Asean1

Available online at www.sciencedirect.com

Journal of Asian Economics 18 (2007) 974–987

Asia/Pacific Regional Trade Agreements: An empirical study Bhavish Jugurnath a, Mark Stewart a,*, Robert Brooks b b

a RMIT University, School of Economics, Finance and Marketing, Melbourne 3000, Australia Department of Econometrics and Business Statistics, Monash University, Caulfield East 3145, Australia

Received 23 January 2007; received in revised form 10 September 2007; accepted 28 September 2007

Abstract At the same time as the General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO) have been encouraging trade liberalized, there has been a proliferation of Regional Trade Agreements (RTAs). These RTAs also aim to reduce trade barriers, but they do so it in a preferential way. There is continued debate as to whether such RTAs are an effective way of achieving free trade, or if increased trade among members causes less trade with non-member countries? If RTAs increase total trade, this is known as ‘trade creation’, whereas if the extra trade occurs at the expense of non-members, this is called ‘trade diversion’. Trade creation implies improved welfare, whereas ‘trade diversion’ may adversely affect welfare. This paper examines five different RTAs using a gravity model to see if they have been trade creating or trade diverting. Annual data from 26 countries covering five RTAs in the Asia and Pacific region for the years 1980–2000 was used. The results show that the effects of the different RTAs varied remarkably. The Association of South East Asian Nations (ASEAN) and the Australian and New Zealand Closer Economic Relations (CER) fostered greater trade with trading partners and with the rest of the world. While the Asian Pacific Economic Cooperation (APEC), the Southern Cone Common Market (MERCOSUR) and the North American Free Trade Association (NAFTA) tended to be trade diverting, that is, they expanded intra-bloc trade at the expense of trade with others. # 2007 Elsevier Inc. All rights reserved. JEL classification: F1; F14; F15 Keywords: Intra-regional trade; Regional trade agreements; Trade creation and trade diversion

* Corresponding author at: RMIT University, Building 108, Level 12, 239 Bourke Street, GPO Box 2476V, Melbourne, VIC 3001, Australia. Tel.: +61 3 9925 5879; fax: +61 3 9925 5986. E-mail address: [email protected] (M. Stewart). 1049-0078/$ – see front matter # 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.asieco.2007.09.003

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

975

1. Introduction This paper addresses the question of whether Regional Trade Agreements (RTAs) enhance welfare. This is examined by using a gravity model to analyse the effect the Association of South East Asian Nations (ASEAN),1 Australian and New Zealand Closer Economic Relations (CER), the Asian Pacific Economic Cooperation (APEC), the Southern Cone Common Market (MERCOSUR) and the North American Free Trade Association (NAFTA) have had on the trade of both members and non-members of these RTAs. In the previous literature the gravity model is usually used to examine only one or two RTAs at a time. This study is unusual in that a comparable method has been used to look at five different RTAs. ‘Trade creation’ occurs when the introduction of an RTA allows an importing country to purchase products at lower cost than was previously the case. In contrast, ‘trade diversion’ is the substitution of a more costly source of supply within an RTA for a less costly source outside. As the introduction of an RTA will generally have both ‘trade creation’ and ‘trade diversion’ effects, it is the net affect that needs to be assessed when deciding whether an RTA hinders or enhances welfare. This paper tests for the existence of ‘trade creation’ and ‘trade diversion’ as a result of five RTAs by using dummy variables within the context of a gravity model. Although the standard gravity model is extended in this paper via the introduction of tax and exchange rate variables (something not often done in previous studies). The rest of the paper is structured as follows. Section 2 describes the analysis of RTAs. Section 3 provides a description of the data used, as well as explaining the estimation procedures. Section 4 reports and discusses the empirical findings, and finally Section 5 provides some concluding remarks.

2. Regional Trade Associations Initially economists saw Regional Trade Agreements (RTAs) as welfare enhancing, as they were a step toward free trade. That is, as long as an RTA did not increase trade barriers to nonmembers they were thought to improve welfare. However, Viner’s (1950) paper changed this idea by noting that RTAs lead to both ‘trade creation’ and ‘trade diversion’. Trade creation occurs when the establishment of an RTA allows an importing country to purchase products at lower cost than was previously the case. Clearly this benefits both the importing country and the world as a whole. In contrast, ‘trade diversion’ is the substitution of a more costly source of supply within the RTA for a less costly source outside, and this would negatively affect welfare. As an RTA will generally have both ‘trade creation’ and ‘trade diversion’ effects, it is the net affect that needs to be assessed when deciding whether it enhances welfare. Johnson (1960) developed a partial equilibrium diagram that explained the economic effects of ‘trade division’ and ‘trade creation’.2 The diagram can be used to show that the affect of an RTAwill be the sum of several effects, and that in markets where trade is diverted countries may be better or worse off. More recently computable general equilibrium models have been used, and the results using these techniques confirm Johnson’s conclusions. Lloyd and MacLaren (2004) have an

