AUSTRALIA-CHINA FREE TRADE AGREEMENT: CAUSAL EMPIRICS AND POLITICAL ECONOMY

Economic Papers Vol. 26 No. 1 March 2008 pp. 19-29

AUSTRALIA-CHINA FREE TRADE AGREEMENT: CAUSAL EMPIRICS AND POLITICAL ECONOMY

TRAN VAN HOA* The launch of negotiations for an Australia-China f.ee trade agreement (ACFTA) started on 18 April 2005, following completion of the joint feasibility stu& that showed substantial economic and trade benefits for the two countries. The paper reassesses these benefits by means of an empirical analysis with a view to providing improved inputs for informed debate on the benefits and costs of an ACFTA *om the perspective of Australia and China. The implications of the findings for policy uses are also discussed.

Keywords: Australia-China Free-Trade Agreement, Trade and Growth Theory, Structural Change, Modelling Economic and Trade Policy. JEL: C30, F15,011,047 1 Introduction

The launch of negotiations for an ACFTA was agreed between Prime Minister John Howard and Premier Wen Jiabao on 18 April 2005, following completion of a joint feasibility study that showed considerable benefits for the two countries. These perceived benefits for Australia included increased trade, higher economic growth, more jobs and higher living standards, both at the national and regional level (DFAT, 2007). The official or publicly announced reasons that have been advanced to support an ACFTA are numerous. They include: China’s emergence as a major industrial power; the complementarity of the two economies; the fact that China is Australia’s secondlargest merchandise export market and her second-largest trade partner overall; China’s high growth rate; and the need for Australia to ‘stay ahead of the pack’. While the feasibility study involved commissioned independent research, business consultation, public submissions, sectoral interests and concerns, and consideration of China’s economic, trade and governance issues (DFAT, 2007), the reliability of its recommendations may be limited due to a lack of exploration of relevant historical data through rigorous quantitative analysis and policy modelling. The paper reassesses the

Centre for Strategic Economic Studies, and Vietnam and ASEAN+Plus Research Program, Victoria University, Melbourne.

19

02006. THE ECONOMIC SOCIETY OF AUSTRALIA. ISSN 08 12-0439

TRAN VAN HOA

ACFTA benefits from an empirically substantive perspective for both Australia and China, using a recent advance in aggregate economic modelling and historical data, with a view to providing improved inputs and more credible outcomes to contribute to the debate on the benefits and costs of an ACFTA for policy use. 2 China’s Recent Trade and Economic Development

Recent trends in China’s trade (as a proportion of its GDP) with its major trading partners in the world-namely, ASEAN, Japan, the United States (US), the European Union (EU) and Australia-over the past two decades are given in Figure 1. The data provide an historical perspective on Australia’s recent contribution to China’s trade and growth and its relevance to an ACFTA. FIGURE 1 REGIONAL SHARES OF CHINA ’S GLOBAL TRADEIGDP. 19862005 12

*

10

I

a

0

a

6

L

4

0

n

Y &TAGY

+TUSY

--e TOZY

2

0

+wNY

Source; Raw data from ICSEAD (2006) databases and author’s calculations Note: TJPY, TKRY, TA6Y, TEUY, TUSY, TOZY, TVNY denote China’s trade/GDP with Japan, South Korea, the ASEAN6, the EU, the US, Australia, and Vietnam, respectively.

We note from the data in the figure that China’s trade with these five partners has generally been growing over time. In 2004, China’s largest trade was with Japan, followed by the EU, the US and ASEAN.. Australia’s trade with China is small in comparison, and has risen only slowly over the sample period. In terms of its volatility or dynamics, China’s trade with Japan and the EU peaked around 1986, 1995 and the turn of the century. China’s trade with the US showed a big surge in the mid-1990s. In terms of growing importance, China’s trade with the US and the EU, which had been trailing trade with Japan in the 1980s, converged in the 2000s. While ASEAN was ranked fourth (slightly ahead of South Korea, which is not the subject of study in this paper) in trading importance with China in Figure 1, its trade has been growing steadily, especially after the 1997 Asian financial crisis. The data also show that China’s trade with its major trade partners was affected by recent developments in Asia (e.g., China’s 1989 Tiananmen Square turmoil, the 1997 Asian financial crisis, and China’s 2001 WTO membership; see Tran Van Hoa, 2000a, 2000b, and Tran Van Hoa and Harvie, 1998). For example, as a result of the Asian financial crisis, China’s trade with these partners fell by a total of 6.8% in 1998. In the case of China-Australia trade, the decline in the same year was 10.8%. These issues should be considered and investigated 20

AUSTRALIA-CHINA FREE TRADE AGREEMENT

empirically to provide improved inputs to informed debates on trade and the economic and political relations between China and its trade partners.

