CONTENTS. Preface... Part I: The Economic Effects of a Korea-U.S. FTA

Transcription

1

2 CONTENTS Preface Part I: The Economic Effects of a Korea-U.S. FTA Conclusions and Implications for Further Research and Policy Excerpt from Economic Effects of a Korea-U.S. Free Trade Agreement* Kozo Kiyota and Robert M. Stern Comments on the Kiyota-Stern Study Jeffrey J. Schott Implications of the U.S.-Korea Free Trade Agreement: A General Equilibrium Approach Renan Zhuang and Won W. Koo Part II: Dynamic Effects of an FTA The Payoff to South Korea From Globalization Gary Hufbauer and Agustín Cornejo How Financial Multilateralism Can Increase Sustainable Output, Employment, and Income in the Pacific Region Douglas H. Brooks and David Roland-Holst Part III: Scope for Dynamic Effects in Korea s Economy Dynamic Consequences of a Korea-U.S. Free Trade Agreement: Foreign Direct Investment Arthur Alexander Is A Free Trade Agreement a Royal Road to Prosperity? Demystifying Trade Regionalism Sungjoon Cho Regional Integration and a Free Trade Agreement Among China, Japan, and Korea Hee-joon Kang

3 Comment: Scope for Dynamic Effects in Korea s Economy Choi Nakgyoon Part IV: Conference Discussion and Conclusions Summary of Proceedings Bernard K. Gordon * Volume 4 in KEI Special Studies Series, published by Korea Economic Institute of America

4 Static and Dynamic Consequences of a KORUS FTA 29 THE PAYOFF TO SOUTH KOREA FROM GLOBALIZATION * Gary Hufbauer and Agustín Cornejo CONTENTS I. Introduction II. Payoff from Past Globalization III. Future Gains IV. Conclusion * Gary Hufbauer and Agustín Cornejo are with the Peterson Institute for International Economics. This paper draws heavily from previous research of Bradford, Grieco, and Hufbauer for Payoff to America from Global Integration in Bergsten (2005).

5 30 The Korea Economic Institute I. Introduction Since the end of the Second World War, eight multilateral bargains under the General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO), coupled with the fall of the Soviet Union, have delivered considerable trade liberalization around the globe. In addition, many nations including South Korea have slashed their barriers and opened their markets unilaterally or through preferential trade agreements. Widespread opening has contributed to the best half century of world economic growth since the time of Jesus Christ (Maddison 2003, table 8b). Dramatic improvements in communications and transportation technology have of course made a major contribution as well. While it is difficult to disentangle the effects of policy from the effects of technology, credible estimates give approximately equal weight to these two forces of world economic integration (Yi 2003). This article summarizes the economic payoff to South Korea from its postwar trade opening (reflecting the combined force of policy and technology) and estimates the potential future gains from more policy opening going forward. To quantify these gains, we survey different methodologies as well as educated guesses available in the literature. Despite their differences, the methods cited may bracket a range that is quite likely to contain the true value of gains. Taken together they suggest that South Korea has enjoyed very substantial past gains and may reap large future payoffs from expanded commercial ties with the global economy. We calculate that trade opening since World War II has added between $90 billion and $100 billion to the South Korean economy, around $2,000 per capita additional income. Our speculative estimates of potential additional gains from removing the rest of trade and investment barriers between South Korea and its economic partners range from $57 billion to $76 billion, around $1,200 to $1,600 per capita. 1 Because trade opening permanently raises national income, these gains are enjoyed annually. Trade and Growth Regressions II. Payoff from Past Globalization The first approach draws on a study designed to identify economic attributes and policies that have the strongest effect on per capita income growth (OECD 2003, 1. Our estimates of potential future gains should be viewed as the gains from wholesale reforms that would move the current trading system to complete liberalization. Ongoing negotiations, such as the WTO Doha Development Round and for various free trade agreements, are not that ambitious.