1

The ASEAN RTA is sometimes known as the Asian Free Trade Association (AFTA). Pomfret (1997) and Schiff and Winters (2003) both have very good expositions of Johnson’s partial equilibrium diagram. 2

976

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

excellent survey article that summaries the developments in the theoretical analysis of RTAs, while Low (2003) discusses some of the practical issues relating to RTAs.3 In terms of the trade organizations analysed in this paper, certainly CER, MERCOSUR and NAFTA could be considered to be conventional RTAs. That is, they have been established to reduce trade protection among member states. However, APEC and ASEAN may be thought of as being a little different. APEC was built on the concept of ‘‘open regionalism’’, which means that any member reducing its trade barriers to other members should offer the same reductions to non-members. The ASEAN Free Trade Association (AFTA) also appears quite similar in that many of its trade concessions have also been multilateralized. 3. Methodology 3.1. Gravity models The most common empirical tool used to estimate the effects of RTAs is a gravity model. A gravity model involves regressing trade on a series of explanatory variables, then using dummy variables to ascertain whether this relationship is affected by the existence of RTAs. Beckerman (1956), Anderson (1979), Bergstrand (1989), Oguledo and Macphee (1994), Frankel and Wei (1995), Kreinin and Plummer (1998), Krueger (1999), Cernat (2001), Havemann, Nair-Reichert, and Thursby (2003), Adams, Dee, Gali, and McGuire (2003), Filippini and Molini (2003) and Tang (2005) are just some of the studies which have used gravity models to estimate the ‘trade creation’ and ‘trade diversion’ effects of various RTAs, including NAFTA, MERCOSUR, CER and ASEAN. Generally, studies on regional trading blocs find that trade volume is directly related to the economic and physical size (GDP, population, land area) of the countries involved, as well as transaction costs which are usually proxied by such things as distance and cultural similarities (a common language is often used for this). This paper examines these factors, as well as adding the cost variables of the exchange rate and taxes to the list. Tinbergen (1962) and Linnermann (1966) provide initial specifications for the gravity model and use it to look at the determinants of trade flows, while Aitken (1973) was one of the first to apply this approach to analysing RTAs. Others to have done this include Bayoumi and Eichengreen (1997) and Frankel (1997), both of whom examined the effect RTAs had on nonmembers as well as members. That is, these papers tried to separate the ‘trade creation’ and ‘trade diversion’ effects of RTAs. 3.2. Dependent variables

3.2.1. Imports Although gravity models typically employ total trade (imports plus exports) as the dependent variable, this paper focuses on imports as they more closely proxy the effects of domestic trade barriers.4 3 The Low paper is of particular relevance as it uses APEC and ASEAN as examples (both of these RTAs are analysed in this paper). 4 Appendix A provides a detailed description of the variables used in his paper, including data sources.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

977

3.3. Independent variables 3.3.1. Gross Domestic Product, GDP The model includes two GDP variables; GDPi is for the importing country and GDPj is for the exporting country. As income and output in each country increase, there would be both greater demand for goods and services as well as increased production, therefore positive relationships between both of these variables and imports to country i would be expected.5 3.3.2. Population, POP Again there are two population variables, POPi for the importing country and POPj for the exporter. It is anticipated that countries with larger populations will both import and export more. However, as was suggested by Aitken (1973), the larger is a country’s population the larger will be the ratio of the domestic market to the foreign market, hence the smaller would be the potential export supply. Nevertheless, Bergstrand (1989) pointed out an inconsistency in this argument. A larger population would allow for economies of scale, which may increase the price competitiveness of the export country’s production, thereby leading to higher exports. Therefore, the sign on the coefficient of the population of the exporting country (POPj) may be indeterminate, while the sign for the importing country (POPi) is expected to be positive. 3.3.3. Distance, DIST The physical distance between trading countries is a proxy for transport costs. It is expected that transport costs would be negatively correlated with trade. DISTij is the geographical distance between the capital cities of the importing country i and the exporting country j. 3.3.4. Surface area, AREA This is a country’s total land area, including areas under inland bodies of water and some coastal waterways. AREAi is the surface area of the importing country, while AREAj is for the exporting country. Generally, it is expected that larger countries will both export and import more. However, it is possible that relative size may also be important for comparative advantage reasons. This being the case the sign on the AREA coefficients may also be indeterminate. 3.3.5. Exchange rate, EXR EXRi is the exchange rate of the importing country, while EXRj is for the exporter. This paper uses a similar approach to that of Soloaga and Winters (2001) when measuring this variable. EXR is defined as the local currency value of one $US multiplied by the US GDP deflator divided by the GDP deflator of the country in question. For country i an increase in EXRi would indicate either a depreciation of the local currency or a fall that country’s relative prices, as such a negative coefficient would be expected. For country j a positive coefficient would be expected. 3.3.6. Taxes on goods and services, TAX TAXi is the International Monetary Fund (IMF) trade tax index for the importing country, while TAXj is the equivalent index for the exporting country. As all taxes (other than lump sum taxes) are distorting, a negative relationship would be expected for both of these variables and 5

Note that through out the analysis, i refers to the importing country, while j is the source of these imports (the exporting country).