3 Crises, Policy Reforms, Trade and Growth in Asia: A Taxonomy The proposed ACFTA will be one of many Asian FTAs signed in recent years. The main focus and objective of the Asian FTAs (as distinct from currency or customs unions) are to promote trade and living standards either among member countries themselves or other economies outside these FTAs such as the US, Mexico, and Chile in the Americas. The currently negotiated Australia-Japan or Australia-Korea FTA is also part of this proliferation in Oceania and East Asia and has a similar focus and objective in the region. While the focus of the FTAs is important, and their objective is ‘plausible’ in a theoretical sense, there have been numerous recent developments in the region that could have impeded or enhanced this objective. These developments include: (a) national resistance to reform in order to maintain the status quo, due to either a ‘fear of the new’ or for reasons of political ideology; and (b) unexpected shocks and crises, and major structural changes and ‘good-in-a-market-economy-sense’ policy reform. The first category encompasses, for example, the Seattle and Singapore issues on agriculture and government procurements, and agricultural subsidies by the US, the EU, and Japan in the current WTO debates. The second category includes the stock market crash of 1987, the Tiananmen Square uprising in 1989, the Gulf War in 1991, the Asia economic and financial crisis of 1997, the SARS or avian flu outbreak of 2004, and the devastating tsunami shock of 26 December 2004 in the Indian Ocean. China’s major state-owned-enterprises(SOE) reforms of the early 1980s and its pro-FDI reform in the early 1990s have been two aspects of ‘good’ structural reform. The roles and impact of these shocks and policy reforms on the growth path of major economies in Asia can be seen in Figure 2. The use of annual data, while telling a compelling story about the interaction between growth and structural change, masks higher-frequency (e.g.. quarterly) movements or information about growth. An example is the impact of the SARS outbreak in early 2003 that produced two consecutive negative quarterly growth rates for that year for what was a solid economy of Singapore. We note that while the terms shocks, crises and structural changes described above are generic, the content in each case could have completely different characteristics and implications. For example, shocks usually refer to a sudden event that can have damaging effects, and structural change in the form of policy reform is often used to indicate a sudden major change in government management or governance that has been gradually developed or constructed to generate beneficial outcomes. In addition, shocks and structural change or policy reform can again have either a short-term impact or they can have lingering, non-decaying, volatile and permanent consequences (Perron, 1997; Tran Van Hoa, 2004). Shocks and structural change may also have selectively national, regional and global implications or ‘contagion’. The ‘Black Friday’ stock market crash of October 1987, for example, was considered significant chiefly for developed countries that had well-developed financial systems and linkages, while it was a minor event for the less developed countries (LDCs) in which a strong financial system was yet to be developed. The Tiananmen Square incident in 1989 may be considered a watershed point in China’s reform processes, but at the time it was only of minimal consequence to other economies in the Asian region and beyond. On the other hand, due to its status as an LDC but with a large population and economy, China’s WTO membership has been regarded as having a global effect especially on countries engaged 21

TRAN VAN HOA

in substantial trade with China or trading in markets in which China is likely to have a competitive edge in trade.

Fi

FIGURE 2 REGIONAL GROWTH, CRISES AND POLICY REFORMS IN ASIA AND AUSTRALIA, 1986200.5

-- I

-

YPH

1986 1987 19BB 1989 1990 1991 1932 1993 193 1995 1986 1

+YIH

WN

-10

-

-Y(P

Source: ICSEAD (2006). Nores: YCN, YJP, YKR, YSG, YID, YPH, YTH, YVN and YOZ denote annual growth in China, Japan, South Korea, Singapore, Indonesia, Philippines, Thailand, Vietnam and Australia, respectively.