6 Static and Dynamic Consequences of a KORUS FTA 31 chapter 2). 2 The OECD study calculates a coefficient of 0.2 for advanced countries for the positive effect on per capita income from a higher long-term level of trade exposure (merchandise exports plus imports). 3 According to this coefficient, a 10 percent rise in a country s long-term trade exposure (imports plus exports) results in a 2 percent increase in the level of annual per capita output, holding everything else equal. Armed with this estimate, we can calculate the effect of increased trade on GDP and per capita income in South Korea. However, the OECD estimate is based on a sample heavily biased toward developed countries, while South Korea was a developing country until recent decades. As a result, the OECD estimate is quite likely lower than the coefficient relevant for most developing countries, and for this reason we do not apply the 0.2 coefficient to decadal changes in South Korean trade exposure. Cline (2004) summarizes the econometric estimates of the long-term elasticity of output per capita to the merchandise ratio. He finds estimates that are much higher than suggested by the OECD, as they range between 0.14 and Accordingly, we apply a low middle of the road coefficient (0.3) in recognition that South Korea was still a developing country during several decades of the period under consideration. 4 With this approach, Table 1 shows the evolution of South Korean trade exposure (merchandise trade), together with the calculated payoff in terms of higher income levels. 5 The eventual increase in per capita income is expressed in 2000 dollars, calculated as a result of rising trade exposure during each decade. The sum of this 2. Among the attributes and policies: physical and human capital, inflation (level and volatility), taxes, government spending, research and development spending, financial sophistication, and trade exposure. 3. The question of causality (i.e., whether faster output growth increases trade or vice versa) and the possibility that omitted variables explain the statistical connection between trade and growth have bedeviled this line of inquiry (Berg and Krueger 2003; Hallak and Levinsohn 2004). Frankel and Romer (1999) used instrumental variables derived from a gravity model to make a strong case that trade causes growth. Berg and Krueger (2003) survey the literature and summarize the pros and cons; they conclude that the literature shows that openness is fairly robustly a cause of growth although there is substantial uncertainty surrounding these estimates. 4. Cline (2004) confirms that 0.5 corresponds to the simple average of the central values of the estimates available in the literature. 5. Since the Second World War, many key variables that determine income levels, such as the size of the workforce, the share of manufacturing value added, and labor productivity, have changed dramatically. For this reason, we break the full period down into several subperiods and then perform intermediate calculations to calculate total postwar gains. The intermediate results that appear in the tables do not reflect gains realized solely during that subperiod. Most of the structural changes set in motion by trade openness take a decade or longer to mature in terms of greater productivity and higher income. Thus, our subperiod figures are an estimation of annual gains that will eventually result from liberalization that occurred in the subperiod. We then add up subperiods to arrive at a final cumulative figure of annual gains caused by opening.

7 32 The Korea Economic Institute Table 1: Growth Regressions: Benefits of Increased Trade Exposure for South Korea Time Trade Period Per capita output Benefit in Pop. Per capita period exposure a change growth due to trade b 2003 b GDP % % % 2000 $ billion 2000 $ million 2000 $ , , , , , ,043 Total 2,110 Sources: IMF-IFS (various years); World Bank (various years). a. Calculated as the average (X + M)/GDP for the decade. b. Subperiod gains are an intermediate calculation, not an estimate of short-term gains from liberalization. Gains due to liberalization may require 10 to 20 years to be fully realized. For the entire period 1960 through 2005, we use a coefficient of 0.3 for the elasticity of trade exposure and growth based on a low midpoint estimate between Cline s (2004) survey for developing countries and the OECD (2003). column, $2,100 per capita, reflects the calculated total increase in per capita GDP resulting from deeper trade exposure between 1959 and The $2,100 payoff translates to an additional $100 billion of GDP, or roughly 17 percent of 2005 GDP measured in 2000 dollars (total GDP gains are calculated by multiplying per capita gains by 2005 population, 47.6 million). Import Variety and Price Benefits Feenstra (1994) and Broda and Weinstein (2006) charted new ground in the study of trade gains by examining expanded product varieties made available through imports. Broda and Weinstein concluded that conventional import price indices, which do not take into account new import varieties, overstate U.S. import inflation by 28 percent. Using product level data on the fraction of imported goods in total consumption, Broda and Weinstein (2006) calculate that the 28 percent increase in consumer purchasing power is equivalent to a gain of 2.8 percent of 2001 U.S. GDP from greater import variety. Feenstra (2006) not only concludes that this estimate can be safely extended to other developed countries, but he also indicates that the likely gain is probably larger for countries with a larger trade exposure (e.g., South Korea). However, applying the 2.8 percent figure to Korean purchasing power suggests a gain of $375 per capita in 2005 GDP (measured in 2000 dollars), or a total of roughly $18 billion (2000 dollars) for the overall economy.