978

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

imports to country i. This variable aims to capture the trade regimes of the countries involved. The IMF trade tax index is the best variable to use for this purpose as it encapsulates many different aspects of government intervention in trade. 3.3.7. Language, LANG As stated by Linnermann (1966) a common language between two countries will influence the volume of trade. Therefore LANGij was assigned a value of one if the two trading countries had the same official language and zero otherwise. The hypothesis is that countries will trade more when they have a common language, so a positive coefficient is expected. 3.4. Model specification Eq. (1) shows how this paper uses the above variables in the basic gravity model. log IMPORTijt ¼ a0 þ a1 log GDPit þ a2 log GDPjt þ a3 log POPit þ a4 log POPjt þ a5 log DISTij þ a6 log AREAi þ a7 log AREAj þ a8 log EXRit þ a9 log EXRjt þ a10 TAXit þ a11 TAXjt þ a12 LANGij þ mijt

(1)

A log specification was used as this has typically given the best results and i is the importing country, while j is the exporting country. To examine the effects of RTAs within the framework of this equation, dummy variables were then added. When others have used this technique a consensus has emerged that RTAs are generally trade creating. For example, Aitken (1973), Bergstrand (1985) and Thursby and Thursby (1987) showed that the European trading blocs increased trade during the 1960s and 1970s. Later, work by Frankel and Wei (1995) and Frankel (1997) found evidence of ‘trade creation’ in Asian and in North American trading blocs. While a recent paper by Rose (2000) also found that regional trade arrangements, in general, were trade creating. However, Hassan (2001) found that the South Asian Association for Regional Cooperation (SAARC)6 countries as a whole traded less with the outside world than would be expected. To look at the effects of RTAs the basic gravity model outlined in Eq. (1) was extended using dummy variables. Table 1 shows the list of included RTAs, their members and starting years. The dummy variables took the value of one if a country was a member of an RTA and zero otherwise, regardless of the membership status of the trading partners. Eq. (2) was used for this. log IMPORTij ¼ the above þ

n n n X X X a13 RTAki RTAkj þ a14 RTAki þ a15 RTAkj k¼1

k¼1

(2)

k¼1

RTAki is unity when the importing country i is a member of the trading bloc k and zero otherwise. A positive coefficient on this variable (a14 > 0) implies that countries that are members of RTAs will import more than an equivalent country that is not a member, and this would indicate that this RTA is trade creating. The second dummy variable, RTAkj takes the value of one if the exporting country j belongs to RTAk and zero otherwise. If a15 > 0, this means that countries import more from other countries 6

The South Asian Association for Regional Cooperation (SAARC) was established in 1985, with members being Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan and Sri Lanka.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

979

Table 1 List of Regional Trade Agreements (RTAs)a PTA

Date of entry

Member countries

ASEAN or the AFTA (Association of South East Asian Nations or the Asian Free Trade Association) APEC (Asia-Pacific Economic Cooperation)

1967

1983

Indonesia, Malaysia, Philippines, Singapore and Thailand Australia, Canada, Chile, China, Hong Kong, China, Indonesia, Japan, Korea, Malaysia, Mexico, New Zealand, Papua New Guinea, Peru, Philippines, Russia, Singapore, Thailand, United States, Vietnam Australia, New Zealand

1991 1994

Argentina, Brazil, Paraguay, Uruguay Canada, Mexico, United States

CER (Australia-New Zealand, Closer Economic Relations Trade Agreement) MERCOSUR (Southern Cone Common Market) NAFTA (North American Free Trade Agreement)

1989

a

Although ASEAN was established in 1967 with these original five member countries, the Asian Free Trade Agreement (AFTA) was initiated in 1993. There have also been additions to its membership. This paper, does not make adjusts to the membership countries, but does estimate the equations for sub-periods in an attempt to take account of changes during the sample.

that are members of RTAk, and again this implies that RTAs are trade creating. That is, positive coefficients on a14 and a15 imply ‘trade creation’, whereas if these coefficients are negative, this means that trade is lower, therefore the formation of trade blocs may be trade diverting. If the dummy variable which is the product of RTAki and RTAkj is unity, this means that both the importing country i and the exporting country j are members of the same RTAk. Therefore, if a13 > 0 this means that countries import more from countries that are members of the same RTA. An indication that an RTA is trade diverting would be that a14 and a15 are significant and negative, while a13 is significant and positive. 4. Estimation procedure and results The data used in this paper covers a period of 20 years from 1980 to 2000 and includes 26 countries from five RTAs (ASEAN or AFTA, CER, APEC, the Southern Cone Common Market or MERCOSUR and NAFTA). The model was estimated for four sub-periods; 1980–1985, 1985–1990, 1990–1995 and 1995–2000. With the pooled data being used to estimate a single regression equation, allowing for the coefficients to be different in the four group periods. The model was also estimated for the entire period using the full panel.7 The results for all of these equations are presented in Tables 2 and 3. 4.1. The basic gravity model results Table 2 reports the basic gravity model results when estimated using panel data for the four groups of observations as well as for the entire sample of 1980–2000. In general, these equations fit the data well, indicating that the proposed explanatory variables were significantly related to bilateral trade. The coefficients of determination (R2) range between 79% and 81%. The F-test 7

This method has been used in many gravity models since Aitken’s (1973) pioneering work.

