Permanent and Transitory Shocks Among Pacific Island Economies - Prospects for a Pacific Islands Currency Union*

ISSN 1178-2293 (Online)

University of Otago Economics Discussion Papers No. 1001

February 2010

Permanent and Transitory Shocks among Pacific Island Economies - Prospects for a Pacific Islands Currency Union*

Willie Lahari

Address for correspondence: Willie Lahari Department of Economics University of Otago PO Box 56 Dunedin 9054 NEW ZEALAND Email: [email protected] Telephone: +64 3 479 8131 Fax: +64 3 479 8174

*I am grateful to Alfred A. Haug for suggesting this study and comments provided. I am also grateful to Arlene Garces-Ozanne and seminar participants of the Economics Department at the University of Otago, for helpful comments and suggestions. Funding for this research was provided through a PhD scholarship award from the New Zealand Agency for International Development.

Abstract

This paper re-kindles the debate on the feasibility of a Pacific Islands currency union in view of the recent expansion and consolidation of regional strategies and agreements such as the ‘Pacific Plan’ and the Pacific Agreement on Closer Economic Relations Plus. These initiatives, including past efforts, have given limited consideration to the subject for a Pacific Islands currency union. This study exploits the optimal currency area theoretical framework and employs the Gonzalo and Ng (2001) decomposition method. This is the first time this method is used in the analysis relating to currency or monetary unions. Newly-constructed quarterly time series data are also applied. This paper investigates the dynamic effects of permanent and transitory shocks on key macroeconomic variables among Pacific Island countries (PICs). Evidence shows that the proposed union of six PICs (Fiji, PNG, Samoa, Solomon Islands, Vanuatu and Tonga) do not meet most of the preconditions for a union. However, further investigation shows evidence for the Melanesian countries (Fiji, PNG, Solomon Islands and Vanuatu) to possibly form a monetary union, preferably with the Australian dollar as the anchor currency. Nonetheless, further costs in terms of the alignment of policies by Melanesian countries are required.

JEL Classification: C3, C5, E3

Key words: Currency union, Gonzalo and Ng (2001) decomposition, Pacific Island countries

1. INTRODUCTION

In recent years, Pacific Island countries (PICs) and development partners in the region have begun developing new and consolidating past strategies aimed towards closer regional economic integration and growth in the region (Forum Economic Ministers Meeting (FEMM), 2000; Asian Development Bank, 2005). However, the issue for a possible Pacific Islands currency union has been given less attention in recent regional strategies and agreements such as the ‘Pacific Plan’ of 2005, and the newly negotiated Pacific Agreement on Closer Economic Relations (PACER) Plus.1 Although many PICs felt that the issue of a currency union was a challenge to national sovereignty, others, such as Fiji, thought the idea should be considered in the future. Australia was of the view that it was premature to emphasize a formal Pacific union given that PICs faced a period where further integration was required prior to any prospects for a Pacific union. While the strategies towards greater regional integration would help address regional problems that include sluggish economic growth, rising unemployment and negative external shocks (e.g., volatility in export commodity prices, and cyclones), such efforts can be seen as a necessary platform on which to re-kindle the debate on the prospects for a Pacific Islands currency union.

There are 22 Pacific Island countries and territories (PICTs) of which only six Pacific Island countries (PICs), namely, Fiji, Papua New Guinea (PNG), Samoa, Solomon Islands, Tonga and Vanuatu are fully independent sovereign states.2 This study will focus on these six PICs. These countries represent about 86% of the combined total population of all the PICTs and about 36% of the total combined gross domestic product (GDP), including 95% of the total land area of the 22 PICTs combined. The choice of these six PICs is relevant to this study given that these PICs have their own currencies (legal tender), exchange rate regimes and central banks. Hence, these six PICs determine their own domestic monetary (and fiscal) policies. Moreover, these six PICs share many common economic and physical characteristics, such as smallness in terms of GDP and population, narrow productive sectors, limited export diversification, geographic isolation, vulnerability to terms of trade fluctuations and frequent natural disasters that are important considerations for a possible currency union (Browne and Orsmond, 2006).

1 Other regional agreements include the Pacific Islands Trade Agreement (PICTA), the Pacific Agreement on Closer Economic Relations (PACER) signed in 2002, and the Melanesian Spearhead Group (MSG) Preferential Trade Agreement of 1993. 2 The 22 PICTs are grouped into three sub-regions of Melanesia (west Pacific), Polynesia (southeast Pacific) and Micronesia (north Pacific), on the basis of their ethnic, linguistic and cultural differences. The Melanesian sub-region consists of Fiji, New Caledonia, PNG, Solomon Islands and Vanuatu. The Polynesian sub-region consists of American Samoa, Cook Islands, French Polynesia (Tahiti), Niue, Pitcairn Islands, Samoa, Tokelau, Tonga, Tuvalu and Wallis and Futuna. The rest consist of Micronesian nations and territories 1

During different phases of economic development since the 1980s, economic (GDP) growth in the Pacific region has been subdued mainly due to negative external and internal shocks.3 Whether the effects of these shocks are asymmetric or not, and how PICs respond to these shocks will have implications for the feasibility of a currency union. Although inter-regional trade flows among the PICs, apart from Australia and New Zealand, has been low, trade has generally increased in the last two decades especially among the Melanesian countries.4 This has been attributed to increased regional efforts such as Pacific Island Countries Trade Agreement (PICTA) that have contributed to strengthening cooperation and efforts to increase trade and regional growth among PICs, Australia and New Zealand. In particular, the Melanesian countries have benefited from inter-country trade both as members of PICTA and their own Melanesian Spearhead Group (MSG) agreement. The continued success or failure of these current efforts will also impact on future considerations for a currency union.

This analysis draws from the Optimal Currency Area (OCA) theory flowing from the seminal contributions of Mundell (1961), McKinnon (1963) and Kenen (1969). A key property of the OCA refers to the asymmetry of shocks. As Mundell (1961) and McKinnon (1963) argued, countries interested in a currency or monetary union will have to be part of a zone where a single currency circulates or, if there are several exchange rates, their exchange rates are fixed to each other permanently. In such a zone, a single monetary authority implements a common monetary policy at the union level. Hence, governments will have limited control over the use of monetary and exchange rate policies to respond to country-specific shocks. A common union-wide response would be effective if the shocks were symmetric in nature across union member countries. However, if this was not the case, and some countries faced asymmetric shocks and were worse-off than the other countries within the same zone, then it would not be beneficial for these countries to forego domestic policy autonomy and form a union. Generally, a currency union refers to a zone consisting of several countries or regions where a single exchange rate regime prevails, a single currency circulates, and where a single monetary authority implements a common monetary policy. A monetary union refers to a group of countries that agree to permanently fix their exchange rates under centralisation of monetary authority. A monetary union may not necessarily be fully centralised or have formal integration but there are commitments for monetary policy coordination among members through arrangements such as currency boards. For the general purpose of this discussion, the term ‘currency union’ will be

3 Examples include the global oil crisis in the 1980s, the decline of major agricultural commodity prices in the mid-1980s, the Asian financial crisis in 1997 and natural disasters (e.g., cyclones). Internal factors relate to civil unrests such as the coups in Fiji and the lack of good governance among many public institutions within the PICs. 4 For instance, from 1990 and 2005, the Solomon Islands imports from PNG increased from 2.8% to 4.1%, Vanuatu imports from Fiji increased from 0% to 6%, and Tonga’s exports to Fiji doubled. 2 used interchangeably to refer to either a currency or monetary union, or both, unless otherwise specified.

The study will employ the method of Gonzalo and Ng (2001) in analysing the dynamic effects of permanent (P) and transitory (T) shocks, denoted as P-T shocks, in a system of key macroeconomic policy variables. Initial investigations in assessing the degree of convergence (cointegration) of the variables will also be undertaken. Specifically, this analysis will examine the extent of (a)symmetry in the behaviour of P-T shocks, and associated costs and benefits as preconditions for PICs forming a currency union. This is undertaken in terms of evaluating the degree of similar (dissimilar) behaviour of P-T shocks pertaining to the following conditions: (a) dominance of either a P or T shock from variance decomposition; (b) direction of the effects of P- T shocks; (c) magnitude of P-T shocks upon initial impact; (d) persistence of P-T shocks, and (e) correlation of P-T shocks. This study will draw insights from recent empirical studies on consumption and wealth that have applied the Gonzalo and Ng (2001) methodology such as Lettau and Ludvigson (2004), Chen (2006), Kishor (2007), de Veirman and Dunstan (2008), and Giancarlo and Konstantinou (2009), but applied instead to the analysis of currency unions. The application of Gonzalo and Ng’s (2001) decomposition to the study of currency or monetary unions is novel. To my knowledge, there are no studies on currency unions that have applied Gonzalo and Ng’s (2001) methodology.

2. RELATED STUDIES FOR THE PACIFIC

There is clearly a lack of comprehensive analysis on currency unions in the Pacific region. Earlier studies centred on appropriate currency regimes, such as Xu (1999) and Duncan and Xu (2000) who argued that PNG should adopt the Australian dollar given the lack of central bank independence during the 1990s and the likelihood of monetary indiscipline that may arise due to internal or external interference. They argued that the economy would perform better with the Australian dollar. Similarly de Brouwer (2000) supported the idea that PICs should adopt the Australian dollar, whilst Duncan (2002) argued more in favour of the Solomon Islands adopting the Australian dollar. However, the arguments of Duncan (2002) and de Brouwer (2000) were policy arguments and lacked advanced empirical analysis. On the other hand, Hughes (2003) advised against any form of currency union in the Pacific prior to real reforms being undertaken. Similar views were expressed by Fichera (2006) and Creane et al., (2006). However, at about the same time that Hughes’ (2003) work was being debated, the Australian Parliamentary Senate Committee (2003) proposed a Pacific economic and political community as a panacea to the economic problems and political instability in the Pacific region. The discussions did not go far,

3 with little debate due to a general lack of interest among political leaders in the region and negative assessments for a currency union (Chand, 2003).5 Furthermore, studies such as those of Jayaraman (2001, 2003, 2004), Bunyaratavej and Jayaraman (2005), Bowman (2005) and Browne and Orsmond (2006) argued that it was premature for PICs to form a currency union given the dissimilarities in the behaviour of macroeconomic policies and key measures such as inflation and exchange rates. A number of these studies, such as Jayaraman (2001, 2003), have been criticised in connection with the application of nominal and effective exchange rates given the varying structure of the weights and issues of endogeneity of currency unions and trade (Duncan, 2005). Bowman’s (2005) test for the relationships of currencies alone was not considered an appropriate test for an OCA given that all the six PICs’ currencies were fixed to individual baskets, except for PNG whose currency is freely (managed) floated, and thus the tests do not assess the extent to which these PICs and Australia share industry-specific shocks (Duncan, 2005). Other related studies focused mainly on regional co-operation and integration such as those by the Asian Development Bank (2005), and Australian Agency for International Development (2008). From the above studies, it is obvious that there is lack of consensus on the issue of currency union for PICs given the mixed results and conflicting arguments.

3. DATA AND EMPIRICAL METHODOLOGY

This section discusses the data aspects of the analysis and the model employed. The discussion of the data will include the newly-constructed series for the PICs and the choice of variables. The Gonzalo and Ng (2001) decomposition method will be discussed in the subsequent sub-section.