8 Static and Dynamic Consequences of a KORUS FTA 33 Import Variety and Productivity Benefits Broda, Greenfield, and Weinstein (2006) extend their methodology to a number of countries, including South Korea, and find that increased import variety accounted for an average of 5 percent of total factor productivity (TFP) during in the typical developed country, and that the South Korean experience was very close to this average. Import variety contributes to TFP because, with a wider range of choices, industrial firms can rightsize their inputs to the exact specifications of their own products. Making a bold assumption that increased import variety accounts for 5 percent of South Korea s total factor productivity growth over the entire postwar period ( ), we can roughly calculate that this single channel contributed about $550 per capita (2000 dollars) to the total gains. 6 This amounts to about $26 billion of GDP in 2005 (measured in 2000 dollars). Export Variety and Productivity Growth The growth in South Korean export variety during the past 50 years has been outstanding. A simple analysis of UN Comtrade data at the highest level of disaggregation available in Standard International Trade Classification (SITC) Revision 1 (the only classification that allows a study of long-term changes) shows that South Korea s exports covered just 570 lines in 1965 but 1,060 in For comparison purposes, the corresponding figures for U.S. exports are 820 and 1,120. Feenstra et al. (1999) study the linkage between export variety and industry productivity growth in South Korea and Taiwan. 7 They find that increased export variety had a positive and significant effect on TFP in 9 of the 16 sectors. In a subsequent study, Feenstra and Kee (2006) estimate that increased export variety can explain about 13 percent of national productivity growth in the industrial sector. 8 We assume that that same coefficient applies to the entire traded goods sector some 32 percent of the Korean economy in recent years, but higher in earlier decades. Applying that coefficient to the fraction of South Korea s per capita income associated with the traded goods sector suggests a gain of almost $400 in per capita income, or $19 billion (in 2000 U.S. dollars). This result previews our estimate in the next section. 6. We obtain this rough approximation by calculating 5 percent of the per capita GDP gain measured in 2000 dollars [($12,043 less $1,110) times 5 percent]. 7. Feenstra et al. (1999) measure variety based on disaggregated exports from Korea and Taiwan to the United States for 16 sectors during Feenstra and Kee (2006) estimates are based on a pool of 34 developed and developing countries (including Korea) and data from 1982 to 1997.

9 34 The Korea Economic Institute Productivity Gains from Sifting and Sorting The sifting and sorting mechanism identified by Richardson 9 provides a theoretically grounded way of thinking about the impact of globalization on productivity. Increased global opportunities that result from declining tariffs and transportation costs, better communications, and other forces can enhance productivity by weeding out weak firms and nourishing strong ones. Highly productive firms tend to expand when they discover opportunities in the global market, while unproductive firms tend to contract under the pressure of import competition. The result is an increase in average productivity of the industry, even though there may be no effect on any single firm s productivity. The sifting and sorting mechanism by which these productivity gains emerge reflects market selection among heterogeneous firms that differ in either their inherent overall productivity or their access to worker pools of varying skills, creativity, and reliability. Some scholars have tried to estimate the firm-level payoff from trade and investment liberalization, where sifting and sorting is the main mechanism. Econometric results are still at a preliminary stage, but a simulation by Bernard et al. (2003) suggests that a 5 percent reduction of global trade barriers (tariffs and other costs expressed on an ad valorem basis) could, through the sorting of resources toward more productive firms, boost overall U.S. manufacturing productivity by 4.7 percent. The coefficient is derived from a model simulation, not a longitudinal estimate. While the actual coefficient could differ quite a bit from one to one, we draw on this result that a one-percentagepoint decrease in trade barriers raises manufacturing productivity by about 1 percent to estimate the gains from trade opening through the sifting and sorting process. 10 To calculate the postwar reduction in South Korean trade barriers, we look at the difference between the simple average applied tariffs in different periods. We take the difference of 31.3 percentage points as a rough estimate of the overall decline in tariff barriers during the past 50 years. 11 Applying the one-to-one coefficient gleaned from the literature that a one-percentagepoint decline in trade barriers will induce a 1 percent improvement in productivity in traded goods sectors we use the constructed series in Table 2 as a measure of induced productivity gains in the traded goods sectors for each period as a result of 9. This section is largely paraphrased from Richardson (2004). 10. We assume that a one-to-one payoff also results when agriculture and mining trade barriers are decreased. Other methods of summarizing gains are possible but all require some degree of speculation. Appendix 2A of Bradford, Grieco, and Hufbauer (2005) (based on Richardson [2004]) presents a separate methodology that examines the growing use of intermediate imports in U.S. manufacturing. 11. This method ignores nontariff barriers, which probably declined as well.