Above, it was suggested that shocks, crises, structural change and policy reform can have a significant impact on trade, development, growth (and even welfare and poverty reduction) for a country, a region, or globally. This impact may outweigh or boost the gains for member countries from liberalised trade and investment or improved cooperation and economic relations as expected from FTAs such as the ACFTA. Unfortunately, existing methodologies or approaches that have been used almost routinely in this kind of study are either unable, or inappropriate, to accommodate this kind of impact in a realistic or historical data-consistent sense. Among these methodologies are ‘computable general equilibrium’ (CGE) models and its variations (e.g., GTAP) or extensions (e.g., the Green and Armington models; see Lloyd and Zhang, 2006), standard ‘gravity theory’ (GT) (see Frankel and Romer, 1999), and ‘panel regression’ models (see Dollar and Kraay, 2004; Eichengreen et al., 2007). These methods are severely restricted either by scope and coverage, temporal and historical features, and a lack of circular causality among economic activities. For example, the CGE/GTAP approach deals only with trade in goods and is structurally heavily calibrated, and essentially static and unable to accommodate crises in modelling (see Productivity Commission [2003]; Eichengreen et al. [2007], for other issues). The GT deals chiefly with cross-sectional data and is also unable to accommodate crises or other recent shocks or economic developments in Asia (and other regions). The popular fixedeffect ‘panel regression’ models exclude country-specific heterogeneity and interdependence between trade and growth (Tran Van Hoa, 2004). In addition, there is no a priori reason to support the assumption of a linear or log-linear fimctional form for this kind of study. In the present study, an appropriate modelling approach (namely, ‘generalised gravity theory’ [GGT]), which has improved features over existing methodologies and is adopted to deal more flexibly with the concept of trade-growth circular causality (endogeneity) and comprehensive trade, that is, trade in goods, FDI and services. The GGT is also able to accommodate more realistically and more accurately the impact of 22

AUSTRALIA-CHmA FREE TRADE AGREEMENT

the so-called Johansen (1982) add- and sub-factors in policy modelling on trade, development, growth, and economic relations for Australia and China. Examples of the add-factors are ‘good’ policy reform and WTO membership, and the sub-factors include natural disasters, domestic, regional and global crises. The causal empirics are then critically compared to the outcomes of the political economy of the feasibility study to provide a better understanding of the benefits and costs of an ACFTA. 4 A Model to Study the Determination of Trade and Growth in Australia and China The development and features of the GGT model have been described more fully elsewhere (see Tran Van Hoa, 2002a, 2004), but they can be briefly described as follows. We consider, for convenience and without loss of generality, a simple generic model of two simultaneous implicit functions comprising and extending the basic postulates and determinants of the standard cross-sectional-data GT linking trade and growth between two trading countries (see Frankel and Romer, 1999). These variables comprise not only the GT’s geographic or demographic attributes for the countries under consideration (i.e., China and Australia), but also, significantly, economic and non-economic factors such as fiscal, monetary and industry policy, crises, and reforms of a regional FTA. Since the geographical attributes such as distance and area in the case of the ACFTA are assumed apriori to be a rationale for setting up this FTA, we can then focus on other relevant demographic, economic and non-economic determinants of trade and growth in our model. In Tran Van Hoa (2002a, 2004), it was demonstrated that this model from its original arbitrary functions can be written equivalently for empirical implementation as a linear structural (1) and a reduced-form (2) equation:

AY=al+ a2A T+ a3A FDI+a4A SV+a5ST+ A T=~cI+ 7 ~A 2 YT+ ~3 A FT+ 7 ~ A 4 MT+x5 A PT+x6 A ERT+x, A IT+IQ A POT +x~ST+EZ

(1) (2)

where Y, T, FDI, SV, ST, YT, FT, MT, PT, ERT, IT, POT are, respectively, China’s GDP, trade, FDI, services, shocks, Australia’s GDP, fiscal policy, monetary policy, inflation, exchange rates, industry policy and population. The a’s and n’s are parameters; the E ’ S are error terms; and A denotes the rate of change. In Equations 1 and 2, in addition to China’s trade with Australia, FDI, services, crises or policy reform, are also assumed to affect China’s growth. But this trade (and services and FDI) is also affected by economic activities, trade-related policies and external or internal shocks (the ‘conditionality’ factors) in China and Australia. Assuming that YT captures Australia’s ‘conditionality’ factors, then Equation 2 simply postulates that China’s trade with Australia is affected by Australia’s GDP, China’s inflation, fiscal policy, monetary policy, trade policy and exchange rates, industry structure, population, and internal or external shocks or policy reform. Major relevant structural changes that can be tested for their effect on China’s trade and growth (see Figure 2) include the 1987 stock market crash, 1989 Tiananmen Square turmoil, the 1991 Gulf War, China’s 1993 reforms, the 1997 Asia crisis, and China’s 2001 WTO membership. In the GGT model, all variables in the model are expressed as their rates of change in consequence of which the units of measurement for the trading countries’ variables are irrelevant. ST is a qualitative time-series variable with discrete values representing internal or external 23