3.1 Data

Time series data for key macroeconomic variables from existing data sources and newly- constructed time series data at a quarterly frequency are employed in this analysis. Given the time series requirements for this study and the general lack of timely and reliable quarterly data, newly-constructed series for relevant variables were derived.6 The time series are mainly from 1980:1 to 2006:4. However, a number of variables had relatively shorter time spans such as the

5 Many PICs were not interested in the idea mainly due to issues of loss of sovereignty (relating to surrendering monetary authority and monetary policy). Also, the PICs’ non-involvement in the initial process of the Australian Parliamentary Senate Committee’s recommendation, implied that PICs did not having ownership of the idea. PICs viewed that the idea was more attune to the interests of Australia and to some extent, New Zealand (see, e.g., Chand, 2003; Duncan, 2005). 6 See Lahari et al. (2008) and the Secretariat of the Pacific Community (2003, 2005, 2007) for discussions about the data issues and problems in the Pacific. 4

GDP series for New Zealand that started from 1987:2 to 2006:4.7 The new data includes the quarterly real GDP series for the PICs constructed by Lahari et al. (2008). The other newly- estimated series include the fiscal (government budget) balance to GDP ratio (FGDP), and the public debt to GDP ratio (DebtGDP). The methodology used by Lahari et al. (2008) is based on the Chow-Lin (1971) framework, with modifications proposed by Fernandez (1981), Litterman (1983), and Abeysinghe and Rajaguru (2004). Following the same approach as in Lahari et al. (2008), the estimation of quarterly series for FGDP and DebtGDP have been undertaken. Data descriptions are provided in Table A2 in the Appendix. The other variables employed include consumer prices (CPI), interest rates (IRATE) and the money supply (M2) whose data descriptions are provided in Table A1 in the Appendix. For our purposes, newly-constructed trade weighted nominal exchange rate indices (TWIs) for each of the PICs were used instead of the exchange rates.8 The TWI for each PIC was constructed following the method used by the Reserve Bank of New Zealand (RBNZ).9 The TWIs are compiled for an independent country currency basis and in the counterfactual monetary union with Australia as anchor currency 10,11 denoted as TWI-AUD, and New Zealand as anchor denoted as TWI-NZD respectively. The latter case involves a hypothetical case to evaluate the choice among PICs for adopting the currency of either one of their closest major trading partners, Australia or New Zealand.12 The details of the TWIs and descriptions are presented in Table A1 in Appendix.

The choice of variables was essentially based on the criteria of the Maastricht Treaty of 1991 (European Union, 2006), applied in the context of the PICs.13 However, the focus in this study is on the underlying policy goals of the Maastricht Treaty, as measured by the variables, and not the

7 The time series data for New Zealand, mainly GDP and fiscal (government budget) balance, started from 1987:2 due to the massive financial reforms in the prior years. Hence, where these variables are analysed within a group of countries including New Zealand, the time period will be based on the series for New Zealand. 8 The use of the TWI based on a common weighting structure of currencies represent a better measure of the degree of economic competitiveness of a country relative to its major trading partners. 9 See the RBNZ website, http://www.rbnz.govt.nz/statistics/exandint/twi/index.html 10 Studies such as Scrimgeour (2001, 2002) have compiled similar TWIs in assessing New Zealand’s experience in a currency union with its major trading partners. 11 Given the lack of consensus on the impact of a currency union on trade, and the controversies around Rose’s (2000) findings, weights are held fixed. Similar arguments have been raised by Scriemgor (2001, 2002) and Grimes, Holmes and Bowden (2000). However, in this analysis, the weights are based on the average value of the total merchandise trade and nominal GDP from 1980 to 2006 to avert any bias from the selection of single- year weights (see, e.g., Sawides, 1996). 12 The nominal TWI is employed throughout this study given our interest on the impact of P-T shocks on the economy. In such a scenario, it is the nominal exchange rate (TWI) that acts as the immediate buffering variable. See also arguments by Grimes, Holmes and Bowden (2000). 13 The Maastricht Treaty criteria, often referred to as the convergence criteria, were based on five key economic pre-conditions for entry into the European Monetary Union. These includes: (i) a country’s budget deficit to GDP ratio should not exceed 3% per annum; (ii) a country’s total public sector public debt to GDP ratio should not exceed 60% per annum; (iii) a country’s exchange rate must be contained within the required fluctuation margins of 2.25% of the EU’s exchange rate mechanism (ERM); (iv) a country’s rate of inflation must not exceed 1.5% above the average rate in the three lowest inflation EU countries, and (v) a country’s long-term interest rates should not exceed 2% above the average of those in the three lowest rate EU countries. 5

specific targets set by the treaty.14 This is relevant in respect to setting a benchmark for comparison for PICs which intend to form a currency union.15 In this context, the measures of FGDP and DebtGDP are seen as key indicators for fiscal policy. The other variables such as the CPI, exchange rate (TWI) and IRATE, are regarded as key measures for monetary policy, although the CPI is also a key measure in assessing fiscal policy. GDP is an output measure of economic performance influenced by both fiscal and monetary policies. The extent to which a group of countries’ policies behave and react to shocks has implications on forming a currency union.

Cluster of Potential Currency Union Groups

The six PICs (Fiji, PNG, Samoa, Solomon Islands, Tonga and Vanuatu) together with Australia and New Zealand are clustered into potential union blocks on the basis of their prevailing trade agreements, culture, language and historical connections. Hence, the following potential groups are proposed: Group 1: Pacific Only (Fiji, PNG, Samoa, Solomon Islands, Tonga and Vanuatu); Group 1A: Pacific with Australia; Group 1B: Pacific with NZ, Group 2: Melanesia (Fiji, PNG, Solomon Islands and Vanuatu); Group 2A: Melanesia (Group3) with Australia; Group 2B: Melanesia with New Zealand. The challenge was to first examine the possibility of countries in Group 1 (the Pacific group) forming a currency union. Evaluating initial prospects for uniting the PICs in a union would help distinguish possible prospects for PICs (least developed/developing economies) first, prior to further assessment with Australia and/or New Zealand (advanced/developed economies).16 Should this hold, then a further examination to form a union with Australia and/or New Zealand is undertaken. If not, then out of the Group 1 countries, a group of Melanesian countries (Group 2) is investigated separately.17 If forming a union is

14 The pre-conditions (or targets), other than the set of measures (variables) are specific for EU countries. The targets are ambitious and may not be reachable by many PICs given the extent of heterogeneity between the economies of the EU (mainly advanced developed economies) compared with PICs (least developed/developing economies). For example, the annual inflation rates for Fiji was 4.2% (2003) and 2.3% (2005) and for the Solomon Islands, 10.3% (2003) and 6.8% (2005). These rates exceed the ‘less than 1.5% of the average’ of the three lowest EU countries. 15 A number of studies (e.g., MacDonald and Taylor, 1991; Hafer and Kutan, 1994; and Haug, 2001) have analysed macroeconomic policy convergences employing similar variables to the Maastricht Treaty. 16 While inter-county trade flows among PICs may be relatively low, except with Australia/NZ, trade has increased among PICs. Thus, arguments against prospects for a union on the basis of low trade alone are debateable because forming a currency union can lead to substantial increases in trade (Rose, 2000), although this is controversial. More importantly, the issue of trade is only one aspect of OCA theory. Among other key OCA properties, the asymmetries of shocks are main concerns for this study. 17 The Melanesian countries of the South Pacific are a sub-regional group established as the Melanesian Spearhead Group (MSG) in 1993. Recently, the MSG Secretariat was established in 2008 to implement the regional developmental agenda of the Melanesian countries. 6

feasible for Group 2 then further investigation for prospects for Group 2 countries forming a currency union with Australia (Group 2A) and/or New Zealand (Group 2B) will be undertaken.18

3.2 VECM MODEL

The Vector Error Correction Model (VECM) incorporates cointegration restrictions on the vector autoregression process in a similar methodological framework to that of Gonzalo and Ng (2001). The Gonzalo and Ng (2001) method provides a systematic framework in identifying and decomposing the permanent and transitory (P-T) shocks among key variables of interest. Thus,

the VECM representation of Zt is presented in the reduced form as:

ΔZt = γα′Zt−1 + Γ(L)ΔZt−1 + et (1)

where ΔZt represents first differences of a 1 vector of I(1) process of Zt; is the number of cointegrated vectors in the system with , and 1, where is the number of variables; is the matrix of adjustment coefficients or error-correction parameters; consist of the cointegrating vectors, and Γ(L) is the lag polynomial.

Permanent and Transitory Decomposition

Following from Gonzalo and Ng (2001), we can generally represent the final-form of the P-T shocks (orthogonalised) from equation (1) as:

~ ~ ΔZ t = D(L)ηt (2)

~ ~ where D(L) is a lag polynomial, and ηt is a 1 vector of the transformed orthogonalised P-T ~ shocks. The ηt s are serially and mutually uncorrelated, and have unit variance. According to

~ p Gonzalo and Ng, the 1 vector of shocks, ηt , are the common stochastic trends or

permanent shocks in the VECM when lim∞ / 0 for the levels of Zt. ~ T Similarly, the 1 vector of shocks, ηt , refers to transitory shocks when lim∞ /

19 ~ 0. The process of decomposing ηt is summarised as follows; the initial phase involves isolating the structural innovations, denoted as ut, from the VECM, distinguished by their

18 Several studies have been conducted on prospects for a currency union between Australia and New Zealand. See, e.g., Grimes et al. (2000, 2004), Grimes (2005), Scrimgeour (2001, 2002) and Haug (2001), among others. 19 See Gonzalo and Ng (2001, p.1530) Definition 1 for discussion. 7

P T ‘unorthogonalised’ permanent and transitory components, denoted as ut and ut . This is attained by a Wold moving-average representation in the form of equation (3). Thus, initially, we can generally represent equation (1) as,

ΔZ t = C(L)e t (3)

where C(L ) is the lag polynomial and et is an 1 vector of the reduced form residuals from the

P T 20 VECM. The task then is to find a matrix G that transforms et into a set of ut or ut shocks. Thus, we set , where γ′ γ = 0, and γ′ is the orthogonal complement to γ . Since γ′ ⊥ ⊥ ⊥ defines the permanent shocks, γˆ can have poor sampling properties (Podivinsky, 1992). Hence, Gonzalo and Ng suggest constraining insignificant estimates of γˆ to zero.21 Not restricting this can result in an unstable process. Thus, we isolate u P and uT by computing G with the residuals t t P of the VECM, et, in the form , where the 1 vector ut = γ⊥′ et and 1

T P T vector ut = α′et . The next process involves ‘orthogonalising’ ut and ut so that they are mutually uncorrelated.22 This allows us to obtain the impulse responses and variance decompositions. This is done through obtaining a lower block triangular matrix, H, by applying 23 the Choleski decomposition to cov(Get). Thus, the transformation converts ∆ to arrive at ∆ as in equation (2) earlier. The transformed orthogonalised P-T shocks are given by . Lastly, the decomposition of the P-T shocks can now be summarised as in equation (4) below, to arrive at equation (2) earlier,

−1 -1 −1 ~ ~ ΔZt = C(L)G HH Get = D(L)HH ut = D(L)ηt (4)

Gonzalo and Ng’s (2001) method is less complex in computation given its explicit derivation process. In contrast to the other approaches, particularly that of King et al., (1991) where economic theory is used to identify the short and long-run dynamics of the model, Gonzalo and Ng’s method allows for restrictions imposed by the data. This approach is supported by arguments from Cochrane (1994a, 1994b) based on the limitation of information for identifying the origins of shocks. A limitation of the Gonzalo and Ng (2001) method involves the

20 The application of G is motivated by the Granger Representation Theorem. See Gonzalo and Ng (2001, p.1531) Proposition 1 for further discussion. 21 Refer to the example in Gonzalo and Ng (2001, section 4.2). 22 For further discussion, see, Gonzalo and Ng (2001, section 2.3). 23 See ‘practical rule’ in Gonzalo and Ng (2001, p.1532) for further discussion. 8

identification of the shocks that are constrained to the cointegration restrictions imposed, as no appropriate rule exists around a subjective choice of H, in respect to orthogonalising the structural shocks.24

4.0 EMPIRICAL RESULTS

The analysis proceeds with tests for unit roots and cointegration of the selected macroeconomic policy variables to assess convergence of macroeconomic policies. This will then form the basis for the OCA analysis of shocks that will involve the variance decomposition and impulse responses where the dominance of P-T shocks, direction of effects of P-T shocks, the magnitude, persistence, and the correlation of impulse responses of P-T shocks will be analysed.

4.1 Tests for Unit Roots

Unit root test results are based on the ADF test, choosing lag augmentations based on Akaike’s Information Criteria (AIC; see Akaike, 1974). The results showed that most variables for all PICs including Australia and New Zealand were I(1) apart from mixed results for a few variables. For example, DebtGDP for Tonga was I(0) and the IRATE for Fiji was I(0). Moreover, the CPI series for Australia and New Zealand were I(2) and transformed to I(1) by taking first-differences to arrive at inflation rates.25 This was also the case for Samoa’s TWI. It was noted that the FGDP data for Samoa, Tonga and New Zealand were for July-ending fiscal years or of shorter span.26 To a considerable extent, the results from the unit root tests show early indications of the nature and behaviour of the macroeconomic policies among PICs, and with Australia and New Zealand.