10 Static and Dynamic Consequences of a KORUS FTA 35 Table 2: Sifting and Sorting: Productivity Benefit of Reduced Import Tariff Barriers for South Korea Years Import Simple Estimated Average share Growth in output per worker Benefit in Memorandum liberaliza- average MFN ad valorem of traded goods in full economy due 2003 f tion rate a tariff at end tariff sectors in to tariff change e Output per Labor of period b reduction c South Korean worker force value added d percent percent percentage percent percent 2000 dollars billion dollars million points dollars , , , Source: World Bank (various years), Dent (2002), Ng (2006). a. The import liberalization rate is the ratio of tariff lines not subject to import licenses to total tariff lines in the Korean schedule. b. Two sources were used for constructing the series: data for are based on data reproduced in Dent (2002), while subsequent data are based on World Bank (Francis Ng database). c. Given the estimated 1-to-1 relation between tariff reduction and productivity gains, these numbers are also the estimated productivity growth in the traded goods sector due to tariff reduction. d. Traded sectors are manufacturing, agriculture, and minerals. e. Subperiod gains are an intermediate calculation, not an estimate of short-term gains from liberalization. Gains due to liberalization may require 10 to 20 years to be fully realized. f. Benefit in 2003 is calculated by multiplying the benefit per worker measured in 2003 dollars by total employment in 2003.

11 36 The Korea Economic Institute tariff cuts. To find the productivity improvement to the entire economy caused by reduced tariff barriers, we multiply this figure by the average share of traded goods sectors in the total economy for different periods. 12 Based on these productivity increases, we estimate an annual GDP benefit of $1,300 per worker for all liberalization between 1960 and 2004 roughly an income gain of $615 on a per capita basis. This equates to a benefit of $29 billion of additional GDP in the South Korean economy in 2005 (measured in 2000 dollars). Since the sifting and sorting estimate measures just productivity gains owing to resource reallocation, we can combine this estimate an income gain of approximately $615 per capita with the other gains enumerated earlier: a $375 per capita gain from greater household purchasing power; a $550 per capita gain from enhanced TFP through rightsizing industrial inputs; and another $400 per capita gain from enhanced TFP through greater export variety. The result of this adding-up exercise is a per capita GDP gain of $1,940 (in 2000 dollars) from postwar Korean opening, which equates to some $92 billion additional GDP for the Korean economy. The adding-up result is in the same ballpark as the growth regression described at the outset $2,100 gain per capita, $100 billion gain for total GDP. Since the growth regression can be interpreted as an umbrella calculation of all the enumerated channels, plus others not identified in this article, the rough correspondence gives reassurance. III. Future Gains So far we have tried to quantify the gains of past trade liberalization for South Korea. Now we turn to the future. While South Korean average tariffs are moderate by developing country standards, they are high when compared with advanced OECD countries (Table 3). A similar conclusion can be drawn from the trade restrictiveness indexes for key areas of the service sector, namely banking and telecommunications (Table 4). While comparable estimates of trade barriers in services are difficult to find, several comparative indices the A.T. Kearney annual survey of corporate executives, the Foreign Direct Investment Confidence Index, is one index indicate that South Korea s attractiveness for inward foreign direct investment (FDI) lags behind the best emerging markets. One study the PricewaterhouseCoopers opacity index suggests that FDI could increase if South Korea s regulation were less opaque. In this section, we draw on scenarios that are contemplated in standard computable general equilibrium (CGE) and gravity models. A caveat is in order. After examining 12. This method assumes that labor moving from traded to nontraded sectors maintains its productivity level.