TRAN VAN HOA

shocks and policy reform having either one-off effects or temporally long-term effects on trade and growth For Australia’s trade, the arguments justifying the model are similar. 5

Substantive Evidence on Expected Benefits and Costs for ACFTA Members

5.1 Data Data, at current prices, were retrieved from the ICSEAD (2006) databases for China and the ABS databanks for Australia. Due to the limitation of the required harmonised data in our studies, all original data are obtained as annual figures and then transformed to their ratios (when appropriate). The ratio variables include: trade (exports and imports), services, FDI, government budget, and money supply (M2) (all divided by GDP), and unemployment rates (open unemploymentAabour force). Other non-ratio variables include RMB/US exchange rates, population and binary variables representing the occurrence of economic, financial and other major crises or policy shifts or reforms over the period 1986 to 2005 that have affected China and Australia. All non-binary variables are then converted to their percentage rates of change. The use of this percentage measurement is a main feature of our modelling and impact approach and avoids the problem of a priori assumed functional forms as well as logarithmic transformations for negative data (such as budget [fiscal] or current account deficits).

5.2. The Estimated Models The trade-growth Equation 1 from the perspective of China and Australia has been estimated with all available data for the period 1986 to 2004 by OLS and 2SLS. When OLS is applied, the effect of trade-growth circular causality is by implication excluded from the equation. This equation can then be regarded as a standard growth regression. The use of 2SLS implies that trade-growth simultaneity is incorporated in the equation. The results for both types of estimation for China and Australia are reported in Table 1. 5.3

Statistical Reliability Interpretation

From the results given in Table 1, we note five important findings. First, both estimated growth models incorporating trade in goods, services, and FDI between China and Australia are statistically significant using the F-test and explain a high level of variance relative to other trade-growth models as reported in previous international studies (see, for example, Frankel and Romer [ 19991). Second, when we look at the dynamic features of the estimated models using either plots or standard diagnostic tests, all estimated models appear free from serious first- or higher-order autocorrelation-inducedor simple Markov scheme inefficiency problems. Third, trade, as defined by total trade/GDP between China and Australia, has a positive but weak impact on China’s growth while having a strongly positive effect on Australia’s economic performance. This finding may be explained partly by the fact that while Australia-China trade/GDP is high from the Australian perspective, its intensity is low from the Chinese trade balance sheet (see also Figure 1). Fourth, the introduction of financial services and FDI into the models shows a stark contrast in magnitude between the impact of trade, services and FDI. More specifically, while China’s goods and services flows have a small, beneficial, and statistically weak effect on China’s growth, FDI is found on the contrary to have a positive and highly significant impact on this growth. From Australia’s perspective, 24

AUSTRALIA-CHINA FREE TRADE AGREEMENT

trade with China is a highly significant contribution to Australia’s growth, thus supporting the official stand on the rationale of an ACFTA. Finally, the introduction of crises, shocks or major policy reform into the models provides substantive evidence about the characteristics of these crises, shocks, or policy reforms, and the roles they have played in, or contributed to, the economic performance of China and Australia in recent years. TABLE 1 IMPACT OF TRADE, SERVICES, FDI AND CRISES ONACFTA GROWTH. GENERALISED GRA VITY THEORY IN FLEXIBLE STRUCTURAL FORM, 19862004 Variable

Constant Openness/GDP ServicedGDP FDVGDP Stock Market Crash, 1987 Tiananmen Turmoil, 1989 Gulf War 1991 China Reforms 1993 Asia Crisis, 1997 WTO Membership R2 F DW - ..