980

Variable C GDPi GDPj POPi POPj DISTij AREAi AREAj LANGij EXRi EXRj TAXi TAXj R2 Adjusted R2 Prob(F-statistic) Cross-sections DW stat Panel observations

Expected signs + + + ?  ? ? +  +  

1980–1985

1985–1990

1990–1995

1995–2000

1980–2000

12.6956 (0.0000)*** 0.9141 (0.0000)*** 1.0748 (0.0000)*** 0.1323 (0.0011)*** 0.1892 (0.0000)*** 1.2489 (0.0000)*** 0.0449 (0.1868) 0.0633 (0.0203)** 0.1632 (0.0000)*** 4.5973 (0.0550)* 0.0485 (0.0000)*** 0.0431 (0.4623) 0.1825 (0.0003)*** 0.7973 0.7958 0.0000 293 0.1932 1604

13.4453 (0.0000)*** 0.9520 (0.0000)*** 1.0385 (0.0000)*** 0.0324 (0.3555) 0.1777 (0.0000)*** 1.2133 (0.0000)*** 0.0365 (0.2520) 0.0191 (0.4556) 0.1531 (0.0000)*** 2.3241 (0.0000)*** 0.0569 (0.0000)*** 0.0139 (0.7756) 0.0358 (0.4086) 0.8139 0.8125 0.0000 333 0.2086 1661

12.8364 (0.0000)*** 0.8901 (0.0000)*** 0.9582 (0.0000)*** 0.0847 (0.0007)*** 0.0223 (0.3481) 1.1087 (0.0000)*** 0.1432 (0.0000)*** 0.1185 (0.0000)*** 0.1518 (0.0000)*** 1.0401 (0.0011)*** 0.0525 (0.0000)*** 0.0741 (0.0653)* 0.0640 (0.0663)* 0.8138 0.8127 0.0000 456 0.1764 1981

13.4753 (0.0000)*** 0.9994 (0.0000)*** 0.9791 (0.0000)*** 0.0052 (0.8383) 0.0632 (0.0071)*** 1.1233 (0.0000)*** 0.1956 (0.0000)*** 0.1791 (0.0000)*** 0.0774 (0.0032)*** 1.0573 (0.0005)*** 0.0268 (0.0078)** 0.1438 (0.0012)** 0.0590 (0.1390) 0.8265 0.8254 0.0000 439 0.1357 1897

13.1686 (0.0000)*** 0.9490 (0.0000)*** 0.9968 (0.0000)*** 0.0642 (0.0001)*** 0.0412 (0.0071)*** 1.1703 (0.0000)*** 0.1472 (0.0000)*** 0.1290 (0.0000)*** 0.1483 (0.0000)*** 1.5067 (0.0000)*** 0.0444 (0.0000)*** 0.0756 (0.0021)*** 0.0895 (0.0001)*** 0.8129 0.8125 0.0000 486 0.1757 6179

Dependent variable: IMPORT. Method: pooled least squares. *p < 0.10, **p < 0.05, ***p < 0.01, significant at the 10%, 5% and 1% level, respectively.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

Table 2 Basic gravity model estimates—pooled regression analysis

Table 3 Extended gravity model estimates on RTAs—pooled regression estimates Expected signs + + + ?  ? ? +  +   + + + + + + + + + + + + + + +

1980–1985

1985–1990

1990–1995

1995–2000

1980–2000

13.3994 (0.0000)*** 1.0758 (0.0000)*** 1.1582 (0.0000)*** 0.1011 (0.1069) 0.2937 (0.0000)*** 1.3013 (0.0000)*** 0.0054 (0.8773) 0.0036 (0.9108) 0.2263 (0.0000)*** 5.1269 (0.0296)** 0.0373 (0.0000)*** 0.0530 (0.3618) 0.1214 (0.0229)** 0.3850 (0.0000)*** 0.2856 (0.0000)*** 0.2184 (0.0145)** 0.0594 (0.3900) 0.1770 (0.0000)*** 0.1729 (0.4802)