12 Static and Dynamic Consequences of a KORUS FTA 37 the predictive power of these models, as applied to past free trade agreements (FTAs) and the Uruguay Round, DeRosa and Gilbert (2005) conclude that neither model has a stellar forecasting record. Table 3: Simple Average Bound and Applied Tariffs, Selected Nations, percentage Country Simple average MFN tariffs Bound Applied Advanced nations Japan Australia United States Canada European Union South Korea Developing nations Indonesia Malaysia South Africa Russia n.a. 9.5 China Brazil Colombia Pakistan India Egypt Source: WTO (2006). Note: Tariff in latest available year, 2005 or Estimates from CGE Models: Global Free Trade Using the Michigan Model of World Production and Trade, Kiyota and Stern (2007) provide the most optimistic scenario of the potential impact of a world without tariffs and nontariff barriers. They estimate that South Korean GDP could expand by almost 12 percent, with most of the gains recorded in the manufacturing sector, followed by services. Not surprisingly, South Korean agricultural GDP would contract. The estimated gains are equivalent to $76 billion in 2005 GDP (measured in 2000 dollars), or almost $1,600 per capita. The key feature of the Michigan model is that, while focusing on static gains, it allows for the modeling of many features of the new trade theory such as product variety, economies of scale, and monopolistic competition.

13 38 The Korea Economic Institute Table 4: Most-Favored-Nation Applied Tariffs on Merchandise Imports and Trade Restrictiveness Indices for Traded Services Country/region Simple average Trade restrictiveness indices (2000) merchandise tariff (2005) Accountancy Architectural Services Banking Distribution Telecom Maritime Legal Australia Canada France Germany Hong Kong Japan Singapore United Kingdom United States South Korea Brazil India n.a Indonesia Malaysia Mexico South Africa n.a. n.a. Source: Productivity Commission (various years). a. The extent of government regulation of a particular service is quantified using a trade restrictiveness indices. Trade restrictiveness indices summarize the nature and extent of restrictions on trade in services for each economy. The more restrictions and the greater their severity, the higher the index and the more restrictive an economy is judged to be.

14 Static and Dynamic Consequences of a KORUS FTA 39 Another advantage of the Michigan model is that it better accounts for potential gains in the service sector. Other CGE models without such features tend to yield lower results. As Table 5 illustrates, gains for South Korea under global free trade are much lower around 2 percent when modeled using the standard GTAP model, as done by Scollay and Gilbert (2001), DeRosa and Gilbert (2004), and Cline (2004). This is equivalent to about $13 billion in 2005 GDP (measured in 2000 dollars). However, as Cline (2004) points out, the GTAP database does not capture barriers in the services sector; therefore, Cline cautions that the results should be interpreted as essentially referring to merchandise trade only. 13 A scenario of global free trade requires hugely successful WTO negotiations of much more ambitious scope than currently envisaged. In fact, Cline (2004) performs different simulations based on various proposals and concludes that South Korea s prospective gains under the Doha Round of negotiations realistically range from 15 percent to 67 percent of potential gains under global free trade. An alternative scenario that is highly valued by economists but dismissed by policymakers is the unilateral dismantling of barriers. Kiyota and Stern (2007) conclude that unilateral liberalization would bring large gains for South Korea, on the order of 5 percent of GDP, or roughly $30 billion dollars in 2005 GDP (measured in 2000 dollars), $630 per capita. However, this gain is just 40 percent of the global free trade scenario, indicating that South Korea also has a major stake in the removal of barriers by its commercial partners. Estimates from CGE models: Bilateral and Regional Approaches Bilateral and regional initiatives may provide a second-best path toward the benefits from global free trade. Table 5 summarizes the results of empirical simulations of several negotiating scenarios for South Korea. The first observation is that, when models allow for dynamic gains (e.g., an increase in capital stock or better productivity as a result of trade liberalization), the gains vastly exceed the results predicted under static models, sometimes twice or three times as large. 14 However, the prospects of dynamic gains from trade liberalization are hotly disputed, and we do not pursue them further in this paper. 13. Brown, Deardorff, and Stern (2003) estimate that the gains from services liberalization could be four times the gains from goods liberalization. 14. The same magnitude of underestimation may well apply to the estimated gains of global free trade reported in the previous section.

