non-visaed migrants:

365 144 R. L. 'ESAU

Pacific islanders in . Journal of Population benefits than harm? In Current trends in south Eccnomics 11: 395-411. Pacific migration, ed. V. Naidu, E. Vasta, and C. Brown, R. P. C., and Connell, J. 2004. The migration of Hawksley, 41-53. Asia Pacific Migration Re doctors and nurses from South Pacific islands search Network Working Paper No. 7. University nations. Social Science and Medicine. In press. of Wollongong, Australia. Brown, R. P. C., and Foster, J. 1995. Some common Fletcher, M. 1995. Pacific islands peoples in the labor fallacies about migrants' remittances in the south market. Labor Market Bulletin l: 124-136. Pacific: Lessons from Tongan and Western Sa Fuka, M. L. A. 1984. The Auckland Tongan commu moan research. Pacific Viewpoint 36(1): 29-45. nity and overseas remittances. Unpublished MA Campbell, I. C. 2001. Island Kingdom: Tonga ancient Thesis, University of Auckland, New Zealand. and modern. 2nd ed. Christchurch: Canturbury Government of Tonga Official Website 2002, http:// University Press. www.pmo.gov.to Chetty, N. K., and Prasad, S. 1993. Fiji's emigration: Government of Tonga Budget Statement. 2002. An examination of contemporary trends and issues. Budget for the year ended 30th 2002. Nu Demographic report 4. : University of the ku'alofa: Government of Tonga. South Pacific, School of Social Economic Devel Haberkorn, G. 1997. A sea of islands: A myriad of opment. indicators on the interface between demography Cook, P., and Kirkpatrick, C. 1998. Labor market and planning in the Pacific islands. Asia Pacific adjustment in small open economies: The case of Viewpoint 38(3): 219-236. Micronesia. World Development 26(5): 845-855. Hau'ofa, E. 1993. Our sea of islands. In A new Connell, J. 1983a. Migration, employment and develop : Rediscovering our seas of islands, ed. E. ment in the south Pacific. Country report no. 18, Waddell, V. Naidu, and E. Hau'ofa, 2-16. A USP Tonga. Noumea, New Caledonia: South Pacific 25th anniversary publication. Suva: School of Commission and International Labour Organiza Social and Economic Development, The Univer tion. sity of the South Pacific in association with Beake Connell, J.1983b. Migration, employment and develop House. ment in the south Pacific. Country report no.22. Hayes, G. 1992. The use of scientific models in the Samoa. Noumea, New Caledonia: South Pacific study of Polynesian migration. Asian and Pacific Commission and International labour organiza Migration Journal 1(2): 278-311. tion. Hayes, G. 1993a. Polynesian migration and the demo Connell, J. 1987. Paradise left? Pacific islands, voyag graphic transition: A missing dimension of recent ers in the modern world. In Pacific bridges: The theoretical models. Pacific Viewpoint 34(1): 1-35 new immigration from Asia and the Pacific islands, Hayes, G. 1993b. MIRAB processes and development ed. J. T. Fawcett and B. V. Carino. New York: on small Pacific islands: A case study from the Center for Migration Studies. southern Massim, Papua New Guinea. Pacific Connell, J. 1990. Modernity and its discontents: mi Viewpoint 34(2): 153-178. gration changes in the south Pacific. In Migra Hooper, A. 1993. The MIRAB transitions in , tion, employment and development in the south Pa Tokelau. Pacific Viewpoint 34(2): 241-264. cific, ed. J. Connell. Pacific Research Monograph James, K. E. 1991. Migration and remittances: a Ton No. 24. Canberra: Australia National University. gan village perspective. Pacific Viewpoint 32 (1): Connell, J., and Conway, D. 2000. Migration and 1-23. remittance in Islands microstates: A comparative James, K. E. 1998. Pacific islands stakeholder partici perspective on the South Pacific and the Car pation in development: Tonga. Pacific islands dis ribean. International Journal of Urban and Re cussion paper series. Washington: World Bank. gional Research 24(1): 53-78. James, K. E. 2002. Disentangling the grass roots in Evans, M. 2001. Persistence of the gift: Tongan tradi Tonga: Traditional enterprise and autonomy in tion in transnational context. Ontario: Wilfred the moral and market economy. Asia Pacific Laurier University Press. Viewpoint 43(3): 269-292. Faeamani, S. U. 1993. Migrant remittances: The engine Kishida, Y. 2000. Tonga no kaigai imin to kyokai. of village development in Tonga. Unpublished MA Unpublished MA Thesis, Kobe University, Japan. Thesis, University, Australia. Laplagne, P. 1997. Dutch disease in the south Pacific: Faeamani, S. 1995. The impact of remittances on evidence from the 1980s and beyond. Economic rural development in Tongan villages. Asian and Bulletin 12(1): 84-96. Pacific Migration Journal 4(1): 139-155. Levick, W., and Bedford, R. D. 1988. Fiji labour migra Fensterseifer, C. 1993. Western Samoa's economic tion to New Zealand in the 1980s. New Zealand development: Contemporary constraints, histori Geographer 44 (1): 14-21. cal conflicts and new opportunities. New Zealand Marcus, G. E. 1974. A hidden dimension of family Journal of Geography 95: 14-18. development in the modern Kingdom of Tonga. Fisi'iahi, F. V. 2001. Migration from Tonga: of more Journal of Comparative Family Studies 5(1): 87

366 International Migration from Tonga 145

102. mea, New Caledonia: Demography/Population Marcus, G. E. 1981. Power on the extreme periphery: Programme, Secretariat of the Pacific Commu The perspective of Tongan elites in the modern nity. http://www.spc.int/demog/. world systems. Pacific Viewpoint 22(1): 48-64. Small, C. A. 1997. Voyages from Tongan villages to Marcus, G. E. 1993. Tonga's contemporary globaliz American suburbs. Ithaca: Cornell University ing strategies: Trading on sovereignty amidst in Press. ternational migration. In Contemporary Pacific Statistics New Zealand. 2001. Population Census. societies: Studies in development and change, ed. V. http://www.stats.govt.nz/ S. Lockwood, T. G. Harding, and B. T. Wallace, Statistics New Zealand. 2002. Pacific progress: A re 21-33. New : Prentice-Hall. port on the economic status of Pacific peoples in McCall, G.1996. Clearing confusion in a disembedded New Zealand. Te Tani Tatau, : Statis world: the case for nissology. Geographische tics New Zealand. Zeitschrift 84(2): 74-85. Sudo, K. 1997. Expanding international migration by McCall, G. 1997. Nissology: A debate and discourse the Tongan people: Strategies and socio-cultural from below. Invited paper at the cultural heri effects on the homeland. In Contemporary migra tage in islands and small states Conference. Val tion in Oceania: Diaspora and network, ed. K. Sudo letta, , 8-10 May. http://www.arts.unsw. and S. Yoshida, Osaka: The Japan Center for Area edu.au/southpacific/nissology.htm. Studies, National Museum of Ethnology. Mohanty, M. 2001. Contemporary emigration from Thomas, F. R. 2002. Self-reliance in Kiribati: Con Fiji: Some trends and issues in the post trasting views of agricultural and fisheries pro independence era. In Current trends in south Pa duction. The Geographical Journal 168(2): 163 cific migration, ed. V. Naidu, E. Vasta, and C. -177. Hawksley, 54-73. Asia Pacific Migration Re Tongamoa, S. T. 1987. Migration, remittance and de search Network Working Paper No. 7. University velopment: A Tongan perspective. Unpublished of Wollongong, Australia. MA Thesis, University of Sydney, Australia. Morton, H. 1996. Becoming Tongan: An ethnography of Tonga Statistics Department. 1986. Population census childhood. : University of Press. 1986. Bulletin no.1. Nuku'alofa: Statistics De Morton, H. 1998. Creating their own culture: di partment. asporic Tongans. The Contemporary Pacific 10(1): Tonga Statistics Department. 1996. Population census 1-30. 1996: Administrative Report and General Tables. Muliana, T. 2001. Remittances, the social system and Nuku'alofa: Statistics Department. development in Samoa. In Current trends in south Tonga Statistics Department. 1999. Tonga population Pacific migration, ed. V. Naidu, E. Vasta, and C. census 1996: Household analysis. Nuku'alofa: Sta Hawksley, 20-40. Asia Pacific Migration Re tistics Department. search Network Working Paper No. 7. University Tonga Statistics Department. 2001. Statistical indica of Wollongong, Australia. tors 2001. Nuku'alofa: Statistics Department. National Reserve Bank of Tonga. 1999. Annual report Treadgold, M. L. 1999. Breaking out of the MIRAB 1999/2000. mould: historical evidence from island. National Reserve Bank of Tonga. 2000. Quarterly Asia Pacific Viewpoint 40(3): 235-249. bulletin, December 11 (4). Trim, A. D. 1987. New Zealand's admission of Asians New Zealand Immigration. 2002a. New Zealand immi and Pacific islanders. In Pacific bridges: The new gration service fact pack, Issue 16. Wellington: immigration from Asia and the Pacific islands, ed. New Zealand Immigration. J. T. Fawcett and B. V. Carino. New York: Center New Zealand Immigration. 2002b. Migrating to New for Migration Studies. Zealand: visa categories. http://www/immigra United States Census Bureau. 2001. Native Hawaiian tion.gov.nz and other population: 2000. Poirine, B. 1994. Rent, emigration and unemploy http://www.census. gov/population/www/cen ment in small islands: The MIRAB model and the 2000/briefs.html French overseas departments and territories. Va'a, L. F. 1992. The future of Western Samoan mi World Development 22(12): 1997-2009. gration to New Zealand. Asian and Pacific Migra Poirine, B. 1998. Should we hate or love MIRAB? The tion Journal 1(2): 313-332. Contemporary Pacific 10(1): 65-105. Vaden, B. D. 1998. Kainga: Tongan families as agents Secretariat of the . 2001. Popula of change. Unpublished PhD Dissertation. A Bell tion and development planning in the Pacific. Nou and Howell Company, Michigan.

367