13.7716 (0.0000)*** 1.0680 (0.0000)*** 1.0193 (0.0000)*** 0.1276 (0.0194)** 0.1083 (0.0282)** 1.2739 (0.0000)*** 0.0059 (0.8638) 0.0284 (0.3725) 0.2210 (0.0000)*** 2.5451 (0.0000)*** 0.0414 (0.0000)*** 0.0651 (0.1880) 0.0917 (0.0434)** 0.4136 (0.0000)*** 0.1489 (0.0016)*** 0.2200 (0.0068)** 0.1442 (0.0104)** 0.2753 (0.0000)*** 0.1544 (0.3821)

12.7043 (0.0000)*** 0.9015 (0.0000)*** 0.9301 (0.0000)*** 0.0328 (0.2870) 0.0231 (0.4064) 1.0800 (0.0000)*** 0.1167 (0.0000)*** 0.1154 (0.0000)*** 0.1509 (0.0000)*** 1.4242 (0.0000)*** 0.0288 (0.0000)*** 0.0720 (0.0829)* 0.0900 (0.0109)** 0.1582 (0.0000)*** 0.0945 (0.0090)** 0.2375 (0.0001)*** 0.0367 (0.3896) 0.2515 (0.0000)*** 0.3430 (0.0131)** 0.1218 (0.0095)** 0.2018 (0.0000)*** 0.2905 (0.0000)*** 0.3269 (0.0000)*** 0.2697 (0.0000)*** 0.4876 (0.0000)***

0.8074 0.8052 0.0000 0.2236 293 1604

0.8275 0.8256 0.0000 0.2066 333 1661

0.8459 0.8440 0.0000 0.2551 456 1981

14.8507 (0.0000)*** 1.0371 (0.0000)*** 1.0256 (0.0000)*** 0.0655 (0.0365**) 0.0356 (0.1985) 0.9932 (0.0000)*** 0.1592 (0.0000)*** 0.1458 (0.0000)*** 0.1439 (0.0000)*** 1.3515 (0.0000)*** 0.0269 (0.0048)*** 0.0273 (0.5641) 0.0257 (0.5424) 0.1317 (0.0002)*** 0.0907 (0.0082)** 0.0454 (0.4272) 0.0370 (0.4418) 0.1533 (0.0000)*** 0.4265 (0.0075)** 0.0001 (0.9991) 0.0862 (0.2350) 0.1643 (0.0214)** 0.2106 (0.0001)*** 0.2058 (0.0001)*** 0.7122 (0.0000)*** 0.0085 (0.8344) 0.1823 (0.0000)*** 0.2596 (0.0038)*** 0.8581 0.8560 0.0000 0.1665 439 1897

13.6084 (0.0000)*** 0.9986 (0.0000)*** 0.9971 (0.0000)*** 0.0209 (0.3344) 0.0336 (0.0788)* 1.1681 (0.0000)*** 0.1009 (0.0000)*** 0.1018 (0.0000)*** 0.1853 (0.0000)*** 1.8366 (0.0000)*** 0.0371 (0.0000)*** 0.0489 (0.0482)* 0.1083 (0.0000)*** 0.2376 (0.0000)*** 0.1253 (0.0000)*** 0.2005 (0.0000)*** 0.0546 (0.0542)* 0.2188 (0.0000)*** 0.2909 (0.0026)*** 0.0245 (0.4177) 0.0723 (0.0193)** 0.1843 (0.0000)*** 0.1924 (0.0000)*** 0.2809 (0.0000)*** 0.4819 (0.0000)*** 0.0346 (0.2890) 0.1521 (0.0000)*** 0.1283 (0.1572) 0.8320 0.8312 0.0000 0.2040 486 6179

981

Dependent variable: IMPORT. Method: pooled least squares. *p < 0.10, **p < 0.05, ***p < 0.01, significant at the 10%, 5% and 1% level, respectively.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

C GDPi GDPj POPi POPj DISTij AREAi AREAj LANGij EXRi EXRj TAXi TAXj ASEANi ASEANj ASEANij CERi CERj CERij APECi APECj APECij MERCOSURi MERCOSURj MERCOSURij NAFTAi NAFTAj NAFTAij R2 Adjusted R 2 Prob(F-statistic) DW stat Cross-sections Panel observations