15 40 The Korea Economic Institute Table 5: Economic Effects of Global Free Trade and Proposed FTAs on South Korea, CGE-Based Estimates, Percentage increase or billions of U.S. dollars Authors Proposed partner Type of model Estimated impact Global free trade and the unilateral benchmark GDP Exports to Imports from Total Total partner partner exports imports percent percent or percent or percent percent change $ billion $ billion change change chnage change Scollay and Gilbert (2001) Global free trade Static CGE DeRosa and Gilbert (2004) Global free trade Static CGE 2.16 Cline (2004) Global free trade Static CGE 2.41 Kiyota and Stern (2007) Global free trade Static CGE (with IRS) Kiyota and Stern (2007) Unilateral free trade Static CGE (with IRS) 4.60 Bilateral and regional scenarios: static and dynamic CGE models Cheong (1999) a Japan Static CGE KIET (2000) a Japan Static CGE 0.07 IDE / JETRO (2000) Japan Static CGE Cheong (2001) Japan Static CGE Brown et al. (2001) Japan Static CGE (with IRS) 0.23 Scollay and Gilbert (2001) Japan Static CGE Nakashima (2002) Japan Static CGE 0.29 Kiyota and Stern (2007) Japan Static CGE (with IRS) 0.30 Nam et al. (2004) China Static CGE Lee H. et al. (2005) China Static CGE Scollay and Gilbert (2001) Japan + China Static CGE Cheong (2003) Japan + China Static CGE 0.94 Lee C. et al. (2005) Japan + China Static CGE

16 Static and Dynamic Consequences of a KORUS FTA 41 Scollay and Gilbert (2001) ASEAN + 3 Static CGE Kim et al. (2005) European Union Static CGE Wang and Cheong (1998) United States Static CGE 1.70 McDaniel and Fox (2001) United States Static CGE Choi and Schott (2001) United States Static CGE DeRosa and Gilbert (2004) United States Static CGE Lee J. and Lee H. (2005) United States Static CGE (manuf.) Schott, Bradford, and Moll (2006) United States Static CGE 3.50 Kiyota and Stern (2006) United States Static CGE (with IRS) 1.26 $6.9 $9.2 Scollay and Gilbert (2001) APEC on MFN basis Static CGE IDE / JETRO (2000) Japan Dynamic CGE b Cheong (2001) Japan Dynamic CGE c 2.88 $3.3 $9.0 Nakashima (2002) Japan Dynamic model 1.09 McKibbin, Lee, and Cheong (2002) Japan Dynamic G-cubed d 3.90 Kawasaki (2003) Japan Dynamic CGE Nam et al. (2004) China Dynamic CGE $ $ Lee H. et al. (2005) China Dynamic CGE Cheong (2003) Japan + China Dynamic CGE 2.45 Lee C. et al. (2005) Japan + China Dynamic CGE Bchir and Fouqin (2006) ASEAN (industrial) Dynamic CGE Bchir and Fouqin (2006) ASEAN (agriculture) Dynamic CGE Bchir and Fouqin (2006) ASEAN India (industrial) Dynamic CGE Bchir and Fouqin (2006) ASEAN India (agriculture) Dynamic CGE

17 42 The Korea Economic Institute Kim et al. (2005) European Union Dynamic CGE Choi and Schott (2001) United States Dynamic CGE Lee J. and Lee H. (2005) United States Dynamic CGE Schott, Bradford, and Moll (2006) United States Dynamic CGE 6.60 Kiyota and Stern (2006) United States Dynamic CGE 8.05 Source: Data compiled by authors. Notes: CGE = computable general equilibrium; IRS = Increasing returns to scale; MFN = most-favored-nation a. Cited in Lee C. et al. (2005). Does not allow for full capital mobility. b. Assumes an FTA-induced productivity growth of 10 percent to 30 percent of total productivity growth, depending on the outlook of the sector. c. Assumes an FTA-induced annual productivity growth of 1 percent for 10 years. d. The model permits the incorporation of rational expectations and forward-looking intertemporal behavior on the part of individual.