Chinese Perspective OLS: Growth Regression 11.16** 0.02 0.002* 0.04**

2SLS: GGT Structural 11.07** 0.03 0.003 0.04**

Australian Perspective OLS: Growth Regression 7.43** 0.11** 0.003 0.001* 8.45**

2SLS: GGT Structural 4.91** 0.22** 0.01* 0.001* 13.45**

-7.45**

-7.47**

-7.11**

-10.04**

6.04** 4.25

5.81** 4 . 71

-7.57* * 3.89**

-8.44** 4,56**

4.97 0.72

4.15 0.62

-1.10

0.92 13.62** 2.71

0.91 6.98** 2.65

-3.14

0.90 11.07** 2.21

0.87 7.88** 2.40

Sources: World Bank (2005) ;ICSEAD (2006) Trade Data. Nores: ** designates significance at the 5% level; designates significanceat the 10% level.

6

Causal Empirics and Implications for ACFTA Political Economy

While the GGT models we used may be simple in their structure, they contain the main ingredients of analysis on causal trade-growth relationships for China and Australia. They are also fairly consistent, for comparative purposes, with similar previous studies of a different more limited kind (e.g., the CGE/GTAP, GT, ‘panel regression’ or other quantitative trade-growth studies). The causal empirics reported in the preceding section also provide a number of interesting insights into bilateral trade-growth causation. Finally, the findings are seen as substantive evidence providing empirical support, at least from Australia’s perspective, for the recent ACFTA initiative and its political economy arguments at the highest political level in the two countries.

25

TRAN VAN HOA

6.1 Does Australia-China Trade Contribute to Australia’s and China’s Growth? Trade-to-growth is an important causality topic in economics that has attracted much attention over the last fifteen years or so (see e.g. Frankel and Romer [1999], for a survey). There are no robust conclusions, especially in the short run for even comparative static and strictly calibrated neo-classical models (see Rees and Tyers, 2004). Our empirical results above show that, in the specific case of China and Australia, there is only weak empirical support for the view that Australia-China trade (when defined as the relative size of openness to its GDP) has a statistically significant and beneficial impact on China’s growth. A partial explanation given above for this weak evidence lies in the relative small size of Australia-China trade to China’s GDP during the period under study (see Guttmann and Richards [2004], for similar evidence on the significance of Australia’s trade on its GDP). My empirical results show, however, that Australia’s growth seems to gain substantially from bilateral trade with China. This effect would give Australia a strong incentive to negotiate an ACFTA.

6.2 Do Financial Services and FDI Impact on Australia’s and China’s Growth?

As mentioned earlier, one of the novel features of the present paper is the introduction of comprehensive trade into the GGT models. The results reported in Table 1 suggest that the effects of service flows to China and Australia are beneficial but weakly significant. In contrast, the important role of FDI on China’s growth can be more robustly ascertained from Table 1 where FDI is seen to be statistically significantly and positively impacting on China’s growth. This impact is, however, statistically weak from Australia’s perspective. This evidence would support the objective of an ACFTA where not only trade in goods is traditionally important but also, and especially, trade in FDI (a major issue of scope and coverage in ACFTA negotiations) would have a more critical role in the economic performance not only of China but also Australia. An implication is that negotiations on FDI liberalisation and promotion under an ACFTA can now assume, with empirical support, a more prominent part in the feasibility and scoping agenda. The evidence also supports the introduction of an effective competition law and policy among the trading partners within this framework to promote international trade, services and FDI for mutual benefit. 6.3 Do Crises and Economic Policy Reform Affect Australia’s and China’s Growth? The specification of shocks, crises and policy reform, either of the sudden or gradual kind and with temporary or long-lasting effects, in our GGT models is one of their significant modelling features (see Edwards [2007] for a recent use). This feature has not been captured adequately by existing methodologies. The shocks have, by definition, the characteristics of a sudden change, while China’s economic policy reforms and WTO membership are assumed, by nature of their implementation or negotiations, to have the feature of a gradual change. Our modelling experiments suggest all shocks and policy reforms have non-decaying long-term effects on trade and growth in Australia and China for the period under study. From Table 1, it appears that the Tiananmen Square incident of 1989, which was considered a minor event by many outside China as far as its impact on world economies was concerned, was a statistically significant impediment with huge damage to China’s growth. In fact, this impact has been the largest (or the most serious) among the other economic (trade in goods, services and investment) and non-economic (major