982

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

( p-value) results show that collectively the models were highly significant. These results are in line with the usual gravity model findings of other papers. Table 2 shows that the coefficients on GDPi and GDPj are all significant and positively signed, as are most of the population coefficients. That is, rich more populace countries tend to trade more. Also, as expected the coefficients on the distance variable, DISTij were all negative and significant. This suggests that transport costs (proxied by geographical distance) play an important role in determining the volume of trade between countries. The coefficients on AREAi and AREAj were all negatively signed and were almost always significant. Frankel (1997) also found this to be the case, and he suggested it was because large countries have more natural resources and tend to trade less with others. The coefficients on LANGij were all positive and significant. This implies that a common language and therefore ‘cultural similarities’ results in trade contacts being easier to make. When it came to the exchange rate, recall that an increase in EXRi or EXRj infers a depreciation of the real exchange rate. In line with expectations the coefficients associated with the importing country, EXRi are all significant and negative, while the EXRj coefficients are all significant and positive. The coefficients on TAXi (the importing countries trade tax index) were mostly negative and significant for the later years 1990–1995, 1995–2000 as well as for the whole period 1980–2000, while they were insignificant for the early years. This implies that tax regimes have become more important in determining trade flows since the 1990s. This may occur as when trade increases (as it has through out our sample period); therefore any remaining barriers become more important. Similarly, the coefficients on the exporting country’s tax rate TAXj were negative and significant for the years 1980–1985, 1990–1995 and for 1980–2000. These results generally imply that higher taxes reduce trade, which is in line with expectations. These tax related trade issues are discussed in detail by Whalley (2002). 4.2. The gravity model including RTAs To address the main questions in this paper, the analysis focuses on the estimated coefficients associated with the ‘trade creation’ and ‘trade diversion’ effects of RTAs. Table 3 reports the extended gravity model results, including the RTA dummy variables, carried out for the four groups of observations; 1980–1985, 1985–1990, 1990–1995 and 1995–2000, as well as for the full panel from 1980 to 2000.8 Again, in general the regression equations fit the data well indicating that the proposed explanatory variables are significantly related to bilateral trade. The coefficients of determination (R2) range between 81% and 86% and the F-test ( p-value) results show that collectively the models are highly significant and explain a large portion of the variation in the data. 4.2.1. ASEAN Concentrating on ASEAN, the cross-sectional results in Table 3 show that the estimated coefficients on ASEANi were all were positively signed and statistically significant. This means 8

Various combinations of these variables were tried. For example, the RTA dummy variables were included separately and then jointly for the case of ASEAN and APEC. This was done because of concerns about multicollinearity as APEC includes almost all of the ASEAN nations. However, as the results were not significantly different, only the results where both sets of dummy variables are included are reported here.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

983

that ASEAN countries import more than would be the case if they were not members of that RTA. Similarly, the ASEANj coefficients are positive and significant, which means that ASEAN counties also tend to export more. For the ASEAN intra-bloc trade variable (ASEANij) the coefficients were mostly negative and significant. The implication is that ASEAN countries have no preference for trade with other ASEAN members; in fact these results imply that they trade less with each other. A possible explanation for this may be that the ASEAN countries have similar characteristics so comparative advantage could see them looking elsewhere for countries to trade with. Notwithstanding the sign of the ASEANij coefficient, all of these results seem to be exactly what would be expected from such strongly outward oriented economies that make up the ASEAN RTA, especially as many of the trade reforms introduced here have been multilateralized. Therefore, the conclusion is that ASEAN is trade creating. These conclusions are in line with Elliott and Ikemoto (2004) and Tang (2005), as they both also used gravity models to show that the formation of ASEAN and the 1992 signing of the AFTA (ASEAN Free Trade Agreement) did not cause ‘trade diversion’. 4.2.2. CER The results in Table 3 indicate that the effects of CER were less conclusive than was the case for ASEAN. The CERi coefficients were all positive, but were only significant for the period 1985–1990, as well as for the entire panel of 1980–2000. All of the export coefficients (CERj) were positively signed and significant. The coefficients for the intra-bloc variable CERij were also all positive, but were only significant for the last two periods, as well as for the entire panel. The inference seems to be that CER had a general ‘trade creation’ effect, but it has also caused Australian and New Zealand’s preference for trade with each other to broadly increase. This is analogous to the findings of Frankel (1997) and Tang (2005) who also found the intra-bloc trade effects of the CER to be positive and significant. Although, like the current paper, Tang also found that CER was associated with Australia and New Zealand increasing their trade with the rest of the world. 4.2.3. APEC The estimates for APEC indicate that this RTA is trade diverting, rather than trade creating. The coefficients on APECi and APECj are mostly negative and significant, while all of the coefficients for the intra-APEC trade variable (APECij) are positively signed and significant. This indicates a strong intra-bloc effect with trade flowing more intensely among economies that are members of APEC than with the rest of the world. This suggests that APEC has not been achieving its goal of open regionalism, and is in fact ‘trade diverting’, which implies reduced welfare. This finding for APEC is in line with the estimates of Frankel (1997) who attributed the large positive coefficients on the intra-bloc variables to the fact that a large share of total world trade is accounted for by APEC member economies. Given that there has not been substantial trade reform among APEC, it is likely that the size effects may dominate, and as such this is likely to be a good explanation. Although some believe that one of APEC’s main objectives was to put pressure on the Europeans to agree to complete the Uruguay round of multilateral trade negotiations (see Bergsten, 1996). Therefore, any judgments about APEC need to be qualified because of its success in pushing the Europeans to reform.