4.2 Tests for Cointegration

The cointegration tests (Johansen, 1988; Johansen and Juselius, 1990) were undertaken for the variables specified with the same order of integration. The Schwarz Information Criterion (Schwarz, 1978) was used to determine the appropriate lag length of the VAR for the cointegration test.27 The maximum lag lenght was set at 4 and upon establishing the optimal lag length, cointegration tests were conducted. Both the trace and the maximum-eigenvalue test

24See Gonzalo and Ng (2001, p.1532) Assumption 1 for further discussion. 25 Although Tonga’s CPI appeared to be I(2), taking first-differences showed it was I(0) rather than I(1). Hence, it was treated as I(1) in levels. Other tests such as Phillips-Perron’s also showed that it was I(1) in levels. 26 The reference year for this study was the calendar year (January to December) as in section 3. 27Pertaining to the cointegraton test, the SIC was chosen given its general preference, although there is no single appropriate criterion for lag length selection (see, e.g., Stock, 1994, p. 2781). 9

Table 1: Summary of Johansen Cointegration Test Results and Degree of Convergence by 28 Proposed Union Groups Number of Cointegrating Vectors (r)a Group 2 Group 2A Group 2B Variables Group 1 (Melanesia) (Melanesia+ (Melanesia + NZ) (Pacific (n=6) (n=4) Aust) (n=5) (n=5)

DebtGDP (%) - 1** 1** 0* ln(CPI) 1** 2**, 3* 4** 4** ln(TWI)b 1** 0* 1**c 1** c ln(real GDP) 4**, 5* 3** 3** 3** ln(real M2) 0* 0* 0* 1*

Degree of Convergenced

Complete Convergence Real GDP Real GDP, CPI CPI CPI (n - r =1) where r ≠ 0 Real GDP, Real GDP, Partial Convergence CPI, TWI DebtGDP DebtGDP, TWI-NZD, (n - r ≥ 2) where r ≠ 0 TWI-AUD Real M2

No Convergence (r =0) Real M2 TWI, Real M2 Real M2 DebtGDP

Notes: a Cointegration results are based on the trace rank test, unless otherwise specified; ** and * represents 5% and 10% significance levels respectively. b The TWIs for Group 2A and Group 2B are based on the Australian dollar as the anchor currency (TWI-AUD) and New Zealand dollar as the anchor currency (TWI-NZD) respectively. Under these arrangements, n=4. c The results were based on the maximum-eigenvalue test as the trace test showed the results to be of ‘full rank’ at both 5% and 10% significance levels. The maximum-eigenvalue test result was considered as an alternative given the sensitivity of the choice of r imposed on Gonzalo and Ng’s method as discussed in section 4.2. See also section 4.3 and footnote 31. d For the purpose of computing the degree of convergence (n-r), when r varies at 5% and 10% significance levels, the optimal r was chosen at the 10% significance level. results were generated but for the purpose of this analysis the trace test results are used.29, 30 Table 1 shows the number of cointegrating vectors (r) within the system by variable and by proposed union groups. The table also shows the degree of convergence (co-movement) in terms of cointegration, drawing from similar applications in empirical studies (e.g., Hafer and Kutan, 1994; Haug, 2001) who have analysed macroeconomic policy convergences regarding currency unions using cointegration analysis. Whilst cointegration depends on r, with 0 cointegrated vectors, ‘complete’ convergence in the long-run, in this discussion, refers to the situation where 1. This indicates one common long-run stochastic trend represented by the permanent shocks shared among the variables in the system. Hence, it is one of the strongest conditions for a union, especially when implementing a union-wide common monetary policy. ‘Partial’ convergence is when , where 2. This indicates more than one common

28 The details of the trace and maximum-eigenvalue test results are available upon request.

29 Although the maximum eigenvalue test has usually more power, as the alternative hypothesis is narrower, than the trace test, there may be differences in small samples where the trace test may in some situations appear superior, although there are some distortions in size (see, e.g., Lu tkepohl et al., 2001) 30 However, due to the implications regarding r, the cointegration results from the maximum eigenvalue test would be considered only in the case where the trace test result leads to a contradiction with the unit root tests (for r = n). 10

permanent stochastic trend within the system. Evidence from Table 1 shows two ‘complete’ convergences in real GDP and the CPI among the Melanesian countries (Group 2). These suggest strong similarities among key macroeconomic policies relating to economic growth and inflation among the Melanesian countries which are ideal conditions for a union. However, no convergence was observed for the exchange rates (TWI) of the Melanesian countries other than when pegged to the Australian and New Zealand currencies. This was expected given the unstable and volatile nature of the currencies of Melanesian countries. Other proposed union groups show partial convergences in variables (e.g., real GDP) and no convergences (e.g., real M2 and DebtGDP). These results indicate some differences in policies among the PICs, and with Australia and New Zealand, respectively.

4.3. Variance Decomposition, Direction, Magnitude & Persistence of Shocks

As suggested by Gonzalo and Ng (2001), insignificant error-correction coefficients, γ , were restricted to zero in our one-lag-order VECM model. Avoiding this would have resulted in an unstable process with implications for variance decomposition and impulse responses. From the variance decomposition results (see Tables A4 to A7 in the Appendix), the dominating or influential shock (either P or T) is represented by its relative share being within the magnitude of 75-100% of the total variance of both P-T shocks, from horizon (h) 1 to 12 quarters. The findings regarding the direction of the effects of the P-T shocks to the initial (h=1) impact, following a +1% shock to the system, are summarised in Table A3 in the Appendix. Figures 1 and 2, and A1 to A3 (see Appendix) show the magnitude and persistence of P-T shocks from the dynamic impulse response functions.

The impulse responses generated by the Gonzalo and Ng (2001) method differ from the standard impulse response functions because they do not have a direct interpretation as a specific shock like a monetary policy shock. For example, in Figure 1 below, the impulse responses for GDP are generated by a positive impact to the system shared by the PICs (Group 1). As a result, the system simultaneously emits one permanent shock and three different transitory shocks that affect PICs. This is in contrast to the conventional impulse responses where impulse responses are generated by the impact of one variable on the other variables within the system. Moreover, as noted earlier, this analysis does not investigate the sources of P-T shocks as these are difficult to identify. However, it is important to note that the permanent and temporary components for a given variable, for instance, real GDP, may be related to supply and demand disturbances. Aggregate supply disturbances may affect the level of real GDP (growth) permanently. This is because aggregate supply shocks are mainly driven by productivity and technological

11 innovations. On the other hand, aggregate demand shocks may only have temporary effects on growth. This is because aggregate demand shocks are driven mainly by factors such as changes in polices, volatility in world commodity prices and consumer preferences.

Group 1 (Pacific): From the variance decomposition results, the dominance of either a P or T shock on growth (real GDP) and prices (CPI) appeared mixed (see Table A4 in the Appendix). However, only the P-shocks were the key drivers for movements in exchange rates (TWI) while the T-shocks were predominant among prices (CPI) for all PICs, excluding Samoa. Nonetheless, the general findings confirm the lack of coherence in macroeconomic policies. Such differences are difficult to ascertain but can be generally linked to supply and demand factors. For instance, Samoa and Tonga share some common features such as their dependence on remittances and tourism receipts, unlike a number of Melanesian countries such as PNG and the Solomon Islands. These factors have a significant influence on Samoa and Tonga’s exchange rates in maintaining a consistent flow of foreign reserves. However, this was to some extent inconsistent with the results that showed that their exchanges rates (TWIs) were affected differently. Samoa’s TWI is affected predominantly by T shocks while Tonga’s is dominated by P-shocks. Such dissimilarities are not conducive for a union. Moreover, a 1% shock to the system resulted in some similarities observed in the direction of the effects of the P-T shocks for variables such as real GDP (positive P-shocks) except for Samoa, CPI (positive T-shocks) and TWI (positive T-shocks), except for Tonga. In terms of the magnitudes of shocks, our results revealed that the initial impacts were relatively low (less than 1%) for all the variables. The low impact is less costly for a union. However, the speed of adjustment of P-T shocks for most of the variables took longer to stabilise or fade away. For example, P-shocks persisted for about 1 year (h=4) to 3 years (h=12) for most of the variables while real GDP took about 5 years (h=20) to just under 8 years (h=32) before reaching new equilibrium levels. The impulse response for GDP is shown in Figure 1. It appears that the third and fifth transitory shocks show no clear pattern of fading. This may be may be attributed to sensitivity of cointegrating restrictions. In terms of the persistence of shocks, most of the P-T shocks took generally longer than one year (h=4). Although the impacts of shocks were minimal, the speeds of adjustment were longer. The latter condition is costly for a union. In all, these findings are consistent with arguments from previous studies such as Hughes (2003), Jayaraman (2003, 2004) and Fichera (2006), among others, that it is premature for PICs to form a currency union.

Group 2 (Melanesia): Evidence from the variance decomposition (see Table A5 in the Appendix) shows a considerable degree of similar behaviour in the dominance of either a P or T shock for the variables among the majority of Melanesian countries. In particular, the T-shocks were the

Figure 1: Dynamic Impulse Response of P-T Shocks for Real GDP - PICs (Group 1)

1.2 Permanent Shock 1 0.04 Transitory Shock 1 1.0 0.03 0.8 0.02 0.6 0.01 0.4 0.2 0.00

Impulse response 0.0 Impulse response ‐0.01 ‐0.2 1 6 11 16 21 26 31 36 1 6 11 16 21 26 31 36 ‐0.02 ‐0.4 ‐0.6 ‐0.03 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

0.7 1.2 Transitory Shock 2 Transitory Shock 3 0.6 1 0.5 0.8 0.4 0.6 0.3 0.4 0.2 0.2 0.1 0 0.0 Impulse response Impulse response ‐0.1 ‐0.2 1 6 11 16 21 26 31 36 ‐0.2 ‐0.4 1 6 11 16 21 26 31 36 ‐0.3 ‐0.6 ‐0.4 ‐0.8 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

0.4 0.3 Transitory Shock 4 Transitory Shock 5 0.3 0.2

0.2 0.2

0.1 0.1

0 0.1 Impulse response Impulse response ‐0.1 0.0

‐0.2 1 6 11 16 21 26 31 36 ‐0.1 1 6 11 16 21 26 31 36 ‐0.3 ‐0.1 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

Notes: Vatu = Vanuatu; PNG = Papua New Guinea; S.I. = Solomon Islands.

key drivers for growth (real GDP) and prices (CPI), while the P-shocks were predominant in debt (DebtGDP). This is consistent with economic expectations of factors that impede economic growth (real GDP) for many Melanesian countries. Some of these factors include volatility of world export commodity/mineral prices, tourism receipts and foreign aid that constitute mainly transitory shocks. Permanent shocks such as productivity and technological advances appear less 13

Figure 2: Dynamic Impulse Response of P-T Shocks for Real GDP – Melanesia (Group 2)

1.2 Permanent Shock 1 0.8 Transitory Shock 1 1.0 0.7 0.6 0.8 0.5 0.6 0.4 0.3 0.4 0.2

Impulse response 0.2 Impulse response 0.1 0.0 0.0 ‐0.1 1234567891011121314 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ‐0.2 ‐0.2 Quarters(h Quarters (h) ) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

0.5 Transitory Shock 3 Transitory Shock 2 1.2

0.4 1.0

0.3 0.8 0.2 0.6 0.1

Impulse response 0.4 0.0 Impulse response 1234567891011121314 0.2 ‐0.1 0.0 ‐0.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ‐0.2 ‐0.3 Quarters (h) Quarters (h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

Notes: See Figure 1.

significant drivers for growth. Our findings also showed common similarities in the direction of the effects of P-T shocks (see Table A3). For instance, the positive effects of P-shocks were predominant among real GDP, CPI and DebtGDP. This showed a degree of similarity in the reactions of P-T shocks upon initial impact. To a considerable extent, these findings are positive indications for a currency union. Furthermore, from a 1% shock to the economy, all P-T shocks for all Group 2 country variables demonstrated minimal initial impact mostly at less than 1%. This impact implies lesser costs for a currency union. In terms of the persistence of shocks, the duration of the effect of a P-T shock on GDP and DebtGDP (P-shocks) was about 1 (h=4) year. The dynamic impulse response for GDP is shown in Figure 2. The overall results meet our expectations, although the third transitory shock did not appear to fade as expected which may be attributed to cointegrating restrictions imposed as noted by Gonzalo and Ng (2001). However, the persistence of shocks for the CPIs (mainly the T-shocks) took between less than 1 year (h=4) to 2 years (h=8) on average, noting the case for the Solomon Islands (T-Shock 2) that took more than 2 years. Overall, the majority of the effects of the P-T shocks showed a degree of similar

dominance, less magnitude and persistence, which are necessary preconditions for a currency union. However, although there is strong convergence (cointegration) in prices (CPI) as shown earlier in Table 1, more efforts in terms of aligning market regulations for instance, is required so that prices become less persistent.