26

AUSTRALIA-CHINA FREE TRADE AGREEMENT

reform policies) determinants of China’s growth incorporated into the models. For China’s policy reforms, which are regarded as involving gradual change and were introduced over a number of years, the implications for China’s growth are remarkable. First, the Gulf War in 1991 coincided with China’s ‘Rectification Program’ of 19881991 in which fixed investment and production were increased even though these were associated with a sharp rise in inflation and in retrenchment measures, especially in the state-owned enterprise and ‘town-village-enterprise’ ( W E ) sectors. From the perspective of developed countries (e.g., Australia) involved in the First Gulf War, the subsequent consequence of this war would be expected to be negative through the crowding-out effect of available resources. But from China’s perspective, the total benefits of its more relevant 1998-1991 program can be interpreted as policy reform in the above context on China’s development and growth. These benefits are significantly captured by our GGT models. Second, from the findings for China’s 1993 reform which reflects the outcomes of the deepening period (1992-2002 based on our data, or 1992present in an economic policy management context) however, our GGT model of tradegrowth for China does not provide results that lend support to the expected benefits of this reform policy during this period. Third, our study shows for the first time empirical support (albeit statistically weak) for the claim that WTO membership has benefited China not only through increasing total trade as official statistics have also indicated but also through other aspects of domestic economic, political and social reforms that China has carried out as a result of its WTO membership and obligations. 6.4 Implications for Trade and Investment Policy in Australia and China

Our empirical findings as given in Table 1 appear to support the view that China’s merchandise trade (i.e., exports and imports of goods), while being considered widely as the most important element of the WTO and all other FTAs (or closer economic relations) by the media and in political debates, constitutes empirically only a negligible contribution to China’s growth. In contrast, FDI is uniformly seen as the most important driver of China’s economic performance over the period under study. The implication is that an overemphasis on WTO-focused trade and trade liberalisation policy and less focus on FDI and services and their promotion policies (e.g., in an ACFTA), while being fashionable in many quarters among WTO advocates, is inappropriate and misguided for China in this case. However, the opposite is true for Australia.

6.5 Implicationsfor China’s Regional FTA Strategy The findings above lead us afortiori to the conclusion that while WTO membership is a prize that many countries have been attempting over many years to get, its benefits are small compared to those attributable to FDI and perhaps to regional co-operation. Since FDI and regional co-operation are major objectives of regional bilateral and multilateral FTAs, China should focus more on these FTAs as its strategic alliances. In this context, an ASEAN + China FTA is not only logically appropriate from a ‘regional’, ‘gravity’ or Asian perspective, but also is empirically supportable from a research-based policy perspective.

7 Conclusion: Do We Have Empirical Support for the Bilateral ACFTA? After the successful negotiations and signing of the Australia-Singapore FTA and the Australia-Thailand FTA in 2003, and the Australia-US FTA early in 2004, the 27

TRAN VAN HOA

Australian Government has proposed official dialogue and negotiations with a view to set up further bilateral FTAs with countries in the region. In addition to the proposed ACFTA, Australia is planning negotiations with Japan, South Korea, and even the Emirates. Given the slow progress of the Doha agreements on WTO agendas, the subsequent dismal outcomes from the WTO Seattle, Cancun and Hong Kong Ministerial Meetings, and growing concerns from the new Singapore Issues, these bilateral FTA proposals are seen as natural developments within a ‘big picture’ (WTO, GATT and GATS) framework and as logical extensions of the current new Asian regionalism. As mentioned earlier, the objectives of setting up an FTA are, in addition to better regional co-operation, political stability and security, to enhance comprehensive trade between the member countries, to improve their growth, development and ultimately welfare, and to reduce income inequality. These objectives necessarily require the validity of the premise that trade in goods, FDI and services do in fact directly and positively affect growth in these FTA members. The findings reported above lend support to the hypothesis that China-Australia trade has been small and has had hardly any impact on China’s growth. The reverse is true for Australia, a small and open economy with large trade with China. The implications of our findings are that for China an ACFTA would serve less as an engine of growth but more as a vehicle for regional political cooperation (or exertion of influence). From Australia’s perspective, its trade with China has been growing rapidly and is making a substantial contribution to Australia’s economic growth. However, the claims that Australia can substantially benefit from more FDI and service exports to China within an ACFTA do not seem to be strongly supported by the empirical evidence.

REFERENCES Baier, S. L. and J. H. Bersgtrand (2007) “Bonus Vetus OLS: A Simple Method for Approximating International Trade-Cost Effects using the Gravity Equation”, Working Paper, Department of Economics, Clemson University, USA. Barro, G. and E. Helpman (1991) “Trade, Knowledge Spillovers and Growth”, European Economic Review, 35, pp. 5 17-536. Coe, D. and E. Helpman (1993) “International R&D Spillovers”, National Bureau of Economic Research Working Paper, Cambridge, MA, No. 4444, August. DFAT (Department of Foreign Affairs and Trade) (2007) http ://www.dfat.gov.adtrade/ ftas.html/. Dollar, D. and A. Kraay (2004) “Trade, Growth, and Poverty”, Economic Journal, 114. Edwards, S. (2007) “Crises and Growth: A Latin American Perspective”, National Bureau of Economic Research Working paper, April. Eichengreen, B., Y. Rhee and H. Tong (2007) “China and the Exports of other Asian Countries”, Review of World Economics, 143, pp. 201-226. Guttmann, S. and A. Richards (2004) “Trade Openness: An Australian Perspective”, (Canberra: Reserve Bank of Australia, Discussion Paper No. 2004-1 1). Frankel, J. A. and D. Romer (1 999) “Does Trade Cause Growth?” American Economic Review, 89, pp. 379-399. Harvie, C. (1999) “Economic Transition: What Can be Learned from China’s Experience”, International Journal of Social Economics, 26, pp. 109 1- 1 1 19. ICSEAD (International Center for the Study of East Asian Development) (2006) httr,://www.icsead.or.iu/indexe.htm1. 28

AUSTRALIA-CHINA FREE TRADE AGREEMENT

Johansen, L. (1982) “Econometric Models and Economic Planning and Policy: Some Trends and Problems”, in M. Hazewinkle and A. H. G. Rinnooy Kan (eds), Current Developments in the Interface: Economics, Econometrics, Mathematics (Boston: Reidel). Kydland, F. E. (2006) “Quantitative Aggregate Economics”, American Economic Review, 96, pp. 1373-1383. Lloyd, P. J. and X. G. Zhang (2006) The Armington Model, Productivity Commission, Staff Working Paper, Melbourne, January. Otto, G., G. Voss and L. Willard (2002) “Understanding OECD Output Correlations”, Seminar Paper, Department of Economics, University of Wollongong, May. Perron, P. (1997) “Further Evidence on Breaking Trend Functions in Macroeconomic Variables”, Journal of Econometrics, 80, pp. 355-385. Productivity Commission (2003) http ://www.pc .gov.adresearchproj ect/200 1I0 10202.html . Rose, A. K. (2000) “One Money, One Market: The Effect of Common Currencies on Trade”, Economic Policy, 30, pp. 9-33. Rees, L. and R. Tyers (2004) “On the Robustness of Short Run Gains from Trade Reform”, CEPR Discussion Paper No. 474, (Canberra: Australian National University). Tran Van Hoa (ed.) (2000a) The Asia Crisis: The Cures, Their Efectiveness and the Prospects After (London: Macmillan). Tran Van Hoa (ed.) (2000b) The Social Impact of the Asia Crisis (London: Macmillan). Tran Van Hoa (2002a) “Korea, China and Japan’s Trade with the World and Its Impact on New Asian Regionalism ASEAN+3”, Paper presented at the Joint Conference of the Association of the Korean Economic Studies, Korea Development Institute, and the Research Center for International Economics, Yonsei University, Seoul, 2 1-22 July. Tran Van Hoa (ed.) (2002b) Economic Crisis Management (Cambridge MA: Edward Elgar). Tran Van Hoa (2004) “Korea’s Trade, Growth of Trade and the World Economy in Post-crisis ASEAN+3 Free Trade Agreement: An Econometric and Policy Analysis”, Journal of the Korean Economy, 5 , pp. 73-108. Tran Van Hoa and C. Harvie (1998) (eds), Causes and Impact of the Asian Financial Crisis (London: Macmillan). World Bank (2005) World Tables (Washington DC: World Bank).

29