18 Static and Dynamic Consequences of a KORUS FTA 43 The second observation from Table 5 is that South Korea could go a long way toward the benefits from global free trade through agreements that create free trade with Japan and China, ASEAN, the European Union, and the United States. According to Scollay and Gilbert (2001), for example, the successful conclusion of ASEAN + 3 would yield gains for South Korea that could be equivalent to 65 percent of total gain under global free trade (as they estimate them). Kim et al. (2005), Schott, Bradford, and Moll (2006), and Kiyota and Stern (2007) show that South Korea could also obtain very significant gains from FTAs with the European Union and the United States. Given that South Korea has already signed an FTA with ASEAN (goods only) and the United States (not yet ratified), and given that talks are ongoing with the European Union and with Japan and China (under the ASEAN + 3 framework), South Korea may well obtain most of the benefits of free trade in the near term especially if the Doha Round is brought to a successful conclusion. Estimates from Gravity Models In the past decade, empirical and theoretical research has revived the gravity model, which posits that trade between two countries is directly proportional to their size and inversely proportional to their distance. 15 The gravity model has been augmented, notably by Frankel (1997) and Rose (2003, 2004), to quantify the effects of common language, shared borders, postcolonial relationships, and other variables that may affect bilateral trade. Rose (2003) used the gravity model to test the influence of international institutions (the GATT and WTO, IMF, OECD, and regional or bilateral FTAs) on bilateral trade (Table 6). 16 Using IMF data on trade between 178 countries over the period , he reported that participation in a regional FTA can increase bilateral trade by 118 percent. 17 Park and Kang (2004, table 6), reach a central estimate very close to Rose (2003). By contrast, DeRosa and Gilbert (2007) introduce several modifications to Rose s (2003) original work importantly, they consider a wider number of bilateral and regional FTAs and obtain a smaller FTA effect, a gain of approximately 62 percent. 15. The gravity model was used as early as 1946 to analyze trade. Linnemann (1966) provided significant refinements to the technique. 16. Rose has made his data set publicly available at The 118 percent coefficient may also be exaggerated because of selection bias the tendency of countries to put priority on FTAs with partners that promise the largest trade gains. In the case of the United States, however, noneconomic factors appear to play a large role in the selection of FTA partners (e.g., Israel, Morocco, Bahrain, and Oman); see Schott (2004, ).

19 44 The Korea Economic Institute Table 6: Gravity Models: Insights into Possible Effects of South Korea s Free Trade Agreements Authors Proposed Type of Insights on trade Other partner model expansion results Gravity models applied to South Korea s negotiating scenarios Sohn and Japan GM Ratio of actual trade to Yoon (2001); potential trade is 67% (1995). Kim (2002) Implied 50% increase of two-way trade with FTA. Lee C. and Asian GM with No evidence of an emerging Park S. (2002) regionalism tariffs and North Asian trading bloc. regional ASEAN + 3 is emerging as dummy a natural trading bloc in the region. They also find evidence of a special relation between ASEAN and the United States. Choi and United States GM U.S.-Korea trade is already Schott (2001) 300% of model prediction. Trade with China and Japan is below model prediction. DeRosa and United States GM with 27% increase in South Korean Gilbert (2004) regional exports to the United States dummy only. Park and United States GM with (see below) U.S.-South Kang (2004) FDI dummy Korea FTA could expand U.S. FDI into South Korea by 14% to 35%, and expand FDI in in South Korea from third parties by 28% to 35%. DeRosa and Global free GM with 54% increase in South Korean Gilbert (2004) trade regional total exports to the world. dummy Gravity models applied to capture typical impact of FTAs Rose (2003) FTAs of GM with 118% increase in bilateral several regional trade between members. countries dummy Park and FTAs of GM with 103% 135% increase in Kang (2004) several regional bilateral trade between countries dummy members. DeRosa and FTAs of GM with 62% increase in bilateral trade Gilbert (2007) several regional between members. countries dummy Source: Compiled from authors data. Note: FDI = foreign direct investment; FTA = free trade agreement; GM = gravity model