984

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

Appendix A. Data, definitions and sources Variables

Expected signs

IMPORT

Definitions

Measures

Sources

It is calculated as the total of bilateral trade that is total trade, comprising merchandise trade and services trade deflated by the GDP deflator

Millions US$

Sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. Data are in constant 1995 US dollars Total population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship, except for refugees not permanently settled in the country of asylum, who are generally considered part of the population of their country of origin Distance between capital cities of pairs of countries

US$

International Monetary Fund (various issues), Direction of Trade Statistics; Department of Foreign Affairs and Trade (DFAT) (2004), APEC Region Trade and Investment World Bank (2001), World Development Indicators Database

GDP

+

POP

+/

DIST



AREA

/?

A country’s total area

km2

EXR

+/

Official exchange rate refers to the exchange rate determined by national authorities or to the rate determined in the legally sanctioned exchange market. It is calculated as an annual average based on monthly averages (local currency units relative to the US dollar). The GDP deflator is used to account for inflation between the two countries then adjusts the exchange rate

Average units

Millions

World Bank (2001), World Development Indicators Database

km

Havemann (2002) International trade website and Fitzpatrick and Modlin (1986) book of direct line distances World Bank (2001) and Philip’s World Atlas and Gazetteer (1994) World Bank (2001), World Development Indicators Database

The other aspect of APEC is that these conclusions may change in time. For example, the industrialized APEC countries have until 2010 to complete their tariff reductions, while the developing countries have until 2018.9

9

Although APEC has these stated objectives, it has established very little else by way of mechanisms whereby they may be achieved. Therefore, the results in this paper may be considered unsurprising.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

985

4.2.4. MERCOSUR The results here show that MERCOSUR has been clearly trade diverting. All of the import and export coefficients (MERCOSURi and MERCOSURj) are negative and significant, while all of the intra-RTA coefficients MERCOSURij are positive and significant. This shows that since its inception in 1991 MERCOSUR has increased trade among its members at the expense of trade with the rest of the world, thereby reducing welfare. This result supports the findings of Yeats (1998) and Clarete, Edmond, and Wallack (2003) who attribute the rise of intra-bloc MERCOSUR trade to the introduction of discriminatory tariffs against non-members. Yeats further noted that the intra-regional trade and export growth among members was concentrated in products that were not competitive outside of the region. 4.2.5. NAFTA These results suggest some form of export diversion, as the NAFTj coefficients are significant and negative, while the intra-region variable NAFTij coefficient is positive and significant for 1995–2000.10 None of the other coefficients are significant. These results are generally in line with the study by Fukao, Okubo, and Stern (2003) who identified NAFTA’s ‘trade diversion’ effects for textiles and apparel imports into the US, as Mexican products are substituted for Asian goods. In general, however, it can be argued that these results show that NAFTA has had little effect during the period covered in this study, which is in line with Kruger’s (1999) conclusion that as yet NAFTA has not had much effect on trade patterns. More recently Tang (2005) also concluded that the formation of NAFTA has had no significant effect on trade flows. 5. Conclusion This study uses a gravity model to examine bilateral trade involving five trading blocs, with data from 26 countries from 1980 to 2000. The estimated coefficients from the basic gravity model show that GDP, population, distance between trading partners, as well as cultural similarity (a common language) and physical area explain much cross country trade. The study also uses some price variables, namely the real exchange rate and taxes and finds that the empirical results line up with prior expectations. That is, real exchange rate movements had the expected affects, as depreciations encouraged exports and discouraged imports. With regard to taxes, the results suggest that taxation decreases bilateral trade. The regression estimates for the effects of the different RTAs varied remarkably. ASEAN and CER were found to foster greater trade worldwide and so were welfare enhancing. However, although APEC, MERCOSUR and NAFTA tended to expand intra-bloc trade, to some extent this was at the expense of their trade with the rest of the world, which implies ‘trade diversion’ and a loss of welfare. References Adams, R., Dee, P., Gali, J., & McGuire, G. (2003). Trade and investment effects of preferential trading arrangements— old and new evidence. Productivity Commission Staff Working Paper, Canberra. Aitken, N. (1973). The effect of the EEC and EFTA on European trade: A temporal cross-section analysis. American Economic Review, 63, 881–892.

10

Note that there are only slight differences between the NAFTA results for 1995–2000 and for 1980–2000, as NAFTA only commenced in 1994.

986

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

Anderson, J. (1979). A theoretical foundation for the gravity equation. American Economic Review, 69, 106–115. Bayoumi, T., & Eichengreen, B. (1997). Is regionalism simply a diversion: Evidence from the evolution of the EC and EFTA. Chicago: University of Chicago Press. Beckerman, W. (1956). Distance and the pattern of intra-European trade. The Review of Economics and Statistics, 28, 31–40. Bergsten, C. F. (1996). Asia-pacific economic cooperation: Summary of the second report of the eminent persons group. In M. Dutta (Series Editor), Research in Asian Economic Studies, Vol. 7, London: JAI Press, Ltd. (p. 5). Bergstrand, J. H. (1985). The gravity equation in international trade: Some microeconomic foundations and empirical evidence. Review of Economics and Statistics, 67, 474–481. Bergstrand, J. H. (1989). The generalised gravity equation, monopolistic competition, and the factor proportions theory in international trade. Review of Economics and Statistics, 71, 43–153. Cernat, L. (2001). Assessing regional trade arrangements: Are south–south RTAs more trade diverting? Policy Issues in International Trade and Commodities Study Series. No. 16 UNCTAD. Clarete, R., Edmond, C., & Wallack, J. S. (2003). Asian regionalism and its effects on trade in the 1980s and 1990s. Journal of Asian Economics, 14, 91–129. Department of Foreign Affairs and Trade (DFAT). (2004). The APEC Region Trade and Investment. Retrieved 16.03.2004, from www.dfat.gov.au/trade/ftas.html. Elliott, R. J. R., & Ikemoto, K. (2004). AFTA and the Asian crisis: Help or hindrances to ASEAN intra-regional trade? Asian Economic Journal, 18, 1–23. Filippini, C., & Molini, V. (2003). The determinants of East Asian trade flows: A gravity equation approach. Journal of Asian Economics, 14, 695–711. Fitzpatrick, G., & Modlin, M. (1986). Direct line distances (international edition). The Scarecrow Press Inc. Frankel, J. (1997). Regional trading blocs in the world economic system. Washington: Institute for International Economics. Frankel, J., & Wei, S. (1995). Emerging currency blocks. Berlin: Springer. Fukao, K., Okubo, T., & Stern, R. M. (2003). An econometric analysis of trade diversion under NAFTA. The North American Journal of Economics and Finance, 14, 3–24. George Philip Limited. (1994). Philip’s World Atlas and Gazetteer. George Philip Limited. Hassan, K. (2001). Is SAARC a viable economic block? Evidence from gravity model. Journal of Asian Economics, 12, 263–290. Havemann, J. (2002). International Trade Data. Retrieved 20.08.2004, from http://www.macalester.edu/research/ economics/PAGE/HAVEMAN/Trade.Resources/TradeData.html. Havemann, J., Nair-Reichert, U., & Thursby, J. (2003). How effective are trade barriers? An empirical analysis of trade reduction, diversion and compression. Review of Economics and Statistics, 85, 480–485. International Monetary Fund. (Various issues). Direction of Trade Statistics. Washington: IMF Publication Services. Johnson, H. G. (1960). The economic theory of customs unions. Pakistan Economic Review, 10, 14–30. Kreinin, M., & Plummer, M. (1998). Ex post estimates of the effects of the European single market program on the exports of developing countries. ASEAN Economic Bulletin, 15, 206–214. Krueger, A. (1999). Trade creation and trade diversion under NAFTA. National Bureau of Economic Research Working Paper, No. 7429. Linnermann, H. (1966). An econometric study in international trade flows. Amsterdam: North Holland. Lloyd, P. J., & MacLaren, D. (2004). Gains and losses from regional trading agreements: A survey. The Economic Record, 80, 445–467. Low, L. (2003). Multilateralism, regionalism, bilateral and cross-regional free trade arrangements: All paved with good intentions for ASEAN? Asian Economic Journal, 17, 65–86. Oguledo, V., & Macphee, C. (1994). Gravity models: A reformulation and an application to discriminatory trade arrangements. Applied Economics, 26, 107–120. Pomfret, R. (1997). The economics of regional trading arrangements. Oxford: Clarendon Press. Rose, A. (2000). One money, one market? The effects of common currencies on international trade. Economic Policy, 15, 7–46. Schiff, M., & Winters, A. (2003). Regional integration and development. Washington, DC: Oxford University Press. Soloaga, I., & Winters, L. (2001). Regionalism in the nineties: What effect on trade? The North American Journal of Economics and Finance, 12, 1–29. Tang, D. (2005). Effects of the regional trading arrangements on trade: Evidence from the NAFTA, ANZCER and ASEAN countries 1989–2000. Journal of International Trade and Economic Development, 14, 241–265. Thursby, J., & Thursby, M. (1987). Bilateral trade flows, the Linder hypothesis and exchange risk. Review of Economics and Statistics, 69, 488–495.

B. Jugurnath et al. / Journal of Asian Economics 18 (2007) 974–987

987

Tinbergen, J. (1962). Shaping the world economy: Suggestions for an international trade policy. New York: Twentieth Century Fund Press. Viner, J. (1950). The Customs Union Issue New York. Carnegie Endowment for International Peace. Whalley, J. (2002). Taxes and trade. Working Paper Series, The University of Western Ontario. World Bank. (2001). World Development Indicators (WDI) online database. Retrieved 28.04.2004, from http:// www.devdata.worldbank.org.mate.lib.unimelb.edu.au/dataonline/. Yeats, A. (1998). Does MERCOSUR’s trade performance raise concerns about the effects of regional trade agreements? The World Bank Economic Review, 12, 1–28.