Group 2A (Melanesia + Australia): The results from the variance decomposition in Table A6 showed some evidence of similar influence in both the P and T shocks among the majority of country variables. A positive observation is the hypothetical case for a possible monetary union for Group 2A between Melanesia and Australia. Evidence show similar behaviour in the exchange rates of the Melanesian countries when fixed to the Australian dollar (anchor currency) as indicated by the TWI-AUD. Findings for the TWI-AUD showed a considerable degree of similar dominance mainly in the T-shocks. Given the highly volatile and weaker currencies of the Melanesian countries, pegging to the Australian dollar would likely bring about a degree of symmetric movement in their exchange rates (see also Table 3 and Figure A2). However, the findings regarding the dominance of either P or T shocks as key drivers for growth (real GDP) appeared mixed. This is expected given that Australian economic growth is dominated by permanent shocks that reflect mainly supply side factors such as productivity and technological innovations, in contrast to the Melanesian economies that are affected considerably by transitory shocks. Regarding the direction of the effects of the P-T shocks, our results were generally mixed, apart from the TWI-AUD that showed positive effects. Hence, the mixed results make it difficult for a common union-wide policy reaction to shocks and thus less conducive for a currency union. In terms of the magnitude of the P-T shocks, our findings showed generally low (less than 1%) impact. However, there were specific cases of relatively high magnitudes observed for a number of variables such as the first P-shock for Fiji’s real GDP that was about 3% and lasted over 3 years, while PNG’s impact was 2-3% that reduced to less than 1% after 2 years. A relatively adverse impact of 6%-7% in magnitude was observed for the third P-shock and the only T-shock for TWI-AUD among the Melanesian countries. This may have cost implications for currency union prospects. Regarding the persistence of P-T shocks, it took generally longer for most of the variables to adjust, such as the P-shocks for real GDP that persisted for 1-2 years and about 2 to 4 years for the TWI-AUD. Under such mixed conditions, where all variables are concerned, it would be costly for union. However, in terms of the similarity of the influence and the direction of shocks relating to TWI-AUD, there is some degree of benefit in pegging the currencies of the Melanesian countries to the Australian exchange rate, but of course, at some cost.

Group 2B (Melanesia + NZ): The results from Table A7 showed obvious dissimilarities in the dominating influence of either a P or T shock for the majority of variables. The findings also 15

showed a lack of cohesion in the direction of the effects of P-T shocks. In all, the dissimilarities reflect a degree of asymmetry of shocks and thus pose problems for the effectiveness of a union- wide common policy. Further, although the magnitudes of the P-T shocks were mainly low (less than 1%), the corresponding adjustment of the shocks took longer. For example, the P-T shocks for real GDP persisted for 1 to 3 years while the T shocks for real M2 for PNG and the Solomon Islands faded away after 6-7 years. These conditions imply that a union would incur huge costs. It was also noted that the variance decomposition and impulse responses could not be further analysed for the TWI-NZD for Group 2B countries based on New Zealand as the anchor currency due to poor results attributed to insignificant speed of adjustment coefficients ( γ ).31

4.4 Correlation of Dynamic Impulse Responses of P-T Shocks

Extending the analysis, the asymmetry of shocks according to the OCA theory can be assessed though the correlations of the P-T shocks where positive correlation suggests symmetry while negative correlation suggests asymmetry of shocks. In a situation where countries decide to form a currency union, this would mean adopting a common union-wide policy for all members. Engaging a common union-wide policy would be effective if shocks among member countries were symmetric in nature. If this was not the case, and countries faced mainly asymmetric shocks, then it would be beneficial not to forsake domestic policy autonomy to be part of a currency union.

Group 1 (Pacific): Mixed correlations of P-T shocks were observed for most of the variables except the CPI (see Table 2). These suggest strong asymmetric behaviour among most of the variables apart from correlations in prices (CPI). For GDP, this was expected given some differences in growth experiences among PICs. Consistent with the earlier findings in section 4.3, the outcome supports earlier arguments from studies such as Hughes (2003), Jayaraman (2004), Fichera (2006), among others, that PICs are not ready to form a currency union.

Group 2 (Melanesia): Our findings revealed evidence of positive correlations corresponding to P and T shocks, amidst some negative correlations (see Table 3). Variables such as CPI (except mainly for T-shock 1) and DebtGDP (P-shock 3 and T-shock 1) showed mainly positive correlations supporting conditions for a union. However, there was mixed correlation of shocks for other variables such as GDP and the TWI. For GDP, positive correlations were found between Fiji and PNG, and Fiji and the Solomon Islands in P-shock 1, T-shock 2 and T-shock 3

31 This was expected given the sensitivity of the cointegration results (see also Table 3). The cointegration results from the trace rank test at both 5% and 10% significance levels showed ‘full rank’, indicating stability in the levels of the series in the VAR. 16

Table 2: Correlation Coefficients of Impulse Responses - P-T Shocks: Group 1 (Pacific) S.I S.I S.I Fiji Fiji Fiji Vatu Vatu Vatu PNG PNG PNG Tonga Tonga Tonga Tonga Samoa Samoa Samoa Group 1 Group (Pacific)

(Real GDP) (CPI) (TWI) Permanent Shock 1 Permanent Shock 1 Permanent Shock 1 Fiji 1 1 1 Vatu 0.88 1 0.99 1 0.74 1 PNG 0.99 0.92 1 0.97 0.95 1 0.99 0.79 1 S.I. 0.97 0.94 0.98 1 0.83 0.87 0.77 1 -0.99 -0.81 -0.99 1 Samoa 0.99 0.91 0.99 0.97 1 0.93 0.95 0.89 0.82 1 0.42 0.92 0.50 -0.53 1 Tonga 0.92 0.97 0.95 0.97 0.93 1 0.98 0.96 0.96 0.80 0.86 1 0.99 0.76 0.99 -0.99 0.47 1 Transitory Shock 1 Permanent Shock 2 Permanent Shock 2 Fiji 1 1 1 Vatu 0.75 1 0.98 1 -0.28 1 PNG 0.61 0.75 1 0.90 0.80 1 0.99 -0.33 1 S.I. 0.57 0.42 0.14 1 0.99 0.97 0.91 1 0.99 -0.34 0.99 1 Samoa -0.1 -0.3 -0.1 -0.3 1 0.90 0.96 0.74 0.92 1 0.89 0.16 0.85 0.84 1 Tonga 0.40 0.59 0.14 0.20 -0.6 1 0.99 0.94 0.94 0.98 0.84 1 0.99 -0.40 0.99 0.99 0.82 1 Transitory Shock 2 Permanent Shock 3 Permanent Shock 3 Fiji 1 1 1 Vatu 0.02 1 0.96 1 0.32 1 PNG 0.61 0.57 1 0.97 0.99 1 -0.97 -0.26 1 S.I. -0.02 0.95 0.50 1 0.98 0.91 0.93 1 -0.99 -0.26 1.00 1 Samoa -0.04 0.98 0.58 0.89 1 0.95 0.99 0.99 0.90 1 -0.68 0.38 0.78 0.76 1 Tonga -0.33 0.92 0.37 0.85 0.95 1 0.99 0.94 0.97 0.98 0.94 1 0.95 0.28 -0.99 -0.98 -0.77 1 Transitory Shock 3 Permanent Shock 4 Permanent Shock 4 Fiji 1 1 1 Vatu -0.61 1 0.95 1 0.82 1 PNG 0.18 0.42 1 0.74 0.73 1 0.98 0.75 1 S.I. 0.04 0.70 0.49 1 0.93 0.79 0.80 1 -0.99 -0.75 -0.99 1 Samoa -0.27 -0.03 -0.7 -0.1 1 0.85 0.97 0.75 0.66 1 0.98 0.90 0.97 -0.96 1 Tonga -0.90 0.65 0.16 -0.1 0.02 1 0.94 0.81 0.77 0.98 0.66 1 0.95 0.63 0.99 -0.98 0.91 1 Transitory Shock 4 Permanent Shock 5 Transitory Shock 5 Fiji 1 1 1 Vatu -0.66 1 0.99 1 0.35 1 PNG -0.23 0.30 1 0.98 0.96 1 -0.68 0.44 1 S.I. -0.96 0.66 0.20 1 0.94 0.91 0.97 1 -0.99 -0.41 0.61 1 Samoa -0.94 0.46 0.31 0.87 1 0.99 0.99 0.94 0.89 1 0.78 0.85 -0.08 -0.82 1 Tonga -0.52 -0.20 -0.02 0.56 0.55 1 0.97 0.95 0.99 0.97 0.93 1 -0.78 0.20 0.95 0.70 -0.28 1 Transitory Shock 5 Transitory Shock 1 Transitory Shock 1 Fiji 1 1 1 Vatu -0.33 1 0.99 1 0.39 1 PNG -0.28 -0.2 1 0.95 0.90 1 0.99 0.33 1 S.I. -0.07 0.57 -0.8 1 0.99 0.99 0.92 1 -0.99 -0.38 -0.99 1 Samoa -0.75 0.73 -0.2 0.60 1 0.79 0.81 0.69 0.81 1 0.49 0.99 0.44 -0.49 1 Tonga 0.48 0.52 0.07 0.04 -0.1 1 0.93 0.88 0.97 0.90 0.81 1 0.10 -0.57 0.16 -0.11 -0.51 1 Notes: Vatu = Vanuatu; PNG = Papua New Guinea; S.I. = Solomon Islands.

respectively, while negative correlations were predominant in T-shock 1. Hence, considerations for a union will come at some cost, such as major reforms and realignment in macroeconomic policies. Moreover, since the exchange rate (TWI) shocks appeared more asymmetric, confirming the volatile nature of the exchange rates of the individual Melanesian countries, it is unlikely that a common Melanesian currency fixed on each other’s exchange rates or a member’s currency should be considered. Instead, a suggestion for an anchor currency such as the Australian dollar should be considered as evident from the strong correlation of P-T shocks of the TWI-AUD (see Table 3 and Figure A2 in the Appendix) including earlier evidence from section 4.3. This would support earlier arguments by Duncan and Xu (2000), de Brouwer (2000), Duncan (2002) in favour of a number of Melanesian countries adopting the Australian dollar. In all, our findings showed that the less synchronised P-T shocks were not prevalent in most cases to warrant any 17

Table 3: Correlation Coefficients of Impulse Responses - P-T Shocks: Group 2 (Melanesia)

S.I S.I S.I Fiji S.I. Fiji Fiji Fiji Vatu Vatu Vatu Vatu PNG PNG PNG PNG Group 2 (Melanesia) (Melanesia) (Real GDP) (CPI) (DebtGDP) (TWI-AUD) Permanent Shock 1 Permanent Shock 1 Permanent Shock 1 Permanent Shock 1 Fiji 1 1 1 1 Vatu -0.99 1 0.93 1 0.95 1 0.99 1 PNG 0.15 -0.22 1 0.98 0.97 1 -0.58 -0.79 1 0.99 0.99 1 S.I 0.86 -0.83 -0.34 1 0.95 0.77 0.88 1 -0.77 -0.91 0.96 1 0.99 0.99 0.99 1 Transitory Shock 1 Transitory Shock 1 Permanent Shock 2 Permanent Shock 2 Fiji 1 1 1 1 Vatu 0.25 1 -0.90 1 -0.98 1 0.99 1 PNG -0.99 -0.26 1 0.45 -0.19 1 0.99 -0.98 1 0.99 0.99 1 S.I 0.03 -0.52 -0.15 1 0.69 -0.73 -0.30 1 -0.99 0.98 -0.99 1 0.99 0.99 0.99 1 Transitory Shock 2 Transitory Shock 2 Permanent Shock 3 Permanent Shock 3 Fiji 1 1 1 1 Vatu -0.97 1 0.99 1 0.95 1 0.99 1 PNG 0.92 -0.83 1 0.99 0.99 1 0.97 0.85 1 0.99 0.99 1 S.I 0.19 0.05 0.46 1 0.71 0.70 0.63 1 0.94 0.79 0.99 1 0.99 0.99 0.99 1 Transitory Shock 3 Transitory Shock 3 Transitory Shock 1 Transitory Shock 1 Fiji 1 1 1 1 Vatu 0.53 1 0.21 1 0.96 1 0.99 1 PNG 0.55 -0.40 1 0.78 -0.41 1 0.10 0.09 1 0.99 0.99 1 S.I 0.78 0.88 -0.04 1 0.72 0.74 0.20 1 0.54 0.29 0.36 1 0.99 0.99 0.99 1 Notes: See Table 2. unlikely chances for a union. Although authors such as Hughes (2003), among others, have argued that such cases would warrant major reforms, prior to any union, the evidence, including findings from section 4.3 that support such arguments appear less overwhelming. Hence, the overall evidence sways in favour of the Melanesian countries forming a union. However, major policy realignment and reforms are needed. Such reforms would appear timely and draw from over two decades of regional efforts to promote economic co-operation and integration through the auspices of the Melanesian Spearhead Group (MSG) formed in 1993. These efforts have contributed to increased inter-regional trade among the Melanesian countries. For example from 1990 and 2005, respectively, Solomon Island’s imports from PNG increased from 2.8% to 4.1% and Vanuatu imports from Fiji increased from 0% to 6%. Recently, in 2008, the Secretariat of the MSG was established in Vanuatu to implement the policies of the MSG. This marks a significant era for the MSG countries to further strengthen and integrate their economies. In view of this background, Group 2 countries have a strong basis for pursuing a currency union in the future.

Group 2A (Melanesia + Australia): Our findings showed generally weak and mixed correlations among the P-T shocks for most of the variables except for the respective country’s (Group 2) exchange rates (TWI-AUD). This confirmed that P-T shocks among key macroeconomic variables were asymmetric in behaviour and predominant. Hence, it would not be feasible for Melanesian countries to form a formal currency union with Australia. However, the exceptional case was observed in the highly positive correlations among the exchange rates of the Melanesian countries based on the Australian dollar as an anchor currency. Hence, to peg to the Australian 18

dollar would be beneficial mainly in terms of the reduction in exchange rate volatility given that the Australian dollar is a stable regional currency. A number of authors such as Duncan and Xu (2000) and de Brouwer (2000) support the idea of adopting the Australian dollar. However, Melanesian countries would have to incur huge costs in terms of harmonising their monetary and fiscal policies to ensure that their policies did not deviate from the monetary policy and exchange rate goals set by Australia. This could be a huge challenge for Melanesian countries in terms of policy discipline and political support.32

Group 2B (Melanesia + NZ): There were mixed correlations of P-T shocks amongst most of the variables indicating asymmetric behaviour of shocks. It was therefore not possible for Melanesian countries to form a currency union with New Zealand. This also suggests that under the current

circumstances, adopting the NZ dollar as an anchor currency would not be feasible.

4.5 Standard Errors of the Impulse Response Functions

The Gonzalo and Ng (2001) standard errors, based on the bootstrap procedure, are derived based on the restricted model.33 The overall results for all union Groups appear efficient with minimal standard errors for the impulse response functions that ensured stability in the model consistent with the recommendation of Gonzalo and Ng (2001). However, there were a few cases where the standard errors of the P-T shocks were relatively high, attributed to the sensitivity of the cointegration restrictions imposed. For example, for Group 2A, the standard errors for CPI were relatively higher for the second and third T-shocks for all countries apart from Australia. When we tested using the cointegrating rank value, r, from the maximum eigenvalue test at r=2 (5% significance) and r=3 (10% significance), we found that using r=3 with the corresponding restrictions on γ resulted in reduced standard errors. Alternatively, this was not the case for the TWI-NZD where testing for an appropriate r resulted in an unstable process. Hence, this led to spurious estimates and so the results were excluded. This confirms the sensitivity of the appropriateness of the choice of r as stated by Gonzalo and Ng (2001) and the proposition that the correct r was required. But, in practice, this may be difficult given the choice of different cointegration tests involved, lag criteria and model specification that may be considered.

5.0 CONCLUSION

32 Adopting the Australian dollar would also have implications on the appropriate choice of exchange rate regimes. Australia has a freely floating exchange rate regime unlike most of the Melanesian countries that have fixed exchange rates systems except for PNG that adopted a freely (managed) floating exchange rate system since late 1994. 33 Both the standard errors for the restricted and unrestricted models are available upon request. 19

This analysis attempts to re-kindle the debate for the feasibility of a Pacific Islands currency union. Although Pacific Island countries continue to strengthen and revitalise their regional development strategies and agreements such as the newly developed ‘Pacific Plan’ and the recent PACER plus, the agenda for prospects of a currency union has gained little or no support at all. Nonetheless, these strategies provide a strong platform for future considerations for a Pacific or sub-regional currency union. This paper employed the optimal currency area theoretical framework and applied the Gonzalo and Ng (2001) decomposition methodology in evaluating the behaviour of the permanent and transitory shocks among key policy variables. The Gonzalo and Ng (2001) decomposition methodology is applied, for the first time, in studying currency or monetary unions. Newly-constructed quarterly time series data were also used in providing new findings to this field. The conditions for a currency union were determined by assessing the nature and extent of the (a)symmetric behaviour of permanent and transitory (P-T) shocks. This involved looking at similar dominance of either a P or T shock, direction of the effects of P-T shocks, magnitude, persistence and the correlations of P-T shocks. The degree of convergence (cointegration) was also considered. These preconditions were evaluated initially for the Pacific Island countries (PICs) (Group 1). Evidence showed that it was not feasible for the PICs to form a currency union. Further analysis was undertaken for the Melanesian countries (Group 2). Our findings showed notable evidence that the Melanesian countries had, to a reasonable degree, met necessary macroeconomic conditions for a currency union. However, more structural adjustment efforts are required in realigning and harmonising policies. Evidence also showed that a favourable choice of a currency anchor for Melanesia would be the Australian dollar. It would appear feasible for the Melanesian countries and Australia (Group 2A) to form a monetary union where common monetary policy coordination was conducted under currency board arrangements, rather than a formal currency union. The evidence for a Melanesian union supports more than two decades of efforts among the Melanesian countries in fostering economic co-operation, even beyond the establishment of the Melanesian Spearhead Group in 1993. The recent setting up of the MSG Secretariat in 2008 only strengthens such efforts for future moves towards a Melanesian union. On the other hand, a currency union of Melanesian countries with New Zealand (Group 2B) would not seem possible at this stage based on the empirical evidence.

REFERENCES

Abeysinghe, T. & Rajaguru, G. (2004). Quarterly Real GDP Estimates for China and ASEAN4 with a Forecast Evaluation. Journal of Forecasting, 23(6), 432-447.

Akaike, H. ( 1974). A New Look at the Statistical Model Identification. IEEE Transactions on Automatic Control, AC-19(6), 716-723.

Asian Development Bank (2005). Chapter 5: Priority Sectors for Regional Action: Regional Economic and Statistical Technical Assistance Facility. Toward a New Pacific Regionalism Volume II: Main Report (Ed). Manila. Asian Development Bank.

Asian Development Bank (2007). Key Indicators, 2007: Inequality in Asia. Retrieved on 31 October, 2007, from: http\\commerce\Users\Economics\wlahari\My Documents\ADB Data\Key Indicators 2007 Inequality in Asia - ADB_org.mht

Australian Agency for International Development (2008). Pacific Economic Survey. Connecting the Pacific. Canberra, ACT: Author.

Australian Parliamentary Senate Committee (2003). A Pacific Engaged. Australia’s Relations with Papua New Guinea and the Island States of the Southwest Pacific. Retrieved on 10 February, 2008, from: http://www.aph.gov.au/SEnate/committee/fadt_ctte/completed_inquiries/2002- 04/png/report/index.htm

Bowman, C. (2005). The Governor or the Sheriff? Pacific Island Nations and Dollarization (EABER Macroeconomics Working Papers 546). Canberra ACT: East Asian Bureau of Economic Research

Browne, C. & Orsmond, D. (2006). Pacific Island Countries: Possible Common Currency Arrangement (IMF Working Paper 234). Washington: International Monetary Fund.

Bunyaratavej, K. & Jayaraman, T.K. (2005). A Common Currency for the Pacific Region: A Feasibility Study (USPEC Working Paper 2005/4). Suva: University of South Pacific Department of Economics.

Chand, S. (2003). An Assessment of the Proposal for a Pacific Economic and Political Community. Pacific Economic Bulletin, 18(2), 117-124.

Chen, J. (2006). Housing Wealth and Aggregate Consumption in Sweden (Uppsala University Economics Working Paper 2006:16). Gavle, Sweden: Uppsala University Economics Department.

Cochrane, J.H. (1994a). Permanent and Transitory Components of GNP and Stock Prices. Quarterly Journal of Economics, 109(1), p. 241-265.

Cochrane, J.H. (1994b). Shocks (NBER Working Paper 4689). Cambridge, MA: National Bureau of Economic Research.

Chow, G. C. & Lin, A. (1971). Best Linear Unbiased Interpolation, Distribution, and Extrapolation of Time Series by Related Series Review of Economics and Statistics, 53(4), 372-375

Creane, S., Kim, J.I., Papi, L., & Isnawangsih, A. (2006). Options for Alternative Exchange Rate Arrangements. In C. Brown (Ed), Pacific Island Economies (pp.56-66). Washington D.C: International Monetary Fund.

De Brouwer, G. (2000). Should Pacific Island Nations Adopt the Australian Dollar? Pacific Economic Bulletin, 15(2), 161 -169.

De Veirman, E. & Dunstan, A. (2008). How do Housing Wealth, Financial Wealth and Consumption Interact? Evidence from New Zealand (RBNZ Discussion Paper DP 2008/05). Wellington: Reserve Bank of New Zealand.

Duncan, R.C. (2002). Dollarising the Solomon Islands Economy. Pacific Economic Bulletin, 17(2): 145-246.

Duncan, R.C. (2005). A Common Currency for the Pacific Island Economies. In Chand, S. (Ed), Pacific Islands Regional Integration and Governance (pp.105-118). Canberra ACT: Asia Pacific Press.

Duncan, R.C. & Xu, Xinpeng (2000). Should Papua New Guinea Adopt a Stronger Exchange Rate Regime? Pacific Economic Bulletin, 15(2): 36-45

European Union (2006). Introducing the Euro: Convergence Criteria. Retrieved on 9 Sept, 2007, from: http://europa.eu/scadplus/leg/en/lvb/l25014.htm

Fernandez, R.B. (1981). A Methodological Note on the Estimation of Time Series. Review of Economics and Statistics, 63(3), 471-476.

Fichera, V. (2006) 2005). The Pacific Islands and the Eastern Caribbean Currency Union: A Comparative Review. In C. Brown (Ed), Pacific Island Economies (pp.42-55). Washington D.C: International Monetary Fund.

Forum Economic Ministers Meeting (2000). Forum Economic Ministers Meeting, 25-16 July, 2000. Record of Meeting Decisions. Alofi, Niue.

Giancarlo, C & Konstantinou, P, T. (2009). What Drives US Foreign Borrowing? Evidence on External Adjustment to Transitory and Permanent Shocks (CEPR Discussion Paper 7134). London: Centre for Economic Policy Research.

Gonzalo, J & Ng, S (2001). A Systematic Framework for Analyzing the Dynamic Effects of Permanent and Transitory Shocks. Journal of Economic Dynamics and Control, 25(10), 1527-1546.

Grimes, A. (2005). Regional and industry cycles in Australasia: Implications for a common currency. Journal of Asian Economics, 16(3), 380-397.

Grimes, A., Bjorksten, N., Karagedikli, O., & Plantier, C. (2004). What Does the Taylor Rule Say About a New Zealand-Australia Currency Union? Economic Record, 80(1), 34-42.

Grimes, A., Holmes, F., & Bowden, R. (Eds). (2000). An ANZAC Dollar? Currency Union and Business Development. Wellington: Institute of Policy Studies.

Hafer, R.W., Kutan, A.M., & Zhou, S. (1997). Linkage in EMS term structures: evidence from common trend and transitory components. Journal of International Money and Finance, 16(4), 595- 607.

Haug, A. A. (2001). Co-Movement Towards a Currency or Monetary Union? An Empirical Study for New Zealand. Australian Economic Papers, 40(3), 307-317.

Hughes, E. (2003). Aid Has Failed the Pacific. Retrieved on 4 November 2008, from: http://www.cis.org.au/issue_analysis/IA33/ia33.pdf

International Monetary Fund (2007). Direction of Trade Statistics. Retrieved in April 2007on various days (24th and 25th) from: http://www.imfstatistics.org/DOT/

International Monetary Fund (2007). International Financial Statistics. Retrieved in April 2007 on various days (2nd, 3rd, 14th, 16th and 18th) from: http://www.imfstatistics.org/imf/imfbrowser.aspx?branch=ROOT

Jayaraman, T.K.(2001). Prospects for a Currency Union in the Pacific: A preliminary Study. Journal of Pacific Studies, 25(2), 173-202.

Jayaraman, T.K. (2003). Is there a Case for a Single Currency for the South Pacific Island Countries?. Pacific Economic Bulletin, 18(1), 41-53.

Jayaraman, T.K. (2004). A Single Currency for the Pacific Island Countries: A Stepwise Approach. Asia-Pacific Development Journal, 11(1), 91-111.

Johansen, J. (1988). Statistical Analysis of Cointegration Vectors. Journal of Economic Dynamics and Control, 12(2-3), 231-254.

Johansen, S. & K. Juselius (1990). Maximum Likelihood Estimation and Inference on Cointegration – With Applications to the Demand for Money. Oxford Bulletin of Economics and Statistics, 52(2), 169-210.

Kenen, P. (1969). Theory of Optimum Currency Areas. An Eclectic View. In R. Mundell & A.K. Swoboda (Eds.) Monetary Problems of the International Economy (pp. 41-60). Chicago: University of Chicago Press.

King, R.G., Plosser, C., Stock, J. and Watson, M. (1991). Stochastic Trends and Economic Fluctuations. American Economic Review, 81(4), 819–840.

Kishor, N.K (2007). Does Consumption Respond More to Housing Wealth Than to Financial Market Wealth? If So, Why? The Journal of Real Estate Finance and Economics, 35(4), 427-448.

Lahari, W., Haug, A. A., & Garces-Ozanne, A. (2008). Estimating Quarterly GDP for the South Pacific Island Nations (Department of Economics Discussion Paper 0805). Dunedin: University of Otago.

Lettau, M. & Ludvigson, S.C. (2004). Understanding Trend and Cycle in Asset Values: Re-evaluating the Wealth Effect on Consumption. The American Economic Review, 94(1), 276-299.

Litterman R.B. (1983). A Random Walk, Markov model for the Distribution of Time Series. Journal of Business and Economic Statistics, 1(2), 169-173.

Lu tkepohl, H., Saikkonen, P. & Trenkler, C. (2001). Maximum Eigenvalue Versus Trace Tests for the Cointegrating Rank of a VAR Process. Econometrics Journal, 4(2), 287–310.

MacDonald, R. & Taylor, M.P. (1991). Exchange Rates, Policy Convergence, and the European Monetary System. The Review of Economics and Statistics, 73(3), 553-558.

McKinnon, R.I. (1963). Optimum Currency Areas. The American Economic Review, 53(4), 717-725.

Melanesian Spearhead Group (1993). Melanesian Spearhead Group (MSG) Trade Agreement. Retrieved on 10 June, 2007, from: http://www.worldtradelaw.net/fta/agreements/msgfta.pdf

Mundell, R. (1961). A Theory of Optimum Currency Areas. The American Economic Review, 51(4), 657-665.

Podivinsky, J.M.(1992). Small Sample Properties of Tests of Linear Restrictions on Cointegrating Vectors and Their Weights. Economic Letters, 39(1), 13-18. 23

Reserve Bank of New Zealand (2008). Trade -Weighted Index. Retrieved on 25 February, 2009, from: http://www.rbnz.govt.nz/statistics/exandint/twi/index.html

Rose, A.K. (2000). One Money, One Market: The Effect of Common Currencies on Trade. Economic Policy, 15(30), 9-34.

Sawides, A. (1996). CFA Franc Zone membership and Exchange Rate Variability. Journal of African Economies, 5(1), 52-68.

Secretariat of the Pacific Community. (2003). Thirteenth Regional Meeting of Heads of Statistics, Regional Meeting of Heads of Statistics 13 List of Meeting Papers, Noumea. SPC / STATS 13 Working Paper 6. Retrieved on 8 April 2008 from: http://www.spc.int/statsen/English/News_and_Events/Stats13/REPORT.htm

Secretariat of the Pacific Community. (2005). Regional Meeting of Heads of Planning and Heads of Statistics, Planners and Statisticians Meeting, Noumea. Retrieved on 8 April 2008 from: http://www.spc.int/statsen/English/News_and_Events/RMHPS2005/RMHPS2005.htm

Secretariat of the Pacific Community. (2007). Regional Meeting of Heads of Planning and Heads of Statistics, Noumea. Retrieved on 8 April 2008 from: http://www.spc.int/sdp/index.php?option=com_content&task=view&id=25&Itemid=1

Schwarz, G. (1978). Estimating the Dimension of a Model. The Annals of Statistics, 6(2), 461-464.

Scrimgeour, D. (2001). Exchange Rate Volatility and Currency Union: Some theory and New Zealand Evidence (RBNZ Discussion Paper 2001/04). Wellington: Reserve Bank of New Zealand.

Scrimgeour, D. (2002). Exchange Rate Volatility and Currency Union: New Zealand Evidence. Journal of Policy Modelling, 24 (7-8), 739–749.

Stock, J. (1994). Unit Roots, Structural Breaks and Trends, in R. F. Engle and D. L. McFadden (Eds.) Handbook of Econometrics, vol. 4, Chapter 46, (pp. 2740-2841). Amsterdam: Elsevier Science Publishing Company.

United Nations Statistics Division (2007). National Accounts Main Aggregates Database. Retrieved on 25 May 2007 from: http://unstats.un.org/unsd/snaama/dnllist.asp

United Nations (2008). UN COMTRADE Database. Retrieved on 8 June 2008 from: http://comtrade.un.org/db/

World Bank (2008). World Development Indicators 2008 (CD-ROM). Washington, DC: World Bank.

Xu, X.(1999). The Exchange Rate Regime in Papua New Guinea? getting it right. Pacific Economic Bulletin, 14(2), 48-60

Table A1.

Variable and Data Descriptions

Variable Description Source

GDP Log of real gross domestic product (GDP) in 2000 prices. Newly constructed series by Lahari For Australia and NZ, nominal GDP from IFS is deflated et al., (2008) using modified using GDP deflator at 2000 prices. Chow-Lin (1971) method; Australia and NZ - IFS (line 99B.CZF and line 99BIRZF) CPI Log of consumer prices index (CPI) in 2000 prices IFS (line 64..ZF) IRATE Interest rates (government bonds) IFS (line 61..ZF) M2 Log of nominal money supply (M2) adjusted by CPI. IFS (line 35L..ZF) FGDP Ratio of nominal fiscal (government budget) balance to Newly constructed series using nominal GDP (in percent). Fiscal (government budget) modified Chow-Lin (1971) method balance is the difference between the nominal government as in Lahari et al. (2008).See also expenditure and revenue. Table A2; series for Australia is from ABS; NZ is from OECD. DebtGDP Ratio of nominal public external debt to nominal GDP (in Newly-constructed series using percent) modified Chow-Lin (1971) method as in Lahari et al. (2008). See also Table A2; series from Australia is from RBA; NZ is from Statistics NZ TWI Log of trade weighted nominal exchange rate index Newly-constructed series based on (March quarter, 1980, base year). The weights are based RBNZ formulae; exchange rates on the average value of total merchandise trade and IFS (line RH..ZF); nominal GDP nominal GDP from 1980 to 2006. The weighting structure IFS (line 99B..ZF and UN); trade consists of the six major trading partners (US, Japan, (UN-COMTRADE). Australia, Singapore, New Zealand and China). TWI-AUD Log of trade weighted nominal exchange rate index Newly-constructed series following (March quarter, 1980, base year) based on Australian from the RBNZ TWI formulae; dollar as anchor currency. The computation of the weights exchange rates IFS (line RH..ZF); is similar to the TWI. Here, the TWI is recomputed using nominal GDP IFS (line 99B..ZF cross rates of Australia and the major trading partners’ and UN); trade (COMTRADE). nominal exchange rates that make up the TWI basket. The country pegging to the anchor currency has a fixed exchange rate with the anchor. TWI-NZD Log of trade weighted nominal exchange rate index Newly-constructed series following (March quarter, 1980, base year. This is similar to TWI- from the RBNZ TWI formulae; AUD except that the anchor currency is the NZ dollar. exchange rates IFS (line RH..ZF); nominal GDP IFS (line 99B..ZF and UN); trade (COMTRADE).

Notes: Log represents natural logarithms; IFS = International Financial Statistics (line numbers refer to the variable code descriptions); NZ = New Zealand; OECD = Organisation of Economic Cooperation and Development; RBA = Reserve Bank of Australia; RBNZ = Reserve bank of New Zealand; UN = United Nations Statistics Division National Accounts; COMTRADE = United Nations Commodity Trade Statistics database.

Table A2 Description of annual and quarterly predictor variables used in constructing quarterly estimates for the PICs - Chow-Lin (1971) framework, with modifications by Fernandez (1981), Litterman (1983) and Abeysinghe and Rajaguru (2004) as in Lahari et al. (2008).

Variable (Annual) by PICs Quarterly (Qtly) Predictor Variables and Time series and Source ( )1 Source ( )1,2

Nominal Government Expenditure

Fiji (ADB) NGDP (Chow-Lin) and Pub.Ext.Debt (Chow-Lin) 1981:1 – 2006:4 PNG (ADB) NGDP (Chow-Lin) and Pub.Ext.Debt (Chow-Lin) 1981:1 – 2006:4 Samoa3 n.a n.a Solomon Is. (ADB) NGDP (Chow-Lin) and Claims-cent/state govt. (line 22a) 1981:1 – 2006:4 Tonga3 n.a n.a Vanuatu (ADB) GREV (Chow-Lin) and Claims-cent/state govt. (line 22a) 1981:1 – 2006:4

Nominal Government Revenue

Fiji (ADB) NGDP (Chow-Lin) and Cent. govt. deposits (line 26d) 1981:1 – 2006:4 PNG (ADB) NGDP (Chow-Lin) and Cent. govt. deposits (line 16d) 1981:1 – 2006:4 Samoa3 n.a n.a Solomon Is. (ADB) NGDP (Chow-Lin) and Cent. govt. deposits (line 16d) 1981:1 – 2006:4 Tonga3 n.a n.a Vanuatu (ADB) NGDP (Chow-Lin), Cent. govt. deposits (line16d) 1981:1 – 2006:4

Nominal Public External Debt

Fiji (ADB) NGDP (Chow-Lin), Cent. govt. deposits (line 26d) 1981:1 – 2006:4 PNG (ADB) NGDP (Chow-Lin), Cent. govt. deposits (line 26d) 1981:1 – 2006:4 Samoa (ADB) NGDP (Chow-Lin), claims-cent/state govt. (line 22a) 1981:1 – 2006:4 Solomon Is. (ADB) NGDP (Chow-Lin), claims-cent/state govt. (line 22a) 1981:1 – 2006:4 Tonga (ADB) NGDP (Chow-Lin), claims-cent/state govt. (line 22a) 1984:1 – 2006:4 Vanuatu (ADB) NGDP (Chow-Lin), Cent. govt. deposits (line 26d) 1981:1 – 2006:4 Notes: 1 Line numbers refer to the International Financial Statistics (IFS) variable code descriptions, in brackets; ADB = Asian Development Bank. 2 See, Lahari, et al. (2008) for discussion of the modified Chow-Lin method and estimation of quarterly nominal and real GDP estimates. 3 Data for Samoa and Tonga data are not available for the calendar years. All data for the variables are expressed in natural logarithms except for net-exports.

Table A3

Direction (+/ -) of Dynamic Impulse Response of P-T Shocks on initial Impact on Selected Variables by Proposed Union Groups Group 1 Real GDP CPI TWI (Pacific) P T P T P T Fiji (+) (-) (+) (+) (+) (+) (+) Vatu (+) (-) (+) (+) (+) (-) (+) (+) PNG (+) (-) (+) (+) (+) (+) (+) S.I. (+) (-) (+) (+) (+) (+) (+) Samoa (-) (-) (+) (-) (+) (+) (-) (+) (+) Tonga (+) (-) (+) (-) (+) (+) (+) (-)

Group 2 (Melanesia) Real GDP CPI TWI-AUD DebtGDP

P T P T P T P T Fiji (+) (-) (+) (+) (-) (+) (+) (+) (+) (+)

Vatu (+) (+) (+) (-) (+) (+) (+) (+) (+)

PNG (+) (-) (+) (+) (+) (+) (+) (+) (+)

S.I (+) (+) (+) (+) (+) (+) (+) (+)

Group 2A (Melanesia Real GDP CPI DebtGDP + Aust) P T P T P T Aust (-) (+) (+) (-) (-) (+) (-) (+) (+) Fiji (+) (-) (+) (+) (-) (+) (+) (+) Vatu (-) (+) (-) (+) (+) (-) (+) (+) (+) PNG (+) (-) (+) (+) (+) (-) (+) (+) S.I (+) (+) (+) (+) (-) (+) (+)

Group 2B (Melanesia Real GDP CPI Real M2 + NZ) P T P T P T NZ (-) (+) (-) (+) (+) (-) (+) (+) (+) Fiji (+) (+) (+) (-) (+) (+) (+) Vatu (-) (+) (-) (+) (+) (-) (+) (+) (+) PNG (-) (+) (-) (+) (+) (-) (+) (-) (+) (+) S.I (+) (-) (+) (+) (+) (+) (+)

Notes: Aust = Australia; NZ= New Zealand; PNG = Papua New Guinea; S.I. = Solomon Islands; Vatu = Vanuatu .

Table A4: Variance Decomposition: Pacific (Group 1)

) h Period ( Fiji Vatu PNG S.I. Samoa Tonga Fiji Vatu PNG S.I. Samoa Tonga Fiji Vatu PNG S.I. Samoa Tonga

P- Shock 1 T- Shock 1 T- Shock 2 Real GDP (r=5)

1 0.45 0.05 0.00 0.59 0.11 0.00 0.00 0.02 0.67 0.02 0.00 0.11 0.03 0.18 0.30 0.26 0.12 0.13

2 0.42 0.05 0.01 0.58 0.09 0.00 0.00 0.05 0.67 0.05 0.12 0.10 0.02 0.13 0.26 0.24 0.17 0.19

4 0.39 0.10 0.01 0.66 0.05 0.02 0.00 0.23 0.51 0.06 0.31 0.16 0.05 0.07 0.31 0.15 0.13 0.20

8 0.35 0.22 0.04 0.75 0.03 0.01 0.01 0.38 0.58 0.06 0.46 0.19 0.11 0.05 0.20 0.10 0.08 0.22

12 0.35 0.33 0.09 0.79 0.03 0.03 0.03 0.32 0.56 0.05 0.37 0.34 0.12 0.04 0.19 0.08 0.07 0.15 T- Shock 3 T- Shock 4 T- Shock 5

1 0.23 0.63 0.00 0.08 0.16 0.00 0.03 0.06 0.02 0.00 0.19 0.60 0.27 0.07 0.01 0.05 0.41 0.16

2 0.24 0.65 0.02 0.10 0.11 0.00 0.03 0.05 0.02 0.00 0.11 0.56 0.29 0.07 0.03 0.03 0.40 0.14

4 0.22 0.51 0.10 0.09 0.08 0.01 0.03 0.03 0.04 0.00 0.09 0.39 0.30 0.06 0.05 0.03 0.34 0.23

8 0.16 0.29 0.12 0.05 0.07 0.04 0.05 0.02 0.04 0.00 0.06 0.24 0.32 0.05 0.03 0.03 0.31 0.30

12 0.14 0.23 0.11 0.04 0.13 0.03 0.06 0.02 0.03 0.01 0.05 0.17 0.31 0.06 0.03 0.03 0.36 0.27 CPI (r=1) P- Shock 1 P- Shock 2 P- Shock 3

1 0.04 1.00 0.00 0.03 0.01 0.00 0.00 0.00 1.00 0.00 0.04 0.01 0.02 0.00 0.00 0.97 0.01 0.01

2 0.03 0.99 0.00 0.04 0.01 0.01 0.01 0.00 0.98 0.00 0.09 0.01 0.01 0.00 0.00 0.93 0.01 0.01

4 0.04 0.97 0.00 0.05 0.03 0.02 0.02 0.00 0.96 0.00 0.15 0.02 0.00 0.00 0.00 0.91 0.01 0.01

8 0.04 0.97 0.00 0.05 0.03 0.03 0.02 0.00 0.95 0.00 0.22 0.02 0.00 0.00 0.00 0.89 0.01 0.01

12 0.04 0.97 0.00 0.05 0.03 0.03 0.02 0.00 0.95 0.00 0.27 0.02 0.00 0.00 0.00 0.89 0.04 0.01 P- Shock 4 P- Shock 5 T- Shock 1

1 0.95 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.03 0.98 0.00 0.00 0.00 0.00 0.91 0.00

2 0.95 0.01 0.01 0.02 0.01 0.00 0.00 0.00 0.01 0.01 0.04 0.97 0.00 0.00 0.00 0.00 0.84 0.00

4 0.93 0.01 0.02 0.03 0.01 0.00 0.01 0.01 0.01 0.01 0.04 0.95 0.00 0.01 0.00 0.00 0.78 0.00

8 0.93 0.02 0.02 0.04 0.01 0.00 0.01 0.01 0.02 0.01 0.08 0.94 0.00 0.01 0.00 0.00 0.65 0.00

12 0.93 0.02 0.02 0.05 0.03 0.00 0.01 0.01 0.02 0.01 0.13 0.94 0.00 0.00 0.00 0.00 0.52 0.00 TWI (r=1) P- Shock 1 P- Shock 2 P- Shock 3

1 0.38 0.74 0.00 0.09 0.01 0.00 0.01 0.00 1.00 0.02 0.01 0.04 0.02 0.01 0.00 0.89 0.00 0.01

2 0.40 0.72 0.00 0.07 0.04 0.01 0.02 0.00 0.99 0.03 0.02 0.04 0.01 0.01 0.00 0.89 0.02 0.00

4 0.42 0.72 0.00 0.05 0.05 0.02 0.02 0.00 0.98 0.05 0.02 0.04 0.01 0.01 0.01 0.88 0.02 0.00

8 0.43 0.72 0.01 0.04 0.06 0.02 0.02 0.00 0.97 0.06 0.02 0.05 0.01 0.02 0.01 0.88 0.02 0.00

12 0.44 0.72 0.01 0.04 0.06 0.02 0.02 0.00 0.97 0.06 0.02 0.05 0.01 0.02 0.01 0.87 0.02 0.00 P- Shock 4 P- Shock 5 T- Shock 1

1 0.59 0.16 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.03 0.95 0.00 0.08 0.00 0.00 0.89 0.00

2 0.56 0.18 0.01 0.01 0.13 0.01 0.01 0.03 0.00 0.00 0.03 0.93 0.00 0.06 0.00 0.00 0.77 0.01

4 0.54 0.19 0.01 0.02 0.19 0.02 0.01 0.05 0.00 0.01 0.06 0.91 0.00 0.03 0.00 0.00 0.67 0.01

8 0.53 0.19 0.01 0.02 0.27 0.03 0.01 0.06 0.00 0.01 0.08 0.90 0.00 0.01 0.00 0.00 0.56 0.00

12 0.52 0.18 0.01 0.02 0.33 0.03 0.01 0.07 0.00 0.02 0.10 0.90 0.00 0.01 0.00 0.00 0.48 0.00

Notes: PNG = Papua New Guinea; S.I. = Solomon Islands; Vatu = Vanuatu.

Table A5: Variance Decomposition: Melanesia (Group 2)

) h Period ( Fiji Vatu PNG S.I. Fiji Vatu PNG S.I. Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

P- Shock 1 T- Shock 1 T- Shock 2 T- Shock 3 Real GDP (r=3)

1 0.03 0.01 0.02 1.00 0.03 0.00 0.94 0.00 0.27 0.76 0.02 0.00 0.68 0.24 0.02 0.00

2 0.02 0.04 0.09 0.98 0.02 0.00 0.84 0.01 0.20 0.70 0.03 0.00 0.77 0.27 0.04 0.00

4 0.01 0.13 0.15 0.99 0.01 0.00 0.77 0.01 0.12 0.52 0.03 0.00 0.85 0.35 0.06 0.00

8 0.02 0.26 0.22 0.99 0.01 0.00 0.67 0.00 0.07 0.32 0.03 0.00 0.91 0.43 0.08 0.00

12 0.02 0.32 0.27 0.99 0.01 0.00 0.60 0.00 0.05 0.23 0.02 0.00 0.93 0.45 0.11 0.00 CPI (r=3) P- Shock 1 T- Shock 1 T- Shock 2 T- Shock 3

1 0.04 0.05 0.01 1.00 0.53 0.19 0.01 0.00 0.27 0.57 0.32 0.00 0.16 0.19 0.67 0.00

2 0.03 0.09 0.01 0.98 0.51 0.14 0.00 0.01 0.30 0.51 0.28 0.01 0.16 0.26 0.71 0.01

4 0.10 0.16 0.03 0.97 0.38 0.11 0.00 0.01 0.31 0.40 0.25 0.01 0.21 0.33 0.73 0.01

8 0.33 0.29 0.06 0.97 0.18 0.09 0.00 0.00 0.20 0.24 0.21 0.01 0.30 0.38 0.73 0.02

12 0.46 0.38 0.10 0.97 0.11 0.07 0.00 0.00 0.12 0.17 0.18 0.00 0.31 0.38 0.72 0.02 TWI-AUD (r=1) P- Shock 1 P- Shock 2 P- Shock 3 T- Shock 1

1 0.00 0.04 0.02 0.00 0.06 0.07 0.10 0.11 0.25 0.26 0.26 0.33 0.69 0.63 0.62 0.57

2 0.00 0.06 0.04 0.01 0.05 0.05 0.08 0.09 0.36 0.36 0.35 0.43 0.60 0.53 0.53 0.47

4 0.01 0.13 0.09 0.03 0.14 0.12 0.17 0.19 0.37 0.36 0.35 0.42 0.48 0.40 0.40 0.36

8 0.06 0.21 0.16 0.09 0.50 0.39 0.46 0.48 0.20 0.20 0.18 0.25 0.25 0.20 0.20 0.18

12 0.08 0.24 0.19 0.11 0.68 0.54 0.61 0.64 0.11 0.11 0.10 0.14 0.13 0.11 0.11 0.10 DebtGDP (r=1) P- Shock 1 P- Shock 2 P- Shock 3 T- Shock 1

1 0.63 0.20 0.03 0.02 0.06 0.14 0.97 0.20 0.10 0.21 0.00 0.77 0.21 0.45 0.00 0.00

2 0.72 0.22 0.02 0.01 0.03 0.22 0.93 0.24 0.07 0.23 0.05 0.74 0.18 0.33 0.00 0.00

4 0.81 0.24 0.01 0.01 0.02 0.32 0.91 0.27 0.07 0.24 0.08 0.72 0.10 0.21 0.00 0.00

8 0.86 0.24 0.01 0.01 0.01 0.40 0.90 0.29 0.08 0.24 0.10 0.71 0.05 0.12 0.00 0.00

12 0.88 0.24 0.01 0.01 0.01 0.44 0.89 0.29 0.08 0.24 0.10 0.71 0.03 0.08 0.00 0.00

Notes: PNG = Papua New Guinea; S.I. = Solomon Islands; Vatu = Vanuatu.

Table A6: Variance Decomposition: Melanesia+Australia (Group 2A)

) h Period ( Period Aust Aust Fiji Vatu PNG S.I. Aust Fiji Vatu PNG S.I. Aust Fiji Vatu PNG S.I. Aust Fiji Vatu PNG S.I. Aust Fiji Vatu PNG

P- Shock 1 P- Shock 2 T- Shock 1 T- Shock 2 T- Shock 3 Real GDP (r=3)

1 0.7 0.3 0.0 0.0 0.1 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.9 0.0 0.0 0.1 0.8 0.1 0.0 0.3 0.6 0.1 0.0 4 0.7 0.3 0.0 0.0 0.0 0.0 0.0 0.3 0.3 1.0 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.6 0.1 0.0 0.3 0.7 0.1 0.0 8 0.7 0.3 0.0 0.0 0.0 0.0 0.0 0.6 0.4 1.0 0.0 0.0 0.0 0.5 0.0 0.0 0.1 0.3 0.1 0.0 0.3 0.7 0.0 0.0 12 0.7 0.3 0.1 0.0 0.0 0.0 0.0 0.7 0.4 1.0 0.0 0.0 0.0 0.5 0.0 0.0 0.1 0.2 0.1 0.0 0.3 0.7 0.0 0.0 CPI (r=4)

P- Shock 1 T- Shock 1 T- Shock 2 T- Shock 3 T- Shock 4

1 0.0 0.0 0.1 0.0 1.0 0.8 0.0 0.1 0.0 0.0 0.2 0.6 0.1 0.0 0.0 0.0 0.0 0.1 0.0 1.0 0.8 0.0 0.1 0.0 4 0.0 0.1 0.1 0.0 1.0 0.8 0.1 0.1 0.0 0.0 0.2 0.4 0.0 0.0 0.0 0.0 0.1 0.1 0.0 1.0 0.8 0.1 0.1 0.0 8 0.0 0.2 0.2 0.1 1.0 0.7 0.1 0.1 0.0 0.0 0.2 0.2 0.0 0.0 0.0 0.0 0.2 0.2 0.1 1.0 0.7 0.1 0.1 0.0 12 0.0 0.3 0.3 0.1 1.0 0.7 0.0 0.1 0.0 0.0 0.2 0.1 0.0 0.0 0.0 0.0 0.3 0.3 0.1 1.0 0.7 0.0 0.1 0.0 DebtGDP (r=1)

P- Shock 1 P- Shock 2 P- Shock 3 P- Shock 4 T- Shock 1

1 0.0 1.0 0.0 0.0 0.1 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 1.0 0.0 0.0 0.0 0.0 0.0 0.9 0.0 0.0 0.9 0.0 4 0.0 0.9 0.0 0.0 0.1 1.0 0.0 0.1 0.0 0.1 0.0 0.0 0.3 1.0 0.0 0.0 0.0 0.2 0.0 0.9 0.0 0.0 0.5 0.0 8 0.0 0.9 0.0 0.0 0.1 1.0 0.0 0.1 0.0 0.1 0.0 0.0 0.4 1.0 0.0 0.0 0.1 0.2 0.0 0.9 0.0 0.0 0.3 0.0 12 0.0 0.9 0.0 0.0 0.1 1.0 0.0 0.1 0.0 0.1 0.0 0.0 0.5 1.0 0.0 0.0 0.1 0.3 0.0 0.9 0.0 0.0 0.2 0.0

Table A7: Variance Decomposition: Melanesia+NZ (Group 2B)

) h Period ( Period NZ Fiji Vatu PNG S.I. NZ Fiji Vatu PNG S.I. NZ Fiji Vatu PNG S.I. NZ Fiji Vatu PNG S.I. NZ Fiji Vatu PNG

P- Shock 1 P- Shock 2 T- Shock 1 T- Shock 2 T- Shock 3 Real GDP (r=3)

1 0.1 0.0 0.1 0.8 0.0 0.0 0.1 0.0 0.0 1.0 0.0 0.0 0.9 0.2 0.0 0.4 0.5 0.0 0.0 0.0 0.5 0.5 0.0 0.0

4 0.4 0.1 0.2 0.8 0.1 0.0 0.1 0.2 0.0 0.9 0.0 0.0 0.6 0.1 0.0 0.2 0.5 0.1 0.0 0.0 0.3 0.3 0.1 0.0

8 0.7 0.3 0.3 0.8 0.1 0.1 0.2 0.3 0.1 0.9 0.0 0.0 0.3 0.1 0.0 0.1 0.3 0.0 0.0 0.0 0.1 0.2 0.0 0.0

12 0.7 0.4 0.3 0.8 0.1 0.2 0.2 0.4 0.1 0.9 0.0 0.0 0.2 0.1 0.0 0.1 0.3 0.0 0.0 0.0 0.1 0.2 0.0 0.0 CPI (r=4) P- Shock 1 T- Shock 1 T- Shock 2 T- Shock 3 T- Shock 4

1 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.0 0.0 0.5 0.2 0.0 0.0 0.0 0.3 0.5 0.3 0.0 0.0 0.2 0.2 0.7

4 0.0 0.1 0.2 0.0 1.0 0.9 0.0 0.0 0.0 0.0 0.1 0.3 0.2 0.0 0.0 0.0 0.4 0.4 0.2 0.0 0.0 0.2 0.3 0.7

8 0.0 0.3 0.3 0.1 1.0 0.9 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.3 0.2 0.2 0.0 0.0 0.3 0.3 0.7

12 0.0 0.5 0.4 0.1 1.0 0.9 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.2 0.2 0.2 0.0 0.0 0.3 0.3 0.7 Real M2 (r=1) P- Shock 1 P- Shock 2 P- Shock 3 P- Shock 4 T- Shock 1

1 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.7 0.0 0.0 0.1 0.2 0.3 0.0 0.0 0.3 0.4 0.0 0.0 0.0 0.6

4 0.0 1.0 0.0 0.0 0.1 0.0 0.0 1.0 0.1 0.0 0.8 0.0 0.0 0.2 0.2 0.2 0.0 0.0 0.2 0.4 0.0 0.0 0.0 0.4

8 0.0 1.0 0.0 0.2 0.2 0.0 0.0 1.0 0.2 0.1 0.8 0.0 0.0 0.3 0.1 0.2 0.0 0.0 0.1 0.4 0.0 0.0 0.0 0.2

12 0.0 1.0 0.0 0.3 0.3 0.0 0.0 1.0 0.3 0.1 0.8 0.0 0.0 0.2 0.1 0.2 0.0 0.0 0.1 0.4 0.0 0.0 0.0 0.1

Notes: Aust = Australia; NZ= New Zealand; PNG = Papua New Guinea; S.I. = Solomon Islands; Vatu = Vanuatu.

Figure A1 –Dynamic Impulse Responses for Selected Variables – Pacific (Group 1)

1.5 CPI‐ Permanent Shock 1 1.5 CPI‐Permanent Shock 2 1.0 1.0 0.5 0.5 0.0 0.0

Impulse response Impulse response ‐0.5 ‐0.5 1611 16 21 26 31 1 6 111621 26 31 ‐1.0 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

1.5 CPI‐ Permanent Shock 3 2.0 CPI‐Permanent Shock 4 1.0 1.0 0.5 0.0 0.0 Impulse response Impulse response ‐0.5 ‐1.0 1 6 111621 26 31 1611 16 21 26 31 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

1.5 CPI‐Permanent Shock 5 1.5 CPI‐Transitory Shock 1.0 1.0 0.5 0.5 0.0 0.0 Impulse response Impulse response 1 61116212631 ‐0.5 1611 16 21 26 31 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

1.5 TWI‐ Permanent Shock 1 1.5 TWI‐ Permanent Shock 2 1.0 1.0 0.5 0.5 0.0 0.0 Impulse response ‐0.5 123456789 10111213 Impulse response ‐0.5 12345678910111213 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

2.0 TWI‐ Permanent Shock 3 2.0 TWI‐ Permanent Shock 4

1.0 1.0 0.0 0.0

Impulse response 123456789 10111213 Impulse response ‐1.0 ‐1.0 12345678910111213 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

1.5 TWI‐ Permanent Shock 5 0.08 TWI‐ Transitory Shock 1 0.06 1.0 0.04 0.5 0.02 0.00 0.0 Impulse response Impulse response ‐0.02 12345678 910111213 ‐0.5 12345678910 11 12 13 14 ‐0.04 Quarters(h) Quarters(h) Fiji Vatu PNG Fiji Vatu PNG S.I. Samoa Tonga S.I. Samoa Tonga

Figure A2 –Dynamic Impulse Responses for Selected Variables – Melanesia (Group 2)

1.5 CPI‐Permenet Shock 1.0 CPI‐Transitory Shock 1 1.0 0.5 0.0 0.5 Impulse response

Impulse response ‐0.5 12345678910111213 0.0 ‐1.0 12345678910111213 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

0.3 CPI‐Transitory Shock 2 10.0 CPI‐Transitory Shock 3 0.2 5.0 0.1 0.0 0.0 Impulse response Impulse response ‐0.1 12345678910111213 ‐5.0 12345678910111213 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

6 TWI‐ AUD Permanent Shock 1 35 TWI‐ AUD Permanent Shock 2 5 30 4 25 20 3 15 2 10

Impulse response 1 Impulse response 5 0 0 ‐1 1 6 11 16 21 1 6 11 16 21 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

10 TWI‐ AUD Permanent Shock 3 8 TWI‐ AUD Transitory Shock 1 8 6 6 4 4 2 2 Impulse response Impulse response 0 0 ‐2 1 6 11 16 21 1 6 11 16 21 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

1.50 DebtGdp‐Permanent Shock 1 1.20 DebtGdp‐Permanent Shock 2 1.00 1.00 0.80 0.50 0.60 0.40 0.00 0.20 Impulse response 0.00 Impulse response ‐0.50 ‐0.20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ‐1.00 123456789101112131415 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

1.50 DebtGdp‐Permanent Shock 3 0.60 DebtGdp‐Transitory Shock 1 0.50 1.00 0.40 0.30 0.50 0.20 0.10 Impulse response Impulse response 0.00 0.00 ‐0.10 ‐0.50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 123456789101112131415 Quarters(h) Quarters(h) Fiji Vatu PNG S.I. Fiji Vatu PNG S.I.

Figure A3 –Dynamic Impulse Responses for Selected Variables – Melanesia + Australia (Group 2A)

4.0 Real GDP‐Permanent Shock 1 1.5 Real GDP‐Permanent Shock 2

1.0 2.0 0.5 0.0 Impulse response Impulse response 0.0 12345678910111213 12345678910111213 ‐2.0 ‐0.5 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

1.0 Real GDP‐Transitory Shock 1 0.6 Real GDP‐Transitory Shock 2 0.8 0.4 0.6 0.2 0.4 0.0 0.2 ‐0.2 0.0 ‐0.4 Impulse response Impulse response ‐0.2 ‐0.6 12345678910111213 12345678910111213 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

0.4 Real GDP‐Transitory Shock 3 1.2 CPI‐Permanent Shock 1 0.3 1.0 0.2 0.8 0.6 0.1 0.4 0.0 0.2 Impulse response ‐0.1 Impulse response 0.0 ‐0.2 12345678910111213 ‐0.2 12345678910111213 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

1.0 CPI‐Transitory Shock 1 1.5 CPI‐Transitory Shock 2 0.5 1.0 0.0 0.5 ‐0.5 0.0

Impulse response ‐1.0 Impulse response ‐0.5 12345678910111213 12345678910111213 ‐1.5 ‐1.0 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

2.0 CPI‐Transitory Shock 3 2.0 CPI‐Transitory Shock 4

1.5 1.5 1.0 1.0 0.5 0.5 0.0 Impulse response 0.0 Impulse response ‐0.5 ‐0.5 12345678910111213 ‐1.0 12345678910111213

Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

Cont…Figure A3 –Dynamic Impulse Responses for Selected Variables – Melanesia+Australia (Group 2A)

1.5 DebtGdp‐Permanent Shock 1 2.0 DebtGdp‐Permanent Shock 2

1.0 0.0 0.5 ‐2.0 0.0 ‐4.0 Impulse response Impulse response ‐0.5 1 2 3 4 5 6 7 8 9 101112131415 123456789101112131415 ‐6.0 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

1.5 DebtGdp‐Permanent Shock 3 2.0 DebtGdp‐Permanent Shock 4 1.5 1.0 1.0 0.5 0.5

Impulse response 0.0 Impulse response 0.0 123456789101112131415 ‐0.5 ‐0.5 1 2 3 4 5 6 7 8 9 101112131415 Quarters(h) Quarters(h) Aust Fiji Vatu Aust Fiji Vatu PNG S.I. PNG S.I.

0.14 DebtGdp‐Transitory Shock 1 0.12 0.10 0.08 0.06 0.04 0.02 Impulse response 0.00 ‐0.02 123456789101112131415 Quarters(h) Aust Fiji Vatu PNG S.I.

Notes: Aust = Australia; NZ= New Zealand; PNG = Papua New Guinea; S.I. = Solomon Islands; Vatu = Vanuatu. Impulse response functions for Group 2B (Melanesia and NZ) can be obtained from the author.