20 Static and Dynamic Consequences of a KORUS FTA 45 We adjust these coefficients downward to take account of trade diversion. Bilateral and regional FTAs increase trade among members through both trade creation (increased trade as a result of relative efficiency) and trade diversion (increased trade as a result of privileged access). Estimates by DeRosa and Gilbert (2004, table A.6) for 14 prospective U.S. FTAs enacted simultaneously suggest that trade diversion measured as the dollar decline in nonpartner trade could account for up to 25 percent of the gross gain in trade with partner countries. In view of this estimate, we reduce the gravity model trade augmentation effect by 25 percent, from 118 percent to 89 percent in the case of estimates by Rose (2003) and from 62 percent to 47 percent in the case of estimates by DeRosa and Gilbert (2007). 18 We use the range of adjusted coefficients to estimate the potential increase in South Korean merchandise trade if FTAs were concluded with all trading partners. This is the same as assuming that all countries eliminated their policy barriers to merchandise trade with South Korea, and vice versa. For simplicity purposes, we use trade data of 2005 that do not reflect the impact of bilateral agreements, with the minor exception of the South Korea Chile FTA. Table 7 drives the reader through the arithmetic. Because we assume no existing preferential trade for South Korea in 2005, bilateral trade of South Korea expands between 47 and 89 percent, leading to a comparable change in the trade exposure ratio. Applying the OECD (2003) estimate for the elasticity of trade exposure to per capita income of 0.2, we obtain a range of estimates for South Korea s possible GDP gains, ranging between 9 percent and 18 percent. 19 The true value may be found somewhere within this range, but we prefer closer to the lower bound estimate because the DeRosa and Gilbert (2007) coefficient is estimated from a larger and more recent database. A GDP gain of 9 percent from future liberalization (about $57 billion based on 2005 GDP measured in 2000 dollars) would translate into per capita benefits of about $1,200 (2000 dollars). 18. All these coefficients apply to two-way bilateral trade (merchandise imports plus exports). DeRosa and Gilbert (2004) performed a slightly different calculation based on exports only. They conclude that, under global free trade, South Korean exports could expand by 54 percent. This estimate suggests that export gains could be considerable. 19. Unlike in our estimates of past gains, we now feel comfortable in using the OECD (2003) estimate because there is no question that South Korea is and will remain a developed country.

21 46 The Korea Economic Institute Table 7: South Korean Merchandise Trade under Global Free Trade, 2005, billions of nominal U.S. dollars South Korean trade Total (2005) With world FTA Lower bound Upper bound South Korean imports n.a. n.a. South Korean exports n.a. n.a. Total trade ,948.9 Memorandum Base Nominal GDP Trade exposure Change in trade exposure Change in GDP 9.4% 17.8% Sources: UN (2005), IMF-WEO (2005), Rose (2003), DeRosa and Gilbert (2007), and authors calculations. Notes: Modeled by assuming worldwide FTAs in a gravity model framework. FTA = free trade agreement. IV. Conclusion Without a doubt, South Korea ranks among the great economic successes of all time. From shambles at the end of the Second World War, South Korea has risen like a phoenix to approach the first rank of nations. Past participation in the process of globalization, abetted by both policy liberalization and technological innovation, clearly contributed to Korea s enormous success. Future participation in the global economy, through complete liberalization of remaining barriers, promises even greater gains.

22 Static and Dynamic Consequences of a KORUS FTA 47 REFERENCES Bchir, Mohamed Hedi, and Michel Fouquin Economic Integration in Asia: Bilateral Free Trade Agreements versus Asian Single Market. Paris: Centre d Etudes Prospectives et d Informations Internationales (CEPII) /wp06-15.pdf. Berg, Andrew, and Anne Krueger Trade, Growth and Poverty: A Selective Survey. IMF Working Paper no. WP/03/30. Washington, D.C.: International Monetary Fund. Bergