Why Don't Pacific Island Countries' Economies Grow Faster.Pdf

Pacific Interactions

Pasifika in New Zealand – New Zealand in Pasifika

Edited by

Alastair Bisley

Institute of Policy Studies

Published in 2008 (online only) http://ips.ac.nz

Institute of Policy Studies School of Government Victoria University of Wellington PO Box 600 Wellington

IPS/Pub/159

This book is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced without the permission of the Institute of Policy Studies.

Copy editor: Belinda Hill Cover design: Alltex Design

Contents

List of Figures...... iv List of Maps...... v List of Tables...... v List of Boxes...... viii Acknowledgements ...... ix

Introduction –Alastair Bisley...... 1

1 Emerging Demographic and Socioeconomic Features of the Pacific Population in New Zealand – Paul Callister and Robert Didham ...... 13

2 Pacific People in the New Zealand Economy: Understanding linkages and trends – Jean-Pierre de Raad and Mark Walton ...... 41

3 Pasifika Mobility: Pathways, circuits, and challenges in the 21st century – Richard Bedford ...... 85

4 Pacific Island Economies: Role of international trade and investment – Bob Warner...... 135

5 Why Don’t Pacific Island Countries’ Economies Grow Faster? – John Gibson and Karen Nero ...... 191 6 Border Management in the Pacific Region – Michael Moriarty ...... 245

Appendix A: Thought Leaders Dialogue, Auckland, New Zealand, 2007...... 289 Appendix B: Keynote Address by Greg Urwin, Secretary General, Pacific Islands Forum Secretariat, at the Pasifika Project: Auckland Dialogue...... 297 Collated References...... 307

Why Don’t Pacific Island Countries’ Economies Grow Faster?

John Gibson1 and Karen Nero2

Contents

Introduction ...... 191 Changing perceptions ...... 193 Pacific systems of wealth ...... 215 Implications and possible interventions ...... 227 Conclusions ...... 236 Annex: Improving development data ...... 233 References ...... 236

Introduction This chapter confronts the issue that the recent rate of growth of the economies of the Pacific Island countries is lower either than what is desired by their governments or what might reasonably be expected due to their proximity to global regions of rapid growth. Pacific governments and Pacific people, both in the islands and in New Zealand, seek ways to achieve sustainable economic growth so that they can fund improvements in education, health, and general living standards. Standard statistical assessments focused on the market economy concur that Pacific economies perform poorly, but may be misleading due to poor data on local food production and distribution that supports up to 84% of some Pacific populations. A better understanding of the value of local food and wealth systems may improve the quality of statistical assessments and assist planners to better support economic growth. We also analyse how poorly the region’s nations are doing in comparison with other similar nations (and how such similarity is best measured) and the degree to which governance issues contribute to poor performance. Could the island economies reasonably be expected to do better? If so, what national, regional, and international policy actions could support such improvements? We consider how best to frame our understandings of the region, using multiple perspectives and aided by geographical information system (GIS) maps of the region that challenge the reader to consider the complex interactions among

1 University of Waikato and Visiting Research Fellow, Macmillan Brown Centre for Pacific Studies, University of Canterbury. 2 Professor Macmillan Brown Centre for Pacific Studies, University of Canterbury.

© Institute of Policy Studies, Victoria University of Wellington, 2008 191 Pacific Interactions multiple factors. Our analyses have a large geographic component. This is appropriate given that this region of small volcanic and atoll islands plus the few large continental islands of the southwest Pacific covers one-third of the globe. Analyses of spatio-economic factors (ie, size, natural endowments, economic remoteness and the type of relationship with and economic strength of former colonial powers) emerge as important factors in explaining growth performance. We present some of these key variables in the form of thematic maps. New Zealand is host to the world’s largest Polynesian population, and has historically focused on the Polynesian world views of the Pacific region of its Māori indigenous population and of its resident populations from Samoa, Tonga and those Pacific nations freely associated with New Zealand (the Cook Islands and Niue) or considering such status (Tokelau). The ways in which New Zealand academics and members of the public frame the questions they use to understand the economic choices made by Pacific peoples and home nations tend to be coloured by their experiences and knowledge of Polynesian world views. We argue that building the economic strength of the entire region is critical to the success of any one nation. In particular, we must consider the large island of New Guinea, with its large population, diverse societies and habitats, and rich mineral and timber resources, as well as neighbouring Solomon Islands and Vanuatu. In the first section, entitled ‘Changing perceptions’, we consider ways to expand the base from which we understand the larger region. We begin with the recognition of western Pacific peoples’ millennia of experience of adapting to changing sea levels and cultivating, domesticating, and hybridising food crops suited to both high island and atoll conditions, as well as adapting to new arrivals and the resources and technologies they bring. A series of GIS maps suggests limits to our perceptions and how we might perceive wider Pacific realities, history, and knowledge more clearly. We assess recent negative analyses of Pacific economies and suggest alternative analyses of existing data. In the second section, entitled ‘Pacific systems of wealth’, we provide overviews and case studies to demonstrate how Pacific wealth systems work. Pacific societies build their social, cultural, and political institutions on the base of food production and distribution. As these foods are transmitted beyond the immediate family through kin, village communities, and trading partners both within and outside their societies, they are in the process transformed and exchanged for social capital as well as for indigenous wealth items and global goods and monies. Over the millennia, Pacific societies have developed their own monies and currencies for storing and transacting wealth across generations and across space. They include textiles such as the well-known tapa and fine mats of Polynesia, as well as hard wealth produced from stone (ie, greenstone, or pounamu in New Zealand) and the shell and bead currencies better known in other parts of the Pacific. Today’s exchanges and interactions at home and abroad draw upon both indigenous and global goods and monies. Mainly because of the difficulty of counting and understanding the economic contributions of these activities embedded in social and cultural institutions, economists and statisticians often characterise these as subsistence activities, and

192 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? their extensions – monetised today in complicated systems that bridge to Western capital and monetary institutions – as forming the ‘informal economy’. Such transactions and wealth are generally uncounted or, at best, sampled from time to time in attempts to estimate their value in Western terms. In this vein, we consider recent assessments of the degree to which agricultural and fisheries production may be undercounted in the region. In Western economic terminology, Pacific people today produce food for subsistence (eg, shellfish), current accounts (eg, lagoon fish sold for routine cash requirements), and capital accounts (eg, pelagic fish sold for major expenditures such as school fees, medical costs, or housing). Some food items may be used in only one such arena but most starch (yam and taro) and protein (fish and pork) foods may be allocated across each arena to suit current needs. Shellfish is also a complicated resource: the meat is locally eaten and rarely sold, but the shells (ie, pearl shell and clam) are sold as well as being made into currencies. Pacific families, spread throughout island nation-states and abroad, transfer indigenous valuables, money, and foods. We, therefore, briefly consider case studies through which families and local communities over time develop transnational capital by negotiating food, commodity, and valuable trading through trans-local members. In the third section, entitled ‘Implications and possible interventions’, we consider the practical implications of these approaches. How might we maximise the impact of interventions? We question whether remittances, especially within national economies, provide a ‘safety net’. Transfers among kin are fraught with risk, especially when the highly mobile and less visible, international currencies that intersect with international banking and transfer mechanisms are used. How are remittances transmitted across national boundaries, and what mechanisms could be used to reduce the costs of such transactions?

Changing perceptions

Pacific region in perspective In our New Zealand focus on Polynesia – the region most recently settled by navigators and settlers who moved into the Fiji–Samoa–Tonga triangle around 3,000 years ago – we often lose track of the history of the earliest settlement in the region and of Pacific adaptations to changing habitats. We also forget that the major Pacific populations, and economic power, are found on the large islands of the Melanesian islands to the west. In fact, the largest single population group in the Pacific is the Highlanders of Papua New Guinea who now number well over two million – as many as the combined total population of all other countries in the Pacific Islands. This population lives far from beaches and coconut trees, at altitudes above 1,200 metres and cultivates many temperate crops. The part of the Pacific that New Zealand knows best – Polynesia – is a tiny and not necessarily representative fraction of the total Pacific. Specifically, in

© Institute of Policy Studies, Victoria University of Wellington, 2008 193 Pacific Interactions terms of population, Polynesia is only 4% of the total (Figure 5.1).3 Although Polynesians are richer on average than Melanesians, even in economic terms Polynesia is dwarfed by the Melanesian nations (Polynesia is home to 9% of regional gross domestic product (GDP)).

Figure 5.1: Pacific population and gross domestic product, ca 2004 and 2005

Population (ca 2005)

Melanesia

Micronesia Polynesia Fiji PNG Solomon Is. Vanuatu

Total GDP (US$ ca 2004)

` Melanesia

Micronesia Polynesia Fiji PNG Solomon Is. Vanuatu

Source: Calculations from data presented in ADB (2006) and AusAID (2006).

Pacific region reconsidered

Near and Remote Oceania The Pacific region covers one-third of the globe and is home to both one of the largest (Austronesian) and most diverse (Papuan or non-Austronesian) language families. It is too large and complex to be understood through simple categories. Early Europeans divided the region into the geographically based sub-regions of

3 These population and economic data are for the 13 Pacific states that are members of the Asian Development Bank, so exclude French Polynesia and New Caledonia, and some of the smallest counties which would make no difference to the regional comparisons illustrated in Figure 5.1.

194 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Melanesia, Micronesia and Polynesia, implying similarities within regions (especially Melanesia and Micronesia) that are extremely diverse. This approach also downplays commonalities throughout the Pacific. In 1973, Pawley and Green (1973) introduced the terms Near Oceania and Remote Oceania to call attention to these two regions with very different resources as well as different times of settlement. At the lower sea levels prevailing in the Pleistocene era, settlers travelled east from the Sunda landmass (including Java, Borneo) through the Indonesian islands before the sea crossing to the Sahul land mass that included the Australian continent, Tasmania and New Guinea. This area of Near Oceania, shaded in Map 5.1, was settled 40,000–45,000 years ago (Green, 1991). Near Oceania (and New Zealand) is entirely to the west of one of the region’s defining geological features, the Andesite Line, which divides the larger continental landmasses made up of felsic andesitic rocks from the smaller oceanic or Pacific plate islands of mafic balsatic volcanic rocks that erupted from the ocean floor of the Central Pacific basin. In contrast to Near Oceania, the smaller islands of the Pacific basin in Remote Oceania were settled within the last 4,000 years, including the more recently established coral atolls and islets that in tropical waters slowly grew up around the submerging volcanic islands.

Map 5.1: Settlement of Near Oceania and Remote Oceania

Long history of adaptation to the land and oceanic habitat The settlers of New Guinea experienced an uninterrupted period of human adaptation to a series of changing environments from 50,000 years ago to the present. Their offshore islands were home to navigators and sailors, who have

© Institute of Policy Studies, Victoria University of Wellington, 2008 195 Pacific Interactions continued their long-distance voyaging to the present day. Recent studies have finally recognised New Guinea as one of the world’s four earliest sites of independent development of agriculture (Mazoyer and Roudart, 2006, p 75). Irrigation structures to support root crops, dating to 9,000 years ago in the Kuk swamp region of Papua New Guinea, were part of networks of food gardens that are still visible from the air. The contributions of the region’s agriculturalists include the development of tuber-based agriculture during the early Holocene period (see Bellwood (1996) and Dewar (2003) for current debates) and the domestication of sugar, one of the world’s major foods, on New Guinea, though these developments are barely mentioned in standard references (ie, Mintz, 1986). Melanesian farmers were the first to domesticate several important species of taro and probably breadfruit. Joined by later settlers from what are now the islands of Taiwan and Indonesia, subsequent groups eventually settled the entire region, successfully adapted to major environmental transformations (ie, the sea-level changes of the Pleistocene and El Niño rainfall variability), and adapted both the plant biota and their agricultural techniques. Nor was agricultural innovation limited to the high islands. Additional innovations were required as around 4,000 years ago settlers began to settle the oceanic islands and coral atolls of Remote Oceania, seeking new plants and adapting their cultivars to suit the atoll and high island environments they encountered. Current research indicates early settlement from island Indonesia into the Palau islands by 1500 BC (Clark et al, 2006), closely followed by the settlement of the Bismarck Islands, Solomon Islands and Vanuatu. By 1000 BC settlers reached New Caledonia, Fiji, and Tonga. They then settled first Samoa and then the distant islands of the Polynesian triangle (Howe, 2006). Archaeological, botanical, and genetic evidence indicates that Polynesian navigators voyaged to South America, bringing back the sweet potato (Ipomoea batatas) found in archaeological sites dating back to 1000–1100 AD in the Cook Islands (Kirch, 2002) and beyond throughout Polynesia (see Ladefoged et al (2005) and Ballard et al (2005)), and reportedly bringing the Polynesian chicken to Chile (Storey et al, 2007). Sweet potato became a staple that supported dense populations and regional exchange systems and transformed New Guinea societies (Ballard et al, 2005). Agricultural innovations were also critically important to atoll dwellers. In the Caroline and Marshall Islands major changes from 1000 AD to 1500 AD are associated with hybridisation to create a salt resistant breadfruit species, leading to a breadfruit revolution in the islands, and major demographic, political, and social transformations. Master navigators maintained trade links connecting the atolls of the Caroline and Marshall Islands with the continental western Micronesian islands of Palau and the Marianas 4,800 km to the west. There they found a salt- resistant type of breadfruit that they hybridised with their own salt-intolerant one. Early voyagers then distributed the new cultigens back throughout the western Micronesian islands and as far as the Polynesian atolls of Tuvalu and Tokelau to the south (Petersen, 2006).

196 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

The long history of Pacific settlement that can be read from archaeological, linguistic, botanical, and oral histories is one of a series of interlinked trading networks. Innovations were quickly transmitted throughout the region through long-distance canoe voyages that ranged from New Guinea in the west to as far east as South America, and connected the north and south Pacific Islands. Despite long-term sea-level and climatic changes, most of the region was successfully settled by growing populations. The relatively recent European exploration of the region beginning in the 16th century introduced a series of demographic, botanical, transportation, political and economic effects. Early population declines due to introduced diseases were widespread and catastrophic in some islands. As Spanish, British, French, German, and Japanese colonial administrations were established in the region, travel was often curtailed, and steamships replaced canoe travel in much of the Pacific. The acceptance of Christianity in the islands brought about significant ideological transformations as indigenous and introduced religions accommodated each together. The introduction of capitalist and monopolist colonial political economic regimes significantly changed the operation of long-established socioeconomic systems focused on agricultural, fisheries, and long-distance trade networks throughout the region. The impacts of these influences differ by islands and by sub-region.

Mapping of regional variables We use a series of GIS maps to engage the reader in reconsidering the complex ways in which geographical factors (ie, size and type of island, proximity to other islands and strong economies, and land and sea resources) intersect with geopolitical factors (eg, the nature of ongoing links to the former coloniser) in this post-colonial world. These maps consider relationships between population density, population growth, income, poverty, urbanisation, education, and migration, and show how these factors play out differently in different economies with different historical linkages. Due to the small and almost invisible outlines of the smaller Pacific states at this map scale, the island nation’s land mass is shown with the 12-mile limit and encircled within its 200-mile exclusive economic zone (EEZ) to depict country borders in all maps.4 One complication of using the EEZ is that the EEZ for Kiribati is not contiguous. Map 5.2, demonstrating per capita GDP of Pacific Islands Forum countries exemplifies the type of map used throughout – on which have colour-coded into five levels the variable/s under consideration.

4 These EEZ geographic data were obtained from the Pacific Islands Applied Geoscience Commission (SOPAC) website: http://www.sopac.org. Population data for these maps are from the Secretariat of the Pacific Community, and economic indicators from the` Development Bank.

Map 5.2: Pacific region by gross domestic product (GDP) per capita (US$)

Further maps invite the consideration of the effects of spatio-economic factors (ie, size, natural endowments, environmental fragility, and vulnerability) in combination with economic remoteness. Socio-cultural factors (ie, size and diversity of residents, rates of migration and overseas workers, and cultural practices) considered include type of relationship and economic strength of former colonial powers. Gross national income (GNI) rather than GDP is generally considered a better indicator for most Pacific nations, so GNI is used in the maps following. Map 5.3 indicates that generally low population density is correlated with low per capita GNI. An important exception is Palau, which enjoys one of the highest per capita GNI, even with a low density of population. Partly this is related to Palau’s proximity to Asia both as a source of low-cost unskilled and professional workers and of incoming tourists from Japan and Taiwan. Geopolitical factors are discussed with Map 5.5.

Effects of urbanisation Another key variable within the Pacific is the degree of urbanisation (Map 5.4). The lowest urbanisation rates are generally in Melanesia and the highest rates in Micronesia, although there are exceptions to this pattern. The population growth rates (or loss rates for those nations with data shown below the bar) that overlay the maps of urbanisation rates also demonstrate the ongoing movement of especially eastern Polynesians overseas. The countries with either declining or static total populations are also among the most urbanised countries.

198 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Map 5.3: Population density (thousands per square kilometre) by gross national income per capita (US$)

Map 5.4: Share of country population in urban areas and population growth rates

The depiction of Pacific Islands Forum Nations by GNI per capita and by urban–rural distribution suggests anomalies in apparent economic performance that may point to complex interactions of both political and geographical realities. For example, in Map 5.5, Palau and the Cook Islands are the top two Pacific Islands Forum nations based on GNI. These two countries are also among the most urbanised. In this view, the larger nations of Papua New Guinea and Solomon Islands that enjoy much richer land resources are at the lowest levels of performance. Both Palau and the Cook Islands are freely associated with their former coloniser (the United States and New Zealand respectively). However, other freely associated states (the Federated States of Micronesia, the Marshall Islands, and Niue) are less successful, so this factor by itself is insufficient to explain the variations in performance.

Map 5.5: Pacific region by gross national income (GNI) and urban–rural distribution

In much of the Pacific, rural populations are considered the poorest and most at risk, as seen in Map 5.6, which refocuses our attention to the percentage of the population identified as being under the poverty line (with national, urban, and rural rates shown separately, if available). The height of the bar shows the poverty rate (high in Kiribati, Palau, Vanuatu, and Papua New Guinea while low in the Cook Islands and Samoa) while the comparison of bars for a given country shows the distribution of poverty (thus, in Papua New Guinea, poverty is high in rural areas, while in Fiji it is somewhat the opposite pattern). A key focus of donor strategies is to reduce poverty. In this regard, Map 5.6 demonstrates some of the political issues associated with data and comparable definitions. In Map 5.5, we saw that Palau and the Cook Islands are the highest-ranking Pacific nations in terms of GNI per capita. Yet, at the same time, Palau is shown as having one of

200 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? the highest poverty rates. This potential anomaly is circled on Map 5.6 to draw attention to it. Does this contrast between high average income levels and high reported poverty indicate extreme inequalities within this small nation? Not really, and instead the contrast highlights problems with the data available for many of these cross-country comparisons. The Asian Development Bank data set on Palauan poverty that year used estimates based on United States poverty thresholds. These refer to a far higher standard of living than the poverty lines used for other countries in the Pacific, so undermine the comparability of this key indicator.

Map 5.6: Population below poverty line on per capita gross national income (GNI) base

Complex interactions of education and emigration Average education levels are the most important correlate of economic growth across countries, according to recent research (Lleras-Money, 2005). We attempted to provide a GIS map of educational levels (Map 5.6), although comparable data on educational attainment is hard to obtain for smaller Pacific Island nations (Docquier and Malfouk, 2006). Therefore, we have used data on adult literacy rates (Map 5.7). In general, the lowest literacy rates are correlated with the lowest per capita income levels (although Kiribati is an exception to this). With the exception of Papua New Guinea, Solomon Islands, and Vanuatu, the gender gap in literacy levels appears to be small.

Map 5.7: Adult literacy rate by gender, on per capita gross national income (GNI) base

Map 5.8: Share of total emigrants by levels of educational attainment

202 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

The other feature that relates to both education and literacy levels, and indeed population mobility, is the potential ‘brain drain’ of talented people from Pacific countries. In this regard, there is clear evidence that emigration is most skill intensive (in terms of being dominated by tertiary educated people) from the countries with the most limited emigration options (Map 5.8). Thus, over one-half of emigrants from Papua New Guinea are tertiary educated and almost one-half from Solomon Islands and Vanuatu, and these three countries have the lowest overall emigration rates. In contrast, less than one-quarter of emigrants from the countries with the highest emigration rates (Samoa and Tonga) are tertiary educated. Thus, paradoxically, restrictions on mobility, which bind most heavily on Melanesia (except for Fiji) might also contribute to the largest losses through ‘brain drain’. Another interesting contrast is with GNI (not shown on this map). Higher average income levels may affect the ability of a country to retain its tertiary- educated population. For example, although Palau has a higher emigration rate than the Federated States of Micronesia and Kiribati, the percentage of FSM or I-Kiribati out-migrants with tertiary education is nearly double that of Palau. With more jobs available in Palau and likely higher pay resulting from the higher GNI, Palau has been able to attract some of its educated former out-migrants to return.

Effects of mobility barriers The rates of emigration, as well as the educational level of those who emigrate are responsive to international incentives and barriers to short- and long-term mobility. In turn, these choices affect both the home communities and the perception of immigrants in their destination countries. Map 5.9 demonstrates the affect of the degree of access to overseas countries for living and employment. Access to employment is critical, especially to those nations including Papua New Guinea, Vanuatu, Solomon Islands, and Kiribati who do not have established rights to emigrate. Emigration rates, even for those with such rights, under free association or territorial status (the Cook Islands, Niue, Tokelau to New Zealand; Palau, the Federated States of Micronesia, the Marshall Islands, Guam, and the Northern Mariana Islands to the United States; the French territories to France; and Rapa Nui to Chile) vary according to effects of other factors, including cultural preferences and economic choices. Interestingly, emigration options do not follow pathways only to the former coloniser (Map 5.9). As expected, out-migration from Palau, the Federated States of Micronesia, and the Marshall Islands is nearly exclusively to the United States. However, over half of the out-migrants from Kiribati also migrate to the United States. It is striking that a high percentage of the small numbers of Solomon Islanders who emigrate go to Europe, nearly equal to the percentage from Vanuatu.

Map 5.9: Migrant destinations on base of migration rate (emigrants per 100 residents)

External assessments of Pacific economies reconsidered Some previous external assessments of Pacific economies, which are discussed below, may have been unduly pessimistic. Therefore, we re-evaluate their performance but with a broader focus. First, we allow a more explicit role for geography and space in an empirical analysis of growth and governance in the region. Second, we attempt to bridge the historical disjuncture between Western understandings and valuation of Pacific food production systems (see ‘Pacific systems of wealth’) and attempts to ‘develop’ the systems, and more recent archaeological, anthropological and development analyses. We expect that with better understanding of these contemporary production and wealth systems, they can be supported and developed through international aid and trade, rather than undermined in ways that retain the legacies of earlier colonial eras through which labourers, minerals (eg, phosphate), and agricultural products (eg, sugar and coconuts) were removed from the region. Third, while recognising hardship and poverty in the Pacific (see especially Abbott and Pollard, 2004) and avoiding simplistic views of rarely attained sustainable rural development resulting in shared wealth (Schoeffel, 1997), we argue that a greater analytical focus on the intersections between indigenous production sectors and national and international transactions may better support effective development strategies, for example, how may remittances from Pacific family members overseas be best understood and maximised? What are the contributions of women and the exchanges of the women’s production of textile wealth? Many Pacific Island economies have found difficulty in achieving sustained economic growth. For example, Sampson (2005) reports that after controlling for

204 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

OECD membership and whether a country is an oil exporter, the Pacific states grew more slowly than countries in any other region of the world over 1995– 2003. Weak governance is often blamed for this performance. For example, according to estimates made by Duncan (2005), in the absence of poor governance, per capita GDP in Papua New Guinea would have been double what it was in 2003 and one-third higher for Fiji. Moreover, this lack of growth may appear puzzling given the high per capita inflows of external finance, in the form of both overseas aid and remittances. Evidence on these in-flows of external finance is shown in Figure 5.2 for the most recent year available for each of the Melanesian and Polynesian countries and for Kiribati. Data on remittances to the other Micronesian countries is less easily available from international sources. Even for the countries with reported data, it is likely that remittances are understated due to some transfers occurring through informal means such as travellers carrying cash back with them. Moreover, the data in Figure 5.2 relates only to monetary remittances rather than the provision of goods in kind, for which there is no comparable cross-country data.

Figure 5.2: Financial flows in the Pacific

45 Aid Remittances Foreign Direct Investment

% of GDP 15

0 Kiribati Cook Samoa Tonga Tuvalu Fiji PNG Solomon Vanuatu -5 Islands Is.

Source: AusAID (2006); ADB (2006).

For the countries in Figure 5.2, aid is equivalent to 16.3% of GDP, on average. This is somewhat higher than remittances, which are 12.5% of GDP, and much higher than foreign direct investment, which is only 0.6% of GDP. Total remittances to the nine countries in Figure 5.2 are about US$350 million per year, with total aid at over US$400 million. Thus, across all Pacific Island economies (including those not in Figure 5.3) it is likely that the combined total external finance is almost NZ$200 per capita, which is very high by world standards. To have such modest economic growth with this generous inflow of external finance has many experts placing the blame on poor institutions (Chand, 2001),

© Institute of Policy Studies, Victoria University of Wellington, 2008 205 Pacific Interactions aid inflows (Hughes, 2003), bad policy settings (Chand, 2003; Gosarevski et al, 2004) and especially governance failures (Duncan, 2005). Smallness and insularity may also be problems in the Pacific. Small size means a small domestic market and dependence on export markets and a narrow range of products, limited ability to influence prices, limited ability to exploit economies of scale, limitations on domestic competition, and problems of public administration (eg, a small labour force from which to draw experienced and efficient administrators). Insularity and remoteness give rise to problems associated with transport and communications such as high per-unit transport costs, uncertainties of supply, and the need to hold relatively large stocks (Briguglio, 1995). Despite these apparent disadvantages, Easterly and Kraay (2000) find that small states have the same range of growth experiences as other states. The absence of a clear growth disadvantage of small states is due to offsetting effects of openness to international trade, which is favourable to growth, and greater output volatility, which harms average growth rates. This output volatility is partly due to the relatively larger terms of trade shocks experienced by small states. Another factor, emphasised by Bertram (2004), is that the level of GDP per capita of small island economies depends directly on the GDP per capita of their metropolitan patron and on the strength of their political ties with the metropolitan patron. Bertram and Karagedikli (2004) show that on average across the Pacific Islands region, politically integrated units exhibit per capita incomes nine times higher than sovereign island states. Similarly, Armstrong and Read (2000) show that dependent territories have higher GNP per capita than the sovereign states, even when controlling for a range of other factors such as economic structure, island status, and aid transfers. Geography also matters. Armstrong and Read (2006) note that the Pacific states are more likely to be smaller, islands, archipelagos, remote, and mountainous than other small states, although these factors are not necessarily a serious growth handicap for the small Pacific states; remoteness, however, almost certainly is. Armstrong and Read measured remoteness as the great circle distance from the capital city of each small state to the nearest one of three global economic hubs (Tokyo, Japan; Washington or Los Angeles, United States; and Brussels, European Union). Since it is arguable that Auckland, Sydney, and even Brisbane are also important hubs for parts of the Pacific, it is clear that this measure of remoteness is only a partial one. It is, therefore, useful to describe just how economically remote the Pacific Island countries are, using more comprehensive measures. In subsequent parts of this chapter, these measures of remoteness are included in cross-country growth models to examine the hypothesis that geographic isolation is a fundamental growth constraint for Pacific Island economies.

Importance of location and remoteness: How remote are the Pacific Islands? The comprehensive measure of country remoteness used here is based on a matrix of bilateral distances for 219 countries with data available. Full details on the

206 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? construction of these distance measures are reported in Gibson (2007). The great- circle distance from each of the 19 Pacific Island states and territories (that have data available) to each of the other 218 countries in the rest of the world (ROW) is then weighted by either the GDP of each of those 218 countries or the population of each of those 218 countries.5 Table 5.1 reports the average distances for three Pacific Island groups, aggregated from country-level information. The average Pacific Island country is 11,500 km from any other randomly selected country (weighted by ROW GDP). Micronesian countries are slightly less remote than either Polynesian or Melanesian countries since more of the world’s GDP is in the northern hemisphere. Out of the 219 countries for which this calculation has been made, the Melanesian and Polynesian countries have an average rank as the 207th most remote, while the Micronesian countries have an average rank of 176th most remote (shown in numerical column 2 of Table 5.1). A useful comparison is with the Caribbean islands; the average Caribbean island is the 100th most remote country, compared with the average Pacific Island, which is the 197th most remote.

Table 5.1: Potential market remoteness measures for the Pacific and Caribbean Islands

GDP- Population- GDP- 2003 GDP- weighted weighted weighted weighted distance distance distance distance rank rank rank Island group (km) (2003) (2003) (1988)

Micronesia 10,377 176 146 179 Polynesia 11,942 207 201 208 Melanesia 11,972 207 170 208 Pacific Islands (mean) 11,456 197 176 199 Caribbean (mean) 8,103 100 176 98

Source: Summary averages for country-level results. Distances are the weighted average distances from the countries in each island group to the countries in the rest of the world (ROW), weighted either by the ROW country’s gross domestic product (GDP) or population. The rank is out of 219 countries, with number 1 the most accessible and number 219 the most remote.

The final two columns of Table 5.1 report the rankings for two different measures of distance – one using population weights rather than GDP and one using GDP 15 years earlier. Pacific Island countries, especially Micronesia, are less remote from population centres than they are from the centres of the world’s GDP. The values in the final column suggest a slight decline in potential remoteness for Pacific Island countries over the 15 years from 1988 to 2003. Their average remoteness rank improved from 209th to 207th, no doubt due to the rising share of world GDP located in Asia.

5 Silva and Tenreyro (2006) also used the GDP-weighted great circle average distance as a measure of remoteness.

Whether proximity to rich (or populous) countries equals actual accessibility depends on transport links. One metric for examining ‘economic distance’ based on transport links is to consider airfares. Map 5.10 shows the average airfares from each Pacific Island country to three ‘hubs’ surrounding the Pacific: Auckland, Sydney and San Francisco. According to these average cost calculations, the Pacific Island countries appear more remote than island countries in other parts of the world. While the Pacific Island countries are, on average, 40% further from the locations of world GDP than the island states in the Caribbean they are much further away in terms of airfare-based measures of distance (Table 5.2).

Map 5.10: Average cost of air travel in the Pacific

Table 5.2: Airfare measures of remoteness for Pacific and Caribbean Islands

Average fare to Average cost of Average fare to most distant of fares to three key closest of three three key Island metropolitan cities key metropolitan metropolitan cities group (US$) cities (US$) (US$)

Micronesia 1,773 1,768 1,361 Polynesia 1,067 727 1,699 Melanesia 905 585 1,586 Pacific Islands (mean) 1,289 1,084 1,534 Caribbean (mean) 545 395 902

208 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

On average, travelling from a Caribbean island to any one of three main metropolitan cities with strong links to the Caribbean (Miami, New York, and London) costs US$545. A similar calculation for the Pacific Islands, but with Auckland, San Francisco, and Sydney as the metropolitan cities, gives an average fare of US$1,289. Thus, by this metric the Pacific Island countries are more than twice as remote as the Caribbean islands. The cost disadvantage is even more apparent when considering travel to the closest metropolitan city, which is likely to act as a deterrent to both freer movement of labour (eg, seasonal migration schemes) and the export of services to in-bound tourists. This example serves as a warning that the potential market access and distance measures used in the rest of this chapter are imperfect proxies for more economic concepts of distance. However, such imperfect proxies have to be relied on because of limited data on better economic proxies for distance.

Importance of neighbours and regional interchanges: Re-analysis of economic growth in the Pacific Islands In this section, we use regression modelling to re-examine the economic growth performance of the Pacific Island economies. This analysis has two notable features. First, it uses the comprehensive measures of geographic remoteness reported above to test the hypothesis that geography is a fundamental constraint on growth in the Pacific. Second, it recognises that economic growth for a given country may depend on the growth of nearby countries (Conley and Ligon, 2002). Surprisingly, despite the spatial nature of cross-country data on economic growth there is little mention of such possible spillovers in the literature. Studies of the economic performance of small states (Easterly and Kraay, 2000; Armstrong and Read, 2002, 2006) ignore this feature, despite small states being clustered in the Caribbean and Pacific Oceans, so that omitted spatial interactions may bias estimates of the effects of ‘smallness’. Similarly, none of the studies of slow growth in the Pacific accounts for possible spatial effects despite the geographic clustering (by definition) of the countries in this region. Table 5.3 describes the variables used in the cross-country growth model and presents regression coefficients with and without spatial spillovers accounted for. A sample of 174 countries has full data available, including the remoteness measures described above and average annual GDP per capita growth rates over 1987–2003. The 15 Pacific Island countries in this sample (Papua New Guinea, Fiji, Solomon Islands, Vanuatu, Samoa, Tonga, Cook Islands, Tuvalu, French Polynesia, New Caledonia, Federated States of Micronesia, Kiribati, the Marshall Islands, Nauru, and Palau) have, on average, grown slightly more slowly in per capita terms than have other countries over 1987–2003. The selection of explanatory variables used in the model is guided by the results of previous studies of economic growth in either ‘small’ or Pacific Island economies, especially Easterly and Kraay (2000), Bertram (2004), and Armstrong and Read (2006). In addition to the distance measures described above, the other geographic variables included in the models are indicators for the Pacific and for landlocked countries. The next variables are the logarithm of population and the

© Institute of Policy Studies, Victoria University of Wellington, 2008 209 Pacific Interactions logarithm of land area, which implicitly gives a measure of population density. Trade openness and output volatility are included, along with the history of colonisation and growth of the former colonial power. Language diversity is included since this may increase the transactions costs of internal commerce, lower the returns to human capital formation, and constitute a proxy for ethnic fractionalisation that may lead to more group conflict (Alesina et al, 2003). Finally, initial income levels are likely to affect growth, although the direction is widely debated. Convergence, whereby countries with higher initial income have lower subsequent growth, appears to be limited to subsets of countries, while across the world as a whole the dominant feature is divergence (Pritchett, 1997). The averages in Table 5.3 suggest that the Pacific Island countries face a number of potential disadvantages in addition to being very remote. Their population and land area are much smaller than for other states, so limited domestic markets may reduce their income (Redding and Venables, 2004). They are also less open to international trade, due either to protectionist policy choices or to their greater remoteness causing transport costs to be a larger source of ‘natural protection’. The Pacific countries also have greater language diversity. While less likely to have been colonised than other small countries their last colonial power grows more slowly than the colonial power for other small countries, which matters if incomes converge to those of the colonial power. On the other hand, between 1988 and 2003 the Pacific became ‘closer’ to the world economy due to the rising share of world GDP produced in surrounding countries. Also, none of the Pacific Island countries is landlocked, which is typically found to be an impediment to growth (Hausmann, 2001). The regression results that do not take into account spatial spillovers suggest that average growth rates of the Pacific Islands are significantly (at the 90% confidence level) lower than for other countries. Among the geographic factors, the negative effect of remoteness (-0.18) seems to be outweighed by the effect of the change in GDP-weighted distance between 1988 and 2003 (-0.42). Since remoteness fell for the Pacific, this suggests a ‘potential growth failure’ from not fully exploiting the opportunity provided by being close to where the greatest growth occurred in the global economy. Note though that the distance measures are based on potential market access. Suitable transport and trade systems need to be in place to turn this into actual market access; being near rapidly growing countries may be of little use if there is no cheap, easy, or reliable way to get to them. The average airfares showed that actual remoteness of the Pacific Islands exceeds what potential remoteness measures indicate and, likewise, actual remoteness may not have fallen by the extent that potential remoteness calculations suggest. Other points of interest from the regression results concern population (larger is better), land area (smaller, meaning greater density, is better), growth volatility (more is bad for average growth rates), and language diversity (more is bad for average growth). However, these conventional regression results do not take account of the significant positive interactions in average GDP growth rates between neighbouring countries. A variety of statistical tests Gibson (2007) reported

210 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? indicate that an appropriate neighbourhood for looking for these is for countries within 35 degrees of latitude or longitude of one another (equivalent to a distance of about 3,900 km at the equator, although these ‘neighbourhoods’ exist for all countries, not just equatorial ones). These tests suggest the appropriate model, where the growth rate of each country is affected by the spatially weighted average of growth rates of other nearby countries, even after controlling for observable factors that might be common for the countries, such as measures of remoteness or dummy variables for belonging to a particular region of the world. In other words, ordinary least squares (OLS) estimates of the growth model that are used in Table 5.3 and in all previous statistical studies of economic growth in the Pacific (and in small countries) are likely to be biased because they omit a relevant variable – the spatially weighted average growth rate of nearby countries. According to the preferred estimates in the last column of Table 5.3, each one point increase in the spatially weighted average growth rate of countries within a radius of 35 degrees raises the GDP growth rate of the country at location i by 0.54 points. This dependence on the growth rate of neighbouring countries is even after controlling for remoteness, other geographic constraints, population, openness and volatility, and initial economic, social, and political factors. Significantly, the use of the spatial model has a dramatic effect on the coefficient on the Pacific Islands dummy variable, which is the focus of attention. This coefficient now is only one-half as large as in the OLS estimates and is statistically insignificant. In contrast to the fragility of the result for the Pacific Islands dummy variable, several other growth determinants have the same effect in the spatial lag model as they had in the OLS estimates. These more robust variables include land area, openness to trade, and the volatility of the growth rate. The positive effect of population becomes even more apparent in the spatial lag results than it was with OLS. Other variables, including language diversity and the change in GDP-weighted distance between 1988 and 2003 become smaller but are still statistically significant. The pattern of spatial effects implies the need for a regional focus in any solutions to growth problems. In the preferred regression model, neighbouring countries’ average growth rate directly enters into the equation predicting the growth of a specific country. In some parts of the world, this spatial lag may reflect cross-border flows of goods and factors of production, but in Pacific intra- regional trade it is not especially important. Instead, Pacific Island countries’ shared use of key assets such as transport infrastructure could create a dependence on the growth of neighbours, since this growth affects the viability of transport links for all the countries sharing the service.6 Possible contagion of civil unrest across borders may also be important in the Pacific. Regardless of the exact source of spatial dependence in growth rates, it appears hard for a single country to have a strong growth performance if it is surrounded by countries that do not.

6 For a discussion of shared use of air and sea transport in the Pacific, see Vitasagavulu (2005).

Table 5.3: Ordinary least squares and spatial lag regression estimates of long-run growth equations

Means and standard Coefficients from deviations growth regressiona Ordinary Pacific Other least Spatial Islands countries squares spillover Average annual growth rate of gross 0.020 0.022 domestic product (GDP) per capita (US$), 1987–2003 (0.023) (0.034) Pacific Island country (= 1, -0.148 -0.066 otherwise = 0) (1.70)+ (0.80) GDP-weighted distance to all other 11.632 8.234 -0.109 -0.039 countries (000 km, 1987–1989) (0.793) (1.764) (1.28) (0.52) Change in GDP-weighted distance -141.767 67.054 -0.318 -0.133 (2003 compared with 1988) (8.519) (87.999) (4.04)** (1.74)+ Landlocked country (= 1, otherwise 0.000 0.201 -0.063 -0.049 = 0) (0.000) (0.402) (0.81) (0.72) Population (logarithm of 000s) 4.704 8.849 0.254 0.262 (1.688) (1.920) (1.80)+ (2.13)* Land area (logarithm of square 7.455 11.697 -0.274 -0.252 kilometres) (2.755) (2.507) (1.88)+ (2.00)* Openness to trade (exports plus 0.715 0.945 0.100 0.114 imports as share of GDP, 1987–2003) (0.181) (0.695) (1.56) (2.00)* Volatility (standard deviation of 0.114 0.149 -0.339 -0.324 annual GDP growth, 1988–2003) (0.071) (0.121) (4.93)** (4.76)** Language diversity (# languages 1.867 1.629 -0.161 -0.107 spoken by at least 20% of population) (0.743) (0.784) (2.71)** (2.07)* Never a colony (= 1, 0.067 0.170 -0.067 -0.071 otherwise = 0) (0.258) (0.377) (0.91) (1.03) Coloniser growth rate 0.031 0.034 0.108 0.071 (0.012) (0.012) (1.34) (0.96) Initial income level ((log) GDP per 7.385 7.309 0.108 -0.008 capita in 1988) (1.104) (1.451) (1.49) (0.13) Spatial weighted average of n/a 0.542 neighbours’ growth rates (5.80)** Constant 0.000 -0.006 (0.00) (0.11) R-squared 0.394 0.492 Log-likelihood -201.811 -190.404

Notes: n/a = not applicable; + = significant at 10%; * = significant at 5%; ** = significant at 1%. a Standardised (beta) coefficients are reported, showing the effect of a one standard deviation increase in the explanatory variable on the standard deviation of average per capita GDP growth over 1987–2003. Robust t-statistics in ( ) for ordinary least squares and robust z-statistics for the spatial model. Source: World Bank and Asian Development Bank for data on GDP, growth rates, population, openness, and volatility. Data on land area, whether landlocked, language diversity and colonisers are from a database constructed by the Centre d’Études Prospectives et d’Informations Internationales.

212 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Of course, another factor that may be shared across neighbouring countries is governance failures, whether due to demonstration effects or to correlated causes. To investigate whether governance in the Pacific differs from the rest of the world, and what impact such differences may have on economic growth, we adapt the growth model used in Table 5.5. In particular, we use indicators for six components of governance from a major World Bank research project that creates consistent indicators from the perceptions of enterprises, citizens and expert respondents in most countries around the world (Kaufmann et al, 2006). These components are: voice and accountability; political stability; government effectiveness; regulatory quality; rule of law; and control of corruption.7 According to the results in Table 5.4, Pacific Island countries have lower levels of governance for three of the six indicators (and in two of these cases the coefficient is only weakly significant). Consequently, only for the ‘control of corruption’ would there be firm grounds for concluding that governance was worse in the Pacific than elsewhere. The results in second and third numerical rows of Table 5.4 show that the addition to each of the governance indicators to a cross-country growth equation has very little effect in explaining any atypical growth performance in the Pacific. Specifically, the difference in average growth rates between the Pacific Island countries and the rest of the world (after taking account of geography, population and land area, trade and volatility, and initial economic, political, and social conditions) is hardly altered as governance indicators are also included in the growth equation. For example, according to the results in the column headed ‘Government effectiveness’ (this reproduces the result in the second-to-last column of Table 5.3) the average growth rate of the 14 Pacific Island economies included in that particular equation is -0.14 standard deviations below that of other countries, after controlling for various features of all of the countries in the sample but not controlling for differences in government effectiveness. Adding government effectiveness to the set of controlling variables reduces the unexplained difference in growth rates only slightly, to -0.115. Moreover, none of these differences in average growth rates is statistically significant. The implication that we take from the evidence in Table 5.4 is that claims of governance failure in the Pacific may be overstated. There is clearly a failure to control corruption but the evidence is less clear-cut for other components of governance. Moreover, the impact of any governance deficiencies is difficult to discern, at least in terms of average economic growth rates. Thus, these results suggest that governance failure may not be as costly to Pacific growth as sometimes suggested. The contrary evidence from Duncan (2005) uses Mauritius as a comparator to Fiji and Botswana as a comparator to Papua New Guinea. This use of comparators recognises that in order to fail, one needs to have some standard of success. Hence, what is required to assess how well the Pacific Island economies are performing and what countries like New Zealand can do to help

7 While these are available for 1996–2005, although with many gaps for individual countries, the analysis uses averages over 1996–2003 since many of the other explanatory variables are dated from about 2003.

Table 5.4: Deviation of average level of governance and growth in Pacific Island countries from overall global average, controlling for geography, population and land area, trade and volatility and initial conditions

Voice and account- Political Government Regulatory Rule of Control of ability stability effectiveness quality law corruption Coefficient on Pacific Island countries in equation for 0.127 -0.042 -0.104 -0.121 -0.077 -0.174 governance (1.51) (0.34) (1.70)+ (1.88)+ (0.97) (2.42)* Coefficient on Pacific Island countries in growth equation without -0.148 -0.131 -0.139 -0.122 -0.122 -0.122 governance indicator (1.63) (0.97) (1.54) (1.31) (1.31) (1.31) Coefficient on Pacific Island countries in growth equation with governance -0.180 -0.117 -0.115 -0.086 -0.106 -0.104 indicator (1.93)+ (0.90) (1.27) (0.91) (1.15) (1.08) Sample size 172 165 173 171 171 171 Pacific countries COK, PYF COK, FSM, PYF NRU, PLW, NRU, PLW, NRU, PLW, missing from KIR, MHL, PYF PYF PYF sample NRU, PYF, TON, TUV

Note: Results are ordinary least squares estimates for models that also include the control variables listed in Table 5.3. The coefficients for the growth rate equations are for models explaining average per capita gross domestic product growth over 1987– 2003, for samples of about 170 countries. In round brackets are t-statistics from robust standard errors; + = significant at 10% level, * = significant at 5% level, ** = significant at 1% level. The Pacific Island economies omitted from each sample (due to lack of governance indicators are): COK = Cook Islands; FSM = Federated States of Micronesia; KIR = Kiribati; MHL = Marshall Islands; NRU = Nauru; PYF = French Polynesia; PLW = Palau; TON = Tonga; TUV = Tuvalu.

However, using the key characteristics of what makes a ‘Pacific Island’ economy (remoteness, small population, and less political independence than in many regions) identifies very few nations that could realistically be considered comparable. This makes the assessment task of how well (or badly) the Pacific Island countries are doing even more difficult. The approach we use is based on a method increasingly used in labour and public economics called propensity score matching. The basic idea is to use a probabilistic model showing the characteristics most closely associated with the propensity to be a Pacific Island country. For example, over 80% of the variation in an indicator variable showing which of the sample of 174 countries are Pacific Island countries comes from a model whose explanatory variables are the GDP-weighted average distance to all other countries in 2003, the logarithm of each country’s population, and a dummy variable for whether the country is politically an independent state. Pacific Island

214 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? countries are more remote, have smaller populations, and are less likely to be politically independent than are other countries. Such a model shows that most countries outside of the Pacific Islands have a very small propensity to be a Pacific Island country. What does this seemingly self-evident statement mean? Intuitively, it says that in terms of remoteness, population size, and political independence very few countries outside of the Pacific are like the Pacific Island countries. Who amongst the non-Pacific countries are the good comparators? The only propensity scores from the non- Pacific Island countries that exceed 0.25 are for New Zealand (0.69), Timor Leste (0.54), and the Seychelles (0.53). In other words, these are the countries most like the Pacific Islands in terms of remoteness, population, and political autonomy. So researchers searching for counterfactuals of how the Pacific Island countries might have grown if they had adopted different policies would do better to look at the growth experience of these three comparators than at some other countries chosen in an ad hoc manner. Of the two comparators used by Duncan (2005) Mauritius has the fourth highest propensity score (0.23) amongst the non–Pacific Island countries, while Botswana has only the 11th highest score (0.07). Thus, they can be considered reasonable, but not necessarily the best, choices of comparators. Does the use of ad hoc methods of selecting comparators matter? The average annual growth rate of GDP per capita (over 1987–2003) of Mauritius and Botswana was 4.0% per year. In contrast, if all 11 non–Pacific Island countries that have propensity scores at least as high as Botswana are used, the average growth rates are only 2.9%. Just restricting attention to the top three comparators of New Zealand, Timor Leste, and the Seychelles, the average growth rate is 3.0%. So a more systematic method of selecting comparators would lead to a lower counterfactual growth rate for evaluating how badly the Pacific Island countries have done compared with choosing Mauritius and Botswana. In fact, once the 15 Pacific Island economies considered are compared with a weighted (by the propensity score)8 average of either their closest comparators or all 159 comparator countries with data available, they are found to have only 0.4 percentage point lower annual growth rate, which is statistically insignificant. Thus, there is no firm evidence that the Pacific Island countries grew more slowly over 1987–2003 than did an appropriate group of comparator countries.

Pacific systems of wealth Pacific economies are increasingly integrated within the global economy through international trade and interactions with aid donors, and international banks and monetary policies; yet, they also have a substantial amount of economic activity that does not fit easily into standard frameworks. To what extent do apparent economic weaknesses of Pacific economies reflect our limited understandings of the ways in which they operate? In this section, we identify and critique the

8 In other words, the counterfactual average growth rate of the control group of countries places most weight on the non- Pacific Island countries with the highest propensity scores.

© Institute of Policy Studies, Victoria University of Wellington, 2008 215 Pacific Interactions disjuncture between current international statistical practices that focus on monetary transactions, and dismiss non-monetised production and distribution of foods and other substances as ‘subsistence’ or ‘informal’ practices too difficult to count. From Pacific perspectives, current data collection is limited by incomplete definitions and perspectives. This results in faulty and biased analyses and missed opportunities for support and expansion.

Pacific and Western differences in wealth systems Pacific systems of generating and storing wealth differ from those of the West in two critical ways. Food production is a foundation on which Pacific wealth systems are built, whereas European stores of value (eg, gold that in the past was the standard on which money was based) are separately produced and based on other types of commodities. As family food productive activities are of little import in European systems today, they are largely ignored or dismissed as part of the ‘informal economy’. In contrast, the production of food remains the primary support for many Pacific peoples, and is the sector through which excess production can be transformed both into Pacific wealth items (hard wealth and soft textile valuables) and into cash through market sales. Undercounting this highly productive sector of Pacific economies leads to a miscalculation of its productive power and its ability to interrelate with monetary economies of the islands and New Zealand and Australia, which host major populations of Pacific people. This is partly due to the radical differences between Pacific crops and better- known world crops such as rice and grains that are regularly harvested and stored. Pacific agriculture focuses upon agro forestry (Clarke and Thaman, 1993), and root crops such as taro and sweet potato (now the fifth most important crop in the developing world). In parts of the Pacific, root and tree crops require little labour and may be planted and harvested throughout the year as needed, indeed the hard taro (Cyrtosperma) may be left in the land for years until needed. Pacific multi- cropping systems of planting, management, and harvesting differ from Western single-cropping chemical fertilisation systems. The second related difference between European and Pacific wealth systems is that to a large extent production of food stuffs (taro, breadfruit, and fish) and of wealth items (fine mats, tapa, adze blades, and bead monies), and their distribution systems are gendered. Generally, women grow root crops and men grow tree crops. Throughout the Pacific, production and exchanges are organised through relationships between sisters and brothers and their respective spouses. Under the early colonial rule of Pacific Islands, only the men were subjected to head taxes and only men were taken as overseas labourers. Europeans observed Pacific women producing mats and tapa but did not recognise their economic value. This lack of attention, at times, permitted women greater flexibility. Much of women’s production is still included in the undercounted and under-analysed household sector. Only in the last decade have United Nations and other major international donors (the Asian Development Bank and World Bank) seriously recognised the economic (and related political) power of Pacific women and

216 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? supported women-oriented projects through Women in Development and other international projects.

Western perspectives: Assessing subsistence For decades academics have noted that local food production and the productive work of women, in the West as well as the islands (Waring, 1999), is undercounted. To what extent is the food production of men and women in Pacific Islands still undercounted and under-represented in national accounts? We start with a discussion of fisheries before turning to agriculture.

Fisheries: A region-wide Asian Development Board assessment In 2001, the Asian Development Bank recognised that the importance of fisheries was not fully appreciated by donors or Pacific nations and that existing data was based on inaccurate estimates, and organised a comparative assessment of Pacific fisheries for the Forum Fisheries Agency and Secretariat of the Pacific Community in conjunction with the World Bank. Gillett and Lightfoot (2001, p 12) assessed the contribution of fisheries to the GDP of Pacific Islands Forum nations, noting: The research reveals that the full value of fisheries is likely to have eluded statisticians, and therefore fisheries authorities, government decision makers, and funding agencies. But the value of fisheries has never escaped the fisher, fish trader, and fish processor. The difference in appreciation, between public and private individuals, must raise the question of whether fisheries are receiving adequate attention from the public sector – including development, extension and training, and sufficient investment. They recognised five common difficulties in calculating these contributions: • the lack of coordination between fisheries agencies and statistical agencies • the lack of data on subsistence fisheries and difficulties in isolating fishing from other subsistence activities • identifying the contribution of fishing processing to exports • official export figures undervaluing exported commodities • data on small-scale fisheries being scarce, as was technical assistance for its analysis (excerpted from Gillett and Lightfoot, 2001, p 21). After assessing official fisheries statistics, Gillett and Lightfoot then re- estimated fisheries contributions for the late 1990s for 14 Pacific nations. They compiled data, including the value of subsistence fishing based on ‘farm gate’ methods (market cost less cost of getting that product to market), coastal commercial fisheries, and offshore local and offshore foreign fisheries. To the degree to which they were available and segregated fishing from agricultural subsistence activities, Household Income and Expenditure Survey data was used. For the Pacific as a whole, the re-estimations “indicated a higher contribution of fishing to GDPs than reported by national statistics”. Rather than contributing an average of 5.4% of GDP across all nations, the re-estimation of the contribution of

© Institute of Policy Studies, Victoria University of Wellington, 2008 217 Pacific Interactions fisheries to GDP was 7.0% (Gillett and Lightfoot, 2001, p 20). The effects of the re-estimation were more striking at the level of individual nation-states. While original reported data and the re-estimations were similar for a few countries (Tuvalu, Tonga, the Cook Islands), for 6 of the 12 nations re-estimations doubled the figures for fisheries production, mainly due to the inclusion of subsistence fishing. Re-estimations for Palau more than quadrupled estimates of fisheries production due to the methods used (the subsistence estimates remained constant). As expected, the Pacific consumption of fish far exceeds the world per capita consumption of 13.0 kg (Westlund, 1995, cited in Gillett and Lightfoot, 2001, p 79). The important observation is that small islands have the highest kilogramme per year per capita (kg/yr/c) fish consumption rates. Kiribati tops the list of annual per capita consumption, with 72.0–207.0 kg/yr/c, followed by Tuvalu (85.0–146.0 kg/yr/c) and Palau (84.0–135.0 kg/yr/c). The large densely populated island nations of Vanuatu (15.9–25.7 kg/yr/c) and Papua New Guinea (18.2–24.9 kg/yr/c) have the lowest annual per capita consumption. Unexpectedly Tonga’s estimated per capita fish consumption rate was also low (25.2– 30.0 kg/yr/c), even lower than that of Solomon Islands (32.2–32.7 kg/yr/c), Marshall Islands (38.9–59.0 kg/yr/c) and Nauru (46.7 kg/yr/c). These statistics demonstrate the range of per capita consumption among islands, and begin to suggest some of the other factors affecting fishing consumption including urbanisation and reduced access to or pollution of nearby resources as well as access to low cost (and often low nutrition) imported foods.

Agriculture: Data from Solomon Islands, Papua New Guinea, and Samoa The percentage of the population that relies primarily on local food production varies throughout the Pacific, with Solomon Islands, Vanuatu, and Papua New Guinea representing the high end of the spectrum. This percentage also changes depending on circumstances. Even residents of small island nations such as Palau for which agriculture and fishing may normally contribute only 38% of GDP (European Community, 2007, p 6) may, like Solomon Islanders, return to farming and fishing to feed their families during periods of political duress. Throughout the Pacific amounts and types of food routinely harvested for consumption and ceremonial events are not routinely documented, nor are the complicated strategies farmers use to allocate the harvest among the household, extended family, and village/church event destinations, and small-market sales. Pacific people attempt to use both local and export food production to access international currencies, with variable success. To our knowledge no comparative regional assessment has been completed of the contributions of agricultural sectors to Pacific economies. Therefore, in this section we include only a few case estimates of agricultural production that demonstrate the discrepancies between official statistics and analysts’ re- estimations. To a large extent, these undercounts are the artefacts of poor access to data. For instance, the estimates of agricultural production in the largest Pacific country, Papua New Guinea, are extrapolations from a 1961/62 Survey of

218 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Indigenous Agriculture. Since up to 84% of the population may derive their livelihoods from a combination of subsistence agriculture and small-scale cash- cropping and marketing, such undercounts are critical barriers to understanding the local economies and areas in which they may be managed and strengthened. If such a large percentage of the people rely on local food production, then one might expect the Regional Assistance Mission to the Solomon Islands and other programmes might be designed to support and augment the agricultural sector. Yet analysts have found that (Bourke et al, 2006, p xi): Although the informal sector is the backbone of the Solomon Islands economy, only 1% of the donor funded development budget in 2004 was targeted towards the agricultural sector, and the current level of donor commitments for future funding is not sufficient to effectively address the constraints identified to improve rural livelihoods. Bourke et al’s (2006) study is based on a major Australian Government and AusAID project. Bourke et al (2006, p xi) are careful to separate “subsistence agriculture and small-scale cash cropping and marketing of fresh produce”. However, it is difficult even for such a careful study to avoid the common practice through which the term subsistence is used beyond its intended meaning of supporting only the household, with no surplus for sale. In volume two of the same study, Subsistence Production, Livestock and Social Analysis, Jansen et al (2006, pp 1–2) state: Subsistence food production has been the traditional way of life in [the Solomon Islands], and is still critical for the livelihood and food security of rural populations … however, for the purposes of this chapter, subsistence food production refers to producing food for local consumption. This includes consumption by the producers themselves (almost all are family based), as well as sharing, exchanging and selling food in the local area…. Subsistence food production makes an important contribution to the rural cash and noncash (or subsistence) economies …. Local markets are found throughout the country; and the sale of food products in local markets is a critical source of regular income for most households. Local markets are particularly important because women are more likely to control this income source and can use it for meeting basic family needs. [Second emphasis added.] This slippage in the definitions of subsistence and the conflation of subsistence production and local market sales is endemic and contributes to analytical confusions. It also means that local market sales, which could be easily measured, remain largely unmeasured because they are put in the ‘too hard’ basket along with subsistence production. In Papua New Guinea, Gibson (2001, pp 19–20) used nation-wide household survey data, and found that “the aggregate value of food production was approximately K1.3 billion … and … the national accounts may understate the value of household agricultural production by almost one-half”. In further research on the sweet potato in Papua New Guinea, Bourke and Vlassak (2004) documented that in 2000 sweet potato provided over 60% of the energy diet in

Papua New Guinea, being the dominant staple for 2.8 million rural people in Papua New Guinea (65% of the rural population) and a sub-dominant staple for another 17% (ie, a total of 3.5 million people or 82% of the rural population). Moreover, the production and demand for sweet potato are not limited to rural households. Bourke and Hide in a recent overview of food security for Papua New Guinea (Bourke et al, 2001) have suggested that the underestimation of Papua New Guinea’s agricultural production (including livestock such as pigs) is much greater, in the magnitude of eight times (not two times) that reported in the official statistics. For the Solomon Islands, the value of the 2004 consumption of staple agricultural crops (calculated by food energy requirements) was estimated at $325 million per year. A comparison with the value of the major copra and cocoa exports of $73 million in 2003 and $70 million in 2004 demonstrates the value of local food production to the Solomons’ economy (Bourke et al, 2006). This does not count crops such as betel nut, and other commodities that are a significant source of cash income. Agriculture proved to be a highly resilient system that was able to meet the basic food needs of the rural two-thirds of Solomon Islanders during the period of ethnic tension from 1998 to 2003. O’Meara (1990) provided an analysis of Samoan agriculture in the early 1980s. He recognised the importance of the social incentives that enter into villager choices, but also documented the poor return received by agriculturalists at that time in comparison with higher wage positions. He offered a full census of village economic activities and values, noting, however, the difficulty in attempting to value subsistence products and fine mats. A more recent study documented Yapese food production and its intersections with global economies (Egan et al, 2006). Urban food production is little documented. In 1990–1991 urban production in the Morata and Waigani suburbs of Port Moresby contributed 49% of household consumption compared with 9% of fresh food purchased for other urban households (Levett, 1992, p 4). This is an area for further development and research.

Subsistence and beyond The studies of Pacific food systems and their complex transactions across the range of production for subsistence, small-scale income-generating activities, prestige economies supporting local and global political and religious leaders, and local and global marketing are predominantly found in the anthropological literature. These studies are ill connected with economic micro- and macro- analyses. Yet Pacific wealth systems remain an important mechanism to mediate relationships between indigenous and introduced money and products. Weiner (1980) and Akin and Robbins’ (1999) edited volume Money and Modernity: State and local currencies in Melanesia demonstrate the processes through which particular Melanesian societies today attempt to regulate intersections between islander and international monies. A brief overview of the ways these systems operate in the region is provided. Local and regional commodity and wealth exchange systems were long established in the Pacific (ie, Papua, Vitiaz Straits, Massim, New Britain, Caroline

220 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Islands, and Samoa/Tonga/Fiji) to distribute scarce commodities (obsidian, pottery, shell and stone tools, dyes) and foods. At the simplest level of analysis, indigenous economies were and are based on the competitive production of local foods and ritualised consumables (betel, kava) upon which a system of valuables and currencies was constructed. Throughout the region cloth valuables (mats, tapa, fine banana fibre weavings and bundles, and fibre clothing) were nearly exclusively made by women. Hard valuables in contrast were generally made by and associated with males – fossilised clam shell fashioned into adzes and ornaments, greenstone, obsidian tools, and shell, stone, and glasseous bead valuables. The valuables and currencies stored value created through food production. As Pacific peoples continue to settle distant shores and encounter new foods, valuables and currencies, they incorporate new materials and currencies into exchange systems that they adapt to new conditions. They include the more portable new currencies to capture their very evident powers.

Gendered production systems The ways that the brother–sister cross-sibling dyad organises production and transfers through their spouses and affines are similar throughout the Pacific, relating to deep cultural roots of these predominantly Austronesian peoples. While the Melanesian islands also include non-Austronesian peoples, similar systems operate in the Papua New Guinea Highlands, connecting male elaborations of hard wealth (including shell traded from the coast) through pigs fed by their wives, and exchanged in regional competitive exchanges (Strathern and Stewart, 1999). Throughout the Pacific, societies elaborate men and women’s productive capacities to competitively transform foodstuffs into wealth, and to link with and mediate introduced foodstuffs and currencies. Malinowski’s (1922) classic example of the long-established Massim regional Kula trade documented its early 20th system of a double circle of delayed exchanges of two shell valuables (necklaces and shell bracelets) between partners on participating islands. Weiner (1976) studied the system as it operated in the late 20th century, analysing the engendering of the wealth system and its integration of introduced money. Simply stated, in the Trobriand system brothers competitively produce and provide yams to their married sisters and their husbands. The stored yams are used as food, as seed, and as the foundation of the male hard wealth items (greenstone blades and the shell valuables). As high chiefs are polygamous and commoners monogamous, chiefs in particular accumulate yam wealth that can be transformed into hard wealth and Kula valuables. In turn, the sisters provide their brothers with the soft wealth (banana fibre bundles and skirts) required for ceremonies associated with major life events and especially mortuary ceremonies that ensure the reproduction of the clan and lineage. In today’s market economy supported by tourism and wood carving, the sister’s husband expends cash to support his wife’s collection of banana fibre bundles (relative to the number of yams his brother-in-law provided). This is an example of the male production of starch foods, here root crops produced and elaborated

© Institute of Policy Studies, Victoria University of Wellington, 2008 221 Pacific Interactions by men, and soft textile wealth produced and elaborated by women, connected through the brother–sister cross-sibling dyad and their respective spouses that serves to transfer and competitively develop food production and wealth. Weiner (1980) argued that Malinowski ignored the role of textile wealth and in particular the importance of banana fibre bundles. She demonstrated how in the late 20th century Trobriand islanders increased their use of readily produced banana fibre bundles, arguing that in the process they protected the greenstone valuables from inflation. They thus addressed one of the classic problems of integrating local and introduced currencies - that of avoiding inflation and destruction of the local currency and economy (Bohannon, 1955; Akin and Robbins, 1999). In Polynesian societies one can see the same key relationships in which the brother–sister dyad is foundational to societal power, and the ways that transactions of food and valuables (in this case especially fine mats and tapa) connect lineages linked through marriage. Indeed the regional component of these links is of long standing, as documented by the circulation of spouses and valuables linking Samoa, Tonga and Fiji in particular (Kaeppler, 1978). The Micronesian examples are also widely regional: studies document trade systems that connect the Carolinian atolls with the high islands of Yap through a hierarchical chiefly system linked by regionally dispersed matrilines to the atolls to the east. The Sawei system linked the small atolls of Eastern Carolines through Ulithi to the high island of Yap, bringing tribute of woven textiles and other items. Although the logic of this Sawei exchange established the supremacy of Yap, it also provided atoll dwellers access to vital high island resources (Lingenfelter, 1975; D’Arcy, 2001). Analysts note that the value of the foods and high islands products given to the atolls in response was greater than that of the tribute received (Alkire, 1989 [1965]). Sawei relationships are still foundational to contemporary practices in this state of the Federated States of Micronesia, although disaster relief is now primarily transported on ships and Air Micronesia planes.

Why consider island production and wealth systems? Why should we study such complicated traditional systems in today’s world? Local variations are constantly developing throughout the Pacific as men and women continue to transfers foods and goods within kin groups at home and abroad, today including class- and work-mates in the transactions. Former prestige offerings (betel nut, kava) and customary exchanges are now partially commoditised in order to earn domestic cash for domestic purposes or to earn export income to generate capital for investments. Pacific exchange systems today also often serve as a point of articulation and negotiation between indigenous/local commodities and value systems and international and local markets.

From prestige item to source of capital: Betel nut The fruit of the betel palm Areca catechu is widely produced, gifted, and traded in the western Pacific. Today it is sold by small entrepreneurs in local markets,

222 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? making it one of the key local products of the ‘informal’ economy that can provide people with income to meet new cash requirements for items such as school fees, and trade store goods such as imported foods. Local market sales are difficult to quantify. The 1995 survey of betel nut chewing in Palau (Ysaol et al, 1996) estimated that the annual market value of the sales of the betel components (betel nut, and minor components leaf and lime less costs of imported cigarettes) exceeded US$9.2 million in Palau, a nation of about 20,000.9 Local market income is jealously guarded, and the incursion of Chinese street merchants into this lucrative small trade in betel nut reportedly contributed to the recent unrest in the Solomon Islands. Increasingly, the value of the export sales of betel nut is being reported in national statistics. In Micronesia betel use is extending into islands where it was not previously chewed, and in 2006 represented 35% of the Federated States of Micronesia’s total exports (Division of Economic Planning and Statistics, FSM, 2006, p 44). These regional exports were sent nearly entirely to neighbouring Guam, the Commonwealth of the Mariana Islands, and the Marshall Islands, but also to Hawai’i. Strong exports from Yap (73.5% of their total exports) have spurred new export production from Pohnpei and Kosrae as well. Market sales of betel nut within Papua New Guinea market are very lucrative. Mekeo entrepreneurs in Papua New Guinea have benefited from betel trading to Port Moresby in recent decades. There are reports of lucrative betel markets operated by grassroots entrepreneurs in other parts of Papua New Guinea, notably in Madang and Boikulu, with a reported multi-million kina betel nut trade from the Ramu region. Spin off transportation businesses by motorised canoe and bus hires also serve the traders. It is expected that the national estimates for betel nut of PGK 78 million plus or minus 18 million, may be low for this nation of over 6 million, especially in comparison with the estimates of US$9 million for Palau alone with a chewing population of less than 15,000. Mekeo used income earned from betel sales to support the building of permanent houses and a church and the purchase of small potentially capital- producing items (vehicles, generators, and freezers) as well as direct consumption items.10 Most notably, “Three voluntary credit societies had been established … replacing old family houses … building a new church” (Mosko, 1999, pp 56–67). However, general control of the new money affected political relationships, and chiefs and sorcerers attempted to reassert their power by taxing the construction of houses of European materials and design on the grounds that they approximated fine chiefly houses (Mosko, 2002, p 100). Economic and political powers are closely entwined in Pacific societies.

9 The 1995 survey of betel nut chewing was based on a sample of 7.9% of the total Palauans aged over 5 years (then resident). Ysaol et al (1996) found that 76% of respondents used betel nut with a slightly larger percentage of female than male users (80% compared with 72%), with a mean number of 15.9 chews per day. 10 Mosko (1999, pp 53–56) did a census of one village of 45 households (350 residents) over a three-month period. Extrapolated to an annual income, for this one village the annual value would have been about 800,000 kina (NZ$1.5 million at the time).

Adapting existing exchange systems To a certain extent the success of Pacific agriculturalists is closely linked to the success of their nation’s monetary economy. As the Palauan economy began to grow and wages increased in the growing market sector, the remaining farmers and fishers could command high prices for betel and local fresh produce demanded for urban households and special events, and the growing tourist market. Palauan entrepreneurs then began to import Asian workers to farm and fish to supply these products. Palauans have adapted existing exchange systems to support the construction of permanent houses. In Palau Planning and Statistics conducted an informal sector census in 1994, covering 44 ocheraol exchange functions (Ysaol et al, 1995). Oriented around family lineage and clan relationships, the ocheraol is an example of the group orientation of Palauan society and its collective values. In the past the event could be used to raise money for a wide range of need; today it is mainly used to cover the cost of house building. A constantly changing custom that adapts to new economic as well as political and social forces, the exchange functions today attract participation and funds from class- and work-mates and from Palauans overseas. In the 1994 census ochearol were held only in the capital city of Koror and were regulated to avoid Christmas and the beginning of schools. The money they earned ranged from US$3,000 to US$100,000 or more, with an average of just under US$28,641. In the past only indigenous bead money was used; today the United States dollar is the national currency and only 12 pieces of Palauan bead money were involved. The financial impact of ocheraol function in 1994 alone was considerable. “These 44 ocheraol functions were associated with construction of approximately 56,600 square feet of new building space valued at close to $2.0 million. New shelter was provided for 237 people” (Ysaol et al, 1995, p 16). The ocheraol provided 69% of the costs of these buildings. The 44 events each averaged 176 participants, for a total of 7,742 people – about half the country’s adult population. This adapted exchange provides important capital, as it is very difficult to obtain bank loans for housing construction since the land tenure system recognises clan/lineage as well as individual ownership, and much of the land in Koror is controlled through a system of leasing through the state government. It makes “a major contribution to the availability of public housing and the economic viability of the private sector small scale construction and food preparation industries” (Ysaol et al, 1995, p 19).

Mediating between ceremonial and market economies An attempt by Tongan women in the 1990s to increase their production of ceremonial textile koloa to earn cash money and Western goods became a classic case study of the intersections between development economics, women’s small- scale business endeavours, and the market and ceremonial economies in Tonga and New Zealand. In 1995 aid donors and a lending agency provided funds to development banks in Tonga to finance micro-credit programmes targeting grassroots rural enterprises designed to foster economic development in rural

224 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? areas, especially to benefit women. Three of the women’s projects accepted by the bank were designed to support local women in the production of tapa to earn money for the families. The funders envisioned that the women would produce tapa for sale through local markets and earn a profit beyond the required repayment of their loan. The women did borrow the money (using existing koloa as collateral), produced more koloa, and repaid the loans. However, they did not sell this tapa at local handicraft markets or only exchange it for other ceremonial textile koloa to earn social capital. Instead, following existing practices within Tonga the producer sold textile koloa at high prices to urban women who did not produce their own textile koloa, or exchanged it for cash and/or Western goods with informal exchange partners. The producer also sold textile koloa overseas through market or retail outlets, or exchanged it for cash and/or Western goods with informal exchange partners. She also directed some of the koloa along ceremonial Katoanga Exchange paths, with an important difference that partners living overseas in return sent the producer cash and/or Western goods rather than the textile koloa that was given in return by a partner in Tonga (see Tongan Ceremonial Economy diagram (Figure 5.3) Access to overseas Western goods through exchange for textile koloa was highly beneficial, as confirmed by family and community willingness to contribute to repay the original loans. The development project was considered a great success by the Tongans in that they had been supported to produce more textile koloa that enabled them to earn cash and Western goods for their families as well as social capital (through Katoanga Exchange). However, the development bank considered the project a failure, because the women refused to sell all of the textiles so produced. The bank did not consider that repaying the loans through fundraising could be considered development, though there had been a net increase in the women’s ‘wealth’ if one considered the integrated ceremonial and monetary economy operating in the islands. Figure 5.3 provides a comprehensive view of the Tongan Ceremonial Economy, from the perspective of the tapa-maker, demonstrating the complex transactions and relationships linking her to Tongans in Tonga and overseas, her interactions with non-Tongans in Tonga and abroad, and with local and overseas market/retail outlets. The transfers began with local transfers of raw materials, cash money and textile goods within the extended families and communities of the producer and her husband, then expanded to include Western goods within the prestation and exchange spheres within Tonga, finally extending to prestations and exchange with overseas migrants. In the process, the producer selectively directs her koloa to informal exchange partners, to mercantile exchange with Tongans or through market/retail outlets overseas, or to Katoanga Exchange partners. This diagram demonstrates the importance of overseas migrants to the tapa producer as a source of Western goods, which she can then use and redirect locally. The sum of the value of these transactions exceeds the value of a simple sale of the koloa once it is produced.

Figure 5.3: Tongan ceremonial economy

Source: Horan (1997).

The increasing value of koloa relates to its increased demand from full-time wage-earning women in Tonga and overseas who cannot make koloa for themselves. One of the factors permitting this to occur is the long-distance exchange and sale of raw materials and completed koloa among women in rural and urban sites in Tonga and Auckland. Horan (1997, 2002) and Schoeffel (1994) argued that this ‘capital’ investment in the local wealth system, rather than international currencies, also made sense within Western economic accounting, but this was not accepted at that time. Although this is a single example, it illustrates the way in which economic agents in the Pacific manage to straddle two economic systems. Within the next decade international donors began to understand the ways islanders could use indigenous wealth to increase family income, and began to support such small- scale projects. In Samoa, the Women in Business Foundation secured overseas funding to support small-scale production of items targeted to the market of Samoans living abroad. These products include coconut oil, soap, and fine mats (textile indigenous wealth) needed for cultural occasions overseas. There are direct connections between the operation of such contemporary Pacific ceremonial economies and the remittances discussed in the next section. McKenzie (2007) concluded that between 25% and 40% of remittance transactions were missing in official statistics focused solely on monetary transfers.11 Migrants also send consumer durable goods, which can be important to small nations with limited supplies of highly priced goods.12 In addition, most remittance data does not include reverse flows of handicrafts and food from Tongans to overseas migrants, some of which was shown in the Tongan Ceremony Exchange diagram. McKenzie’s Tongan data suggests these reverse remittances equalled 43% of the value of remittances sent by New Zealand migrants. This indicates how substantial are the reversal flows of the type shown in Figure 5.3 that captures the complexity of these transfers. One problem of most discussions of remittances, such as that in the following section (Table 5.5), is that they focus on the one-way direction of cash transfers only. They do not include reverse flows, or the important transfers of goods going to the islands. The issue of reducing remittance costs is an example of the sorts of transaction costs (transportation, communication) that can be reduced to better aid Pacific agents maximise their responses to opportunities in local and global economies.

11 While recorded remittances are an important contribution to the GDP of many Pacific nations, most do not capture all the cash remitted, since some is carried through informal channels by friend or family travellers. Based on data collected from migrants who identified remittance of both cash and goods, cash averaged 75% of total remittances sent, and 63% or all remittances received in Tonga (not only from New Zealand). 12 In September 2005, McKenzie collected prices of 14 durable goods in New Zealand and Tonga, and found that the price in Tonga averaged 1.7 times the price in New Zealand.

Implications and possible interventions The purpose of this chapter is to draw attention to both conceptual and data quality issues of concern in assessing our understandings of Pacific economies and suggest possible interventions. As the effects differ at community and national levels, we have made no attempt to substitute existing data with re- estimations. In this final section, we briefly consider some of the implications of the critiques we have raised. In the Annex to this chapter, we suggest a few approaches that could assist in the collection of better data of comparative value for the region, and the benefits and costs of not developing better economic data on which to base national programmes. The need for regional approaches is recognised by regional organisations (ie, Pacific Islands Forum and South Pacific Community and major international donors (ie, the Asian Development Bank, World Bank, and European Community). Much of the work needs to be done at the regional level, designed and implemented by Pacific and international partners.

Safety net implications of remittances and inter- household transfers in the Pacific The complex wealth systems discussed above involve substantial exchange and redistributive activity. Economists are interested in such activity for at least three reasons. First, to the extent that redistribution reduces longer-term inequalities in living standards it is of interest because most evaluations of social welfare consider both the level or growth in income (or consumption) and the distribution of that income or consumption across the population. Moreover, to the extent that there is an existing informal safety net, the introduction of any public transfers programme (which may be partly funded by aid, so should be of interest to the donor community) may be less effective than expected because public transfers could crowd out the existing informal safety net (Dercon, 2002). Second, remittances and other transfers may act like insurance in the sense that they provide assistance for those people who find themselves in temporary distress (Cox and Jimenez, 1990). In other island regions, remittances have been shown to provide partial insurance for the damage caused by natural disasters (Clarke and Wallsten, 2003). Similarly, Connell and Brown (2005) suggest that remittances in the Pacific may play a key role in reducing national and household vulnerability. For example, in the two months after Cyclone Ofa, which struck Samoa in 1989, remittances were 70% higher than in the same months of the previous year as immigrants who had not remitted for some time sent money to friends and relatives (Macpherson, 1994). The insurance role of transfers is also illustrated during the 1997 food shortages in Papua New Guinea that were caused by frost and drought during the El Nino conditions. Consumption of imported rice soared by 58,000 tonnes (40% above the previous level) during this event, and much of this reflected private transfers rather than supply by aid donors, who started food distribution after the peak of the crisis (Whitecross and Franklin, 2001). Rural people with relatives in urban areas received money that was used to help pay for imported food during the crisis but people in poor and isolated places

228 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? with no urban kin were unable to compensate for the loss of their traditional foods (Allen et al, 2005). Third, a large literature argues that remittances to Pacific Island countries have undesirable economic effects. In particular, remittances are believed to contribute to over-valued real exchange rates by pushing up domestic wages and production costs, which reduces international competitiveness (Ahlburg, 1991). In this sense, remittances may have an effect like those of other booming sectors, such as mineral exports in Papua New Guinea, leading to the so-called ‘Dutch disease’ where resources flow out of other sectors that in the longer-run may have greater comparative advantage (Sturton, 1992). For example, Faeamani (1995) suggests that remittances are associated with a reduction in garden size and food production in Tonga. In addition, Borovnik (2006) gives an example where the difficulty of transferring remittances to the outer islands in Kiribati has contributed to the urbanisation of South Tarawa, with consequent environmental problems. On the other hand, a more positive view of the developmental impact of remittances in the Pacific is reported by the World Bank (2006). Despite the considerable literature on remittances and inter-household transfers in the Pacific, there is no consensus on the impact of these transfers on the distribution of income or on the responsiveness of transfers to short-term shocks. Pessimistic views about the ability of informal transfers to act as a safety net are reported by Morauta (1983, 1984) and Mounsell-Davis (1993). In contrast, Ward (1977) and Norwood (1984) argue that reciprocity still operates and income is spread widely through the community, and Borovnik (2006) argues that remittances from seafarers act as a safety net in Kiribati. Highlighting this lack of consensus may be important if policy-makers and external observers presume that just because transfer networks are so prevalent this will automatically give a built-in social safety net. Louise Morauta (1983, p 8) cogently argued that this is not necessarily more than 20 years ago: not everyone who needs transfers receives them or receives enough. This is because transfers in the wantok system are not transfers of charity or in a state welfare programme.13 They are part of a system of personal obligation, and some people who badly need transfers have nobody to help them, either because they have no relatives living near or because the relatives they do have are too poor or otherwise unwilling to help. Therefore, in this section we describe patterns of remittances and inter- household transfers in three locations in the Pacific (urban and rural Papua New Guinea and Tonga). These locations are chosen to illustrate several contrasts, including between international remittances and domestic transfers, between

13 A wantok is someone from the same extended family, clan, village or province (or in more modern settings even from the same workplace or school). The term comes from neo- Melanesian pidgin, from ‘one talk’ meaning people speaking the same language. With over 800 languages in Papua New Guinea the average group size for speaking the same language is only 6,000 people.

© Institute of Policy Studies, Victoria University of Wellington, 2008 229 Pacific Interactions urban and rural areas, and between urban squatter settlements and more formalised urban residential areas. Full details are provided in Gibson (2006a).14 The empirical framework is based on a seminal study of transfers in Indonesia that shows how an econometric model can be used to infer the attitude of remitters with regards to inequality among the recipients (Ravallion and Dearden, 1988). The key result from this model is an elasticity, which shows what the percentage change in transfers would be for a 1% change in the recipient household’s income. If this elasticity equals one, transfers rise or fall proportionate to recipient households’ change in income, and therefore transfers are a constant share of recipient household income and the distribution of income is not altered. If the elasticity is less than one, transfers reduce disparities in income and make the distribution more equal because they do not fall as quickly as income falls (and so are rising shares of recipient income). If the elasticity exceeds one, transfers increase disparities in income and make the distribution more unequal than it was before. Table 5.5 presents descriptive statistics on the remittance and transfer receipts. Approximately 90% of households in the 2005 Tongan and 1996 Papua New Guinea samples were receiving transfers. About two-thirds of the urban Papua New Guinea households in the late 1980s were receiving transfers, noting that the questionnaire used for this survey only covered two weeks of transfers rather than a whole year, which would tend to reduce the estimated participation rate. Households in urban squatter settlements, who have been highlighted in some studies as the most vulnerable and least connected to traditional exchange networks (Mounsell-Davis, 1993), appear to be more likely to be receiving (and making) transfers, even when compared with a group having the same (low) income level. In all three surveys, the median ratio of transfers to total expenditures (as a measure of long-run income) is about 10% for net recipients. Table 5.5 reports the key results for the elasticity of transfer receipts with respect to recipient income, which enables the impact on inequality to be evaluated. While the pattern of remittance receipts in Tonga suggests some reduction in inequality, since they do not fall as quickly as income falls (and so are rising shares of recipient income), this is not apparent in the rural sector of Papua New Guinea for transfers (which are almost all sourced from within Papua New Guinea rather than abroad). Indeed, for rural Papua New Guinea, which is easily the largest population in the Pacific, it appears that transfers are actually inequality increasing. In the urban sector in Papua New Guinea, transfers appear to be inequality reducing. The last two bars in Figure 5.4 look at comparisons just inside the urban sector of Papua New Guinea (and from a different survey than that used for the other Papua New Guinea results, so there is no comparison with the second and third bars on the chart). It is apparent that transfers reduce

14 Many outsiders are likely to view these squatter areas as some of the most deprived places in the Pacific, although this view is not always supported by statistical analyses of poverty, which find that living conditions are significantly worse in the rural sector (Gibson and Rozelle, 1998).

230 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? inequality in this data from urban Papua New Guinea, although mainly outside the squatter settlements.

Table 5.5: Importance of remittance and private transfer receipts in the Pacific

Percent of households Receipts as a percent of receiving transfers or total expenditure (for remittances (%) net recipients) (%)

Tonga 89.1 10.3 Rural Papua New Guinea 92.6 10.1 1996 Urban Papua New 89.0 6.9 Guinea 1996 Urban squatters, Papua 78.5 12.0 New Guinea 1985/90 Other urban, Papua New 60.0 7.6 Guinea 1985/90

Note: Results are comparable within but not between surveys. There are separate surveys for Tonga, for Papua New Guinea in 1996, and for urban Papua New Guinea in 1985/90. Source: Summarised from results presented in Gibson (2006a).

Figure 5.4: Elasticity of transfer receipts with respect to recipient income

-1

-2

-3 Elasticity of Transfers w.r.t. Income

-4 Tonga Rural PNG Urban PNG Settlements Other Urban

Note: PNG = Papua New Guinea. Source: Based on detailed regression results reported in Gibson (2006a).

What is clear from this analysis is that in the largest population studied (the rural sector of Papua New Guinea), informal transfers either leave the distribution of income unchanged or make it even more unequal. Thus, policy-makers may be wrong to assume that high rates of participation in informal exchange networks

© Institute of Policy Studies, Victoria University of Wellington, 2008 231 Pacific Interactions throughout the Pacific imply equally good informal safety nets in all areas. The adverse distributional effects of various policies cannot be ignored under the assumption that informal transfers will ameliorate the consequences. Furthermore, it also may be wrong to assume that these safety nets fail the most in settings that seem the most underprivileged (such as urban squatter settlements). Instead, it appears to be among the more invisible poor in the rural sector that informal transfers are not acting as an effective safety net.

Increasing value from financial transactions: Possible interventions Remittances are the fastest growing source of external capital for the Pacific Island economies and as described above, now total about NZ$600 million per year from all sources. Remittances can be expected to grow even further as short- term labour mobility is used as a mechanism to deal with seasonal labour shortages in New Zealand. As shown for the Tongan case in Figure 5.4, these international remittances may also help to reduce inequality. However, the transactions costs of sending remittances to the Pacific Islands are very high for the most widely used methods – in the order of 15–20% – and this effectively acts as a tax on households in the region (including the sending households in the Pacific communities of New Zealand and Australia). These transaction costs are much higher than elsewhere in the world. Sending money from the United States to Central America and the Caribbean typically costs less than 10% (Gibson et al, 2006). Finding ways to reduce the transactions costs of sending money to the Pacific is obviously of interest to policy-makers in the Pacific Islands. However, it is also of interest to the Pacific populations in New Zealand, since they are one of the main sources of remittances. Moreover, remittance costs are an issue that intersects with both population mobility and international trade, so are a good example of the cross-cutting issues at the heart of the current study. Evidence for the possibility of costs being lowered comes from the variation in costs among the different methods for sending money to the Pacific (Figure 5.5). For the New Zealand to Tonga remittance corridor illustrated in the figure the most popular methods are Western Union, Melie mei Langi (a church- based company), and bank transfers. On average, transactions costs for each of these popular methods take 15–20% of the amount sent, for the median transfer of $200. In contrast, transaction costs are less than 5% when the recipient uses an automated teller machine (ATM) card to withdraw funds from an account set up by the person sending money. It is notable that for smaller transfers, Melie mei Langi is more competitive than either the international money transfer operators or the banks (except for using an ATM card). Melie mei Langi, is run by the Tokaikolo Fellowship church but can be used by people of any denomination. Money is transferred in under one hour, and can be received at two branches in Tongatapu and at branches in Vava’u, Ha’apai and ‘Eua, while there is one branch in Auckland. The fact that a new entrant without the reputational and financial infrastructure advantages of

232 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster? incumbent banks and money transfer operators can provide a competitive service is suggestive of the presence of rents being earned by incumbent operators.

Figure 5.5: Transactions costs (as percentage of amount remitted) for the New Zealand to Tonga remittance corridor

60 Remittance Costs as a Percentage of the Amount Remitted

40 Median transfer

30 Percentage 20

0 50 100 150 200 250 300 350 400 Amount Remitted (NZ$) Bank Draft ATM Bank TT Moneygram Western Union Melie Mei Langi

Notes: ATM = automated teller machine; TT = telegraphic transfer. Source: Gibson et al (2006).

Extrapolating from the spread between the most popular and the cheapest remittance methods in Figure 5.5 to the entire Pacific, avoidable transactions costs across the Pacific Island economies (and their diaspora in the developed countries) may total NZ$60 million per year (Gibson, 2006b). This amount might be justified as the price of services such as accessibility, familiarity, and security that are provided by the incumbent operators. However, it can also be viewed as a surplus able to be used by the Pacific populations in New Zealand and the islands if lower cost operators introduce cheaper methods into these remittance corridors while still providing sufficient accessibility and security. Thus, understanding more about money transfer costs and choices is an important issue for the broader Pacific community.

Annex: Improving development data Another implication of the research reported here is the need for better distributional data in the Pacific, which means enhanced programmes of household surveys and improved access for researchers to this data. In particular, the results on inter-household transfers for urban Papua New Guinea that were presented above are based on data that is 20 years old, and it would be useful to see if the apparent strength of the urban informal safety net has persisted in the face of considerable shocks from macroeconomic and social disturbances.

More generally, information on the response of households to various opportunities, such as emigration and remittance receipt, cash cropping and improved initiatives for food production, would be valuable for assessing the impact on the welfare of the population in the Pacific Island economies. New Zealand and Australia are currently providing governmental support to technical solutions to reduce the transaction costs of remittances, with a goal of a reduction of costs to 7%, from the current 15–20% that prevails for most methods. These interventions include creating a Pacific equivalent to the SendMoneyHome.org website (http://www.sendmoneyhome.org), created as part of a remittance project of the Department of International Development in the United Kingdom, with the aim of improving transparency and promoting informed choice for remitters and competition among remittance providers. While not typically the subject of news releases, such interventions will eventually affect understanding and cooperation.

Developing the System of National Accounts household sector manual The System of National Accounts (SNA-93) has already provided the framework in which the contributions of the household sector could be incorporated to better understand the national economy. International donors are supporting statistical training and development in the region; an increased attention to this sector and developing regional standards in this area would assist. Household income and expenditure surveys are an important tool: establishing comparable collection protocols, for instance that separate agricultural and fisheries ‘subsistence’ production, would provide better comparable bases. One of the internationally recognised systems for preparing national accounts is the System of National Accounts produced by the Inter-Secretariat Working Group on National Accounts that represents the Commission of the European Communities, International Monetary Fund, OECD, United Nations, and World Bank. The system relies on being able to collect and compile data on all parts of the economy and to organise the measurement units in a codified and structure manner. The System of National Accounts is an accounting system that provides a format for comprehensive analysis, but how detailed the analyses can be depends somewhat on the detail provided in the accounts that a country compiles. Households are one of five institutional units used to define the total economy. According to the System of National Accounts manual prepared for the Palau Government (no date, p 6), “the sector is diverse and can actually be the largest productive sector in an economy”. The household sector includes as a subsector the informal sector. Unfortunately, as the manual notes (pp 7–8): This is an ongoing field of development for statisticians and the definitions of the informal sector remain fairly vague and therefore it is really up to Palau how best to define the sector. It may be desirable at some stage to come up with criteria to apply to the characteristics described in subparagraphs 5 (1) and (2) of the Annex to Chapter 4 in the SNA93 manual or to see if this has been done for other developing Pacific countries.

234 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Since this was written there have been additional reports on informal sector statistics; additional conceptualisation and protocols could assist in the collection of comparable data in Household Income and Expenditure Survey studies, especially ensuring data on agricultural and fishing and other subsistence sectors is segregated.

Benefits (and costs) of better data on food production The benefits permitted by developing projects from a better database and understanding of the existing conditions, or the costs incurred without such data, are especially borne by local communities and Pacific nations. As Pacific communities and governments assess potential development projects, or the costs of existing projects or disasters, they need to begin with a full understanding of the economic values of current practices. Often the true values become visible only when they are lost, such as the case when one of the three main municipalities of Yap State of the Federated States of Micronesia was affected by an oil spill from the Kyowa Violet (Nero, 2006). Due to the spill their reef area and mangrove swamp were closed for all purposes in 2002, and the community sought compensation for lost income due to the cessation of reef and lagoon fishing and shellfish collection. Only because this community was included in an earlier anthropological study of Micronesian food systems was data available in a form that was acceptable to the court to make the case. In 2004, the community received compensation for the documented loss of subsistence and market production and damage to the lagoon and mangrove swamp region affected. Future work is required to ensure that compensation can be received to cover the long-term effects of any extended loss of access to resources that results in the loss of the vital inter-generational transfer of knowledge from elders to their descendants concerning the management, conservation, fishing, and cultural distribution processes in Pacific fishing economies. These issues are critical practically, because a focus on the monetary market and import/export economy without valuing local food production creates an environment in which nations seek export dollars to increase their GNP. An alternate approach would be to better account for the food production that is currently not counted. At a wider governmental level, as Pacific governments institute export control of the small-scale export of reef fish (by air freight or accompanied coolers to kin, colleagues or buyers overseas) planners gain access to better data that also helps governments better manage lagoon fish resources and protect them from over- utilisation. Murray (2001) assessed the local socioeconomic and environment impacts of Tonga’s non-traditional pumpkin exports to Japan and Niue’s taro exports to New Zealand. With regard to Tonga, Murray noted (p 144): The sector is currently facing mutually re-enforcing problems of economic and environmental non-sustainability, and rural inequality has been further exacerbated through the boom.

With regard to Niue, Murray warned (p 145): [although] taro is one of the more successful Niuean exports … there have been few efforts by the government or other agents to provide the basis for the long-term sustainable development. … From an economic sustainability point of view the crop is highly vulnerable. … More worrying are the environmental implications of increasing taro production … with a devastating effect on ecosystems … No work has been undertaken to establish … [the] effect this has on the water lens. Murray (2001) argues that the export focus supports the further penetration of Pacific Island economies by the global agro-food complex, which has already had very negative impacts on Pacific livelihoods and environments.

Conclusion Island nations will benefit from closer documentation and a critical examination of island economies in ways that incorporate understandings of local food production for subsistence and market and the roles of households in the island nations’ economies, and more robust ways to assess contributions of Pacific economies and workers in the region.

References Abbott, D, and S Pollard (2004) Hardship and Poverty in the Pacific: Strengthening poverty analyses and strategies in the Pacific, Manila: Pacific Department, Asian Development Bank. ADB (2006) Key Indicators for Asia and the Pacific, Manila: Asian Development Bank. Ahlburg, D (1991) Remittances and their Impact: A Study of Tonga and Western Samoa, Pacific Policy Papers 7, Canberra: National Centre for Development Studies, Australian National University. Akin, D, and J Robbins (1999) Money and Modernity: State and Local Currencies in Melanesia, Pittsburgh: University of Pittsburgh Press. Alesina, A, A Devleeschauwer, W Easterly, S Kurlat, and R Wacziarg (2003) ‘Diversity’, Journal of Economic Growth 8(2), pp 155–194. Allen, B, M Bourke, and J Gibson (2005) ‘Poor rural places in Papua New Guinea’, Asia Pacific Viewpoint 46(2), pp 201–217. Alkire, W (1989 [1965]) Lamotrek Atoll: Inter-island socioeconomic ties, Prospect Heights, IL: Waveland Press. Armstrong, H, and R Read (2000) ‘Comparing the economic performance of dependent territories and sovereign micro-states’, Economic Development and Cultural Change 48(2), pp 285–306. Armstrong, H, and R Read (2002) ‘The Phantom of Liberty? Economic growth and the vulnerability of small states’, Journal of International Development 14(4), pp 435–458.

236 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Armstrong, H, and R Read (2006) ‘Geographical “handicaps” and small states: Some implications for the Pacific from a global perspective’, Asia Pacific Viewpoint 47(1), pp 79–92. AusAID (2006) Pacific 2020: Challenges and opportunities for growth, Canberra: Australian Government and AusAID. http://www.ausaid.gov.au/publications/pdf/pacific2020.pdf. Ballard, C, P Brown, R Bourke, and T Harwood (eds) (2005) The Sweet Potato in Oceania: A reappraisal, New South Wales: University of Sydney Press. Bellwood, P (1996) ‘The origins and spread of agriculture in the Indo-Pacific region: Gradualism and diffusion or revolution and colonization?’, in J Fox (ed), Indonesia: The making of a culture, pp 57–74, Canberra: Research School of Pacific Studies, Australian National University. Bertram, G (2004) ‘On the convergence of small island economies with their metropolitan patrons’, World Development 32(2), pp 343–364. Bertram, G, and O Karagedikli (2004) ‘Core-periphery linkages and income in small Pacific Island economies’, in J Poot (ed) On the Edge of the Global Economy, pp 106–121, Cheltenham: Edgar Elgar. Briguglio, L (1995) ‘Small island developing states and their economic vulnerabilities’, World Development 23(10), pp 1,615–1,632. Bohannan, P (1955) ‘Some principles of exchange and investment among the Tiv’, American Anthropologist 57(1), pp 60–70. Borovnik, M (2006) ‘Working overseas: Seafarers’ remittances and their distribution in Kiribati’, Asia Pacific Viewpoint 47(1), pp 151–161. Bourke, R, M Allen, and J Salisbury (eds) (2001) Food Security for Papua New Guinea: Proceedings of the Papua New Guinea Food and Nutrition 2000 Conference, ACIAR Proceedings 99, Papua New Guinea University of Technology, Lae, 26–30 June 2000, Canberra: Australian Centre for International Agricultural Research. Bourke, R, and V Vlassak (2004) Estimates of Food Crop Production in Papua New Guinea, Land Management Group, Canberra: Australian National University. Bourke, R, A McGregor, M Allen, B Evans, B Mullen, et al (2006) Solomon Islands Smallholder Agriculture Study: Vol 1. Main findings and recommendations, Canberra: Australian Government and AusAID. http://www.ausaid.gov.au/publications/pubout.cfm?ID=4088_5412_1071_619 3_2813. Chand, S (2001) ‘Lessons for development from Pacific Island countries’, Pacific Economic Bulletin 16(1), pp 95–108. Chand, S (2003) ‘Economic trends in the Pacific Island countries’, Pacific Economic Bulletin 18(1), pp 1–15. Clark, G, A Anderson, and D Wright (2006) ‘Human colonization of the Palau Islands, Western Micronesia’, Journal of Island and Coastal Archaeology 1, pp 215–232.

Clarke, G, and S Wallsten (2003) ‘Do remittances act like insurance? Evidence from a natural disaster in Jamaica’, mimeo, World Bank. Clarke, W, and R Thaman (eds) (1993) Pacific Island Agroforestry: Systems for sustainability Tokyo: United Nations University Press. Conley, T, and E Ligon (2002) ‘Economic distance and cross-country spillovers’, Journal of Economic Growth 7(2), pp 157–187. Connell, J, and R Brown (2005) Remittances in the Pacific: An overview, Manila: Asian Development Bank. http://www.apgml.org/frameworks/docs/8/ADB_remittances- pacific_Mar2005.pdf. Cox, D, and E Jimenez (1990) ‘Achieving social objectives through private transfers: A review’, World Bank Research Observer 5(2), pp 205–218. D’Arcy, P (2001) ‘Connected by the sea’, Journal of Pacific History 36, pp 163– 182. Dercon, S (2002) ‘Income risk, coping strategies and safety nets’, World Bank Research Observer 17(2), pp 141–166. Dewar, R (2003) ‘Rainfall variability and subsistence systems in southeast Asia and the western Pacific’, Current Anthropology 44(3), pp 369–388. Division of Economic Planning and Statistics, Department of Economic Affairs, FSM National Government (August 2007) International Trade Publication, Federated States of Micronesia, 2006, Palikir, Pohnpei, FM: Federated States of Micronesia. http://www.spc.int/prism/country/fm/stats/Publications/Annual/Trade/2006_% 20Intl_Trade_Pubn.pdf Docquier, F, and A Marfouk (2004) Measuring the International Mobility of Skilled Workers (1990–2000), policy research working paper 3381, World Bank. Docquier, F, and A Marfouk (2006) ‘International migration by educational attainment (1990–2000)’, in C Ozden and M Schiff (eds), International Migration, Remittances and the Brain Drain, New York: Palgrave-Macmillan. Duncan, R (2005) Quantitative Assessment of the Cost of Governance Failure in Fiji Islands, Nauru, Papua New Guinea and Solomon Islands, working paper 6, Asian Development Bank–Commonwealth Secretariat Joint Report to the Pacific Islands Forum Secretariat on Toward a New Pacific Regionalism. Easterly, W, and A Kraay (2000) ‘Small states, small problems? Income, growth and volatility in small states’, World Development 28(11), pp 2,013–2,027. Egan, J, M Burton, and K Nero (2006) ‘Building lives with food: Production, circulation, and consumption of food in Yap’, in R Wilk (ed) Fast Food/Slow Food: The cultural economy of the global food system, pp 31–47, Society for Economic Anthropology monographs, Lanham: Altamira Press.

238 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

European Community (2007) The Republic of Palau: European Community Country strategy paper and national indicative programme for the period 2002–2007, Brussels: European Commission. http://ec.europa.eu/development/body/csp_rsp/csp_en.cfm?cid=pw&status=1# pw. Faeamani, S (1995) ‘The impact of remittances on rural development in Tongan villages’, Asian and Pacific Migration Journal 4(1):139–156. Gibson, J (2001) ‘The economic and nutritional importance of household food production in Papua New Guinea’, in L Bourke, M Allen, and J Salisbury (eds), Food Security for Papua New Guinea, pp 37–44, ACIAR Proceeding 99, Canberra: Australian Centre for International Agricultural Research. Gibson, J (2006a) Are there Holes in the Safety Net? Remittances and inter- household transfers in Pacific Island Economies, working paper number 1, Pasifika Interactions project. http://ips.ac.nz/events/completed- activities/Pasifika%20project/Pasifika%20WP1.pdf. Gibson, J (2006b) Efficient Financial Services for Development in the Pacific, working paper number 2, Pasifika Interactions Project. http://ips.ac.nz/events/completed- activities/Pasifika%20project/Pasifika%20WP2%20.pdf. Gibson, J (2007) ‘Is remoteness a cause of slow growth in the Pacific? A spatial- econometric analysis’, Pacific Economic Bulletin 22(1), pp 83–101. Gibson, J, D McKenzie, and H Rohorua (2006) ‘How cost-elastic are remittances? Estimates from Tongan migrants in New Zealand’, Pacific Economic Bulletin 21(1), pp 112–128. Gibson, J, and S Rozelle (1998) ‘Results of the household survey component of the 1996 Poverty Assessment for Papua New Guinea’, mimeo, Population and Human Resources Division, World Bank. Gillett, R, and C Lightfoot (2001) The Contribution of Fisheries to the Economies of Pacific Nations, Manila: Asian Development Bank. http://www.adb.org/documents/reports/contribution_fisheries_pacific_econom ies/contribution_fisheries.pdf. Gosarevski, S, H Hughes, and S Windybank (2004) ‘Is Papua New Guinea viable?’, Pacific Economic Bulletin 19(1), pp 134–145. Green, R (1991) ‘Near and Remote Oceania: Disestablishing “Melanesia” in culture history’, in A Pawley (ed), Man and a Half: Essays in Pacific anthropology and ethnobiology in honour of Ralph Bulmer, pp 491–502, Auckland: Polynesian Society. Hausmann, R (2001) ‘Prisoners of geography’, Foreign Policy 122(1), pp 44–53. Horan, J (1997) ‘Production of Textile Koloa as “Development” in the Kingdom of Tonga: A case study of culture, development and anthropology’, thesis for Master of Arts in anthropology, University of Auckland, Auckland. Horan, J (2002) ‘Indigenous wealth and development: Micro-credit schemes in Tonga’, Asia Pacific Viewpoint 43(2), pp 205–221.

Howe, K (ed) (2006) Vaka Moana: Voyages of the ancestors, Auckland: Bateman. Hughes, H (2003) ‘Aid has failed the Pacific’, Issue Analysis 33 http://www.cis.org.au/issue_analysis/IA33/ia33.pdf. Jansen, T, B Mullen, A Pollard, R Maemouri, C Watotot et al (2006) Solomon Islands Smallholder Agriculture Study: Vol 2. Subsistence Production, Livestock and Social Analysis, Canberra: Australian Government and AusAID. http://www.ausaid.gov.au/publications/pubout.cfm?ID=4088_5412_1071_619 3_2813. Kaeppler, A (1978) ‘Exchange patterns in goods and spouses: Fiji, Tonga and Samoa’, Mankind 11(3), pp 246–252. Kaufmann, D, A Kraay, and M Mastruzzi (2006) Governance Matters. V: Governance Indicators for 1996–2005, policy research working paper, Washington, DC: World Bank. Kirch, P (2002) On the Road of the Winds: An archaeological history of the Pacific Islands before Europe contact, Berkeley, CA: University of California Press. Ladefoged, T, M Graves, and J Coil (2005) ‘The introduction of sweet potato in Polynesia: Early remains in Hawai’i’, Journal of the Polynesian Society 114(4), pp 359–373. Levett, M (1992) ‘Urban gardening in Port Moresby Part II: Inputs of land and labour, and value of production’, Fresh Produce News 72 (November). Lingenfelter, S (1975) Yap: Political leadership and cultural change in an island society, Honolulu: University Press of Hawaii. Lleras-Muney, A (2005) ‘The relationship between education and adult mortality in the United States’, Review of Economic Studies 72, pp 185–221. Macpherson, C (1994) ‘Changing patterns of commitment to island homelands: A case study of Western Samoa’, Pacific Studies 17(3), pp 83–116. Malinowski, B (1922) Argonauts of the Western Pacific, London: Routledge and Kegan Paul. Mazoyer, M, and L Roudart (2005) A History of World Agriculture: From the Neolithic to the current crisis, New York: Monthly Review Press. McKenzie, D (2007) `Remittances in the Pacific’, in Susan Pozo (ed) Immigrants and their International Money Flows, pp 99–121, Kalamazoo, MI: WE Upjohn Institute for Employment Research. Mintz, S (1986) Sweetness and Power: The place of sugar in world history, New York: Penguin Books. Morauta, L (1983) ‘Preliminary field report, Madang Phase of IASER Project on Interhousehold Transfers in PNG Towns’, mimeo, Institute of Applied Social and Economic Research. Morauta, L (1984) Left Behind in the Village, Institute of Applied Social and Economic Research monograph 25, Port Moresby: Papua New Guinea.

240 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Mosko, M (1999) ‘Magical money: Commoditization and the linkage of maketsi (“market”) and kangakanga (“custom”) in contemporary North Mekeo’, in D Akin and J Robbins (eds) Money and Modernity: State and Local Currencies in Melanesia, pp 41–61, Pittsburgh: University of Pittsburgh Press. Mosko, M (2002) ‘Totem and transaction: The objectification of “tradition” Among North Mekeo’, Oceania 73(2), pp 89–109. Mounsell-Davis, M (1993) Safety Net or Disincentive? Wantoks and Relatives in the Urban Pacific National Research Institute discussion paper 72, Port Moresby: Papua New Guinea. Murray, W (2001) ‘The second wave of globalization and agrarian change in the Pacific Islands’, Journal of Rural Studies 17, pp 135–148. Nero, K (2006) Assessing the Values of ‘Subsistence’: Lessons learned from litigating loss from an oil spill, working paper, mimeo, Macmillan Brown Centre for Pacific Studies: University of Canterbury. Norwood, H (1984) Port Moresby Urban Villages and Squatter Areas Waigani, Papua New Guinea: University of Papua New Guinea Press. O’Meara, T (1990) Samoan Planters: Tradition and economic development in Polynesia, Fort Worth: Holt, Rinehart and Winston. Palau Government (no date) National Accounts: The system and methods. http://www.palaugov.net/stats/PalauStats/Economic/NatAccts/SNAManual.pdf. Pawley, A, and R Green (1973) ‘Dating the dispersal of the oceanic languages’, Oceanic Linguistics 12(1/2), papers of the First International Conference on Comparative Austronesian Linguistics, 1974: Oceanic (Summer–Winter, 1973), pp 1–67. Petersen, G (2006) ‘Micronesia’s breadfruit revolution and the evolution of a culture area’, Archaeology in Oceania 41(2), pp 82–92. Pritchett, L (1997) ‘Divergence, big time’, Journal of Economic Perspectives 11(3), pp 3–17. Ravallion, M, and L Dearden (1988) ‘Social security in a “moral economy”: An analysis for Java’, Review of Economics and Statistics 70(1), pp 36–44. Redding, S, and A Venables (2004) ‘Economic geography and international inequality’, Journal of International Economics 62(1), pp 53–82. Sampson, T (2005) Notes on the Economic Performance of Small States 1995– 2003, working paper 2, Asian Development Bank–Commonwealth Secretariat Joint Report to the Pacific Islands Forum Secretariat on Towards a New Pacific Regionalism. Schoeffel, P (1994) ‘Women as beneficiaries of New Zealand ODA to the Tongan Development Bank: Gender and women in development issues’, unpublished consultancy report in Horan’s files. Schoeffel, P (1997) Myths of Community Management: Sustainability, the State and Rural Development in Papua New Guinea, Solomon Islands and Vanuatu, State, Society and Governance in Melanesia Discussion Paper 97/8, Canberra: Research School of Pacific and Asian Studies, Australia National University.

Silva, J, and S Tenreyro (2006) ‘The log of gravity’, Review of Economics and Statistics 88(4), pp 641–658. Sodhi, G (2008) Five out of Ten: A performance report on the Regional Assistance Mission to the Solomon Islands (RAMSI), Issue Analysis St Leonards, NSW: Centre for Independent Studies. http://www.cis.org.au/issue_analysis/IA92/ia92.pdf. Storey, A, J Ramirez, D Quiroz, D Burley, D Addison et al (2007) ‘Radiocarbon and DNA evidence for a pre-Columbian introduction of Polynesian chickens to Chile’, Proceedings of the National Academy of Sciences 104, pp 10,335– 10,339. Strathern, A, and P Stewart (1999) ‘Objects, relationships and meanings: Historical switches in currencies in Mount Hagen, Papua New Guinea’, in D Akin and J Robbins (eds), Money and Modernity: State and Local Currencies in Melanesia, pp 164–191, Pittsburgh: University of Pittsburgh Press. Sturton, M (1992) Tonga: Development through Agricultural Exports Economic Report 4, Honolulu: Pacific Islands Development Program. Vitusagavulu, J (2005) Retrospective Cost-Benefit Analyses of Air Pacific and Pacific Forum Line, working paper 5, Asian Development Bank– Commonwealth Secretariat Joint Report to the Pacific Islands Forum Secretariat on Toward a New Pacific Regionalism. Ward, R (1977) ‘Internal migration and urbanization in Papua New Guinea’, in R May (ed) Change and Movement: Readings on internal migration in Papua New Guinea Canberra: Australian National University Press. Waring, M (1999) Counting for Nothing: What men value and what women are worth, Toronto: Toronto University Press. Weiner, A (1976) Women of Value, Men of Renown, Austin: University of Texas Press. Weiner, A (1980) ‘Stability in banana leaves: Colonization and women in Kiriwina, Trobriand Islands’, in M Etienne and E Leacock (eds) Women and Colonization: Anthropological Perspectives, pp 270–203, Praeger Special Studies, New York: Begin & Garvey. Whitecross, N, and P Franklin (2001) ‘The role of rice in the 1997 PNG drought’, in M Bourke, M Allen, and J Salisbury (eds) Food Security for Papua New Guinea, proceedings of the Papua New Guinea Food and Nutrition 2000 Conference, Papua New Guinea University of Technology, Lae, 26–30 June 2000, ACIAR Proceedings 99, pp 255–259, Canberra: Australian Centre for International Agricultural Research. Women in Business Foundation (no date). http://www.intracen.org/bsrt/ casestudies/cssamoawibf.pdf. World Bank (2006) At Home and Away: Expanding Job Opportunities for Pacific Islanders through Labour Mobility, Washington, DC: World Bank.

242 © Institute of Policy Studies, Victoria University of Wellington, 2008 Why Don’t Pacific Island Countries’ Economies Grow Faster?

Ysaol, J, J Chilton, and P Callaghan (1995) Republic of Palau 1994 Census of Ocheraol Customs, Koror: Office of Planning and Statistics. Ysaol, J, J Chilton, and P Callaghan (1996) ‘A survey of betel nut chewing in Palau’, ISLA: Journal of Micronesian Studies 4(1), pp 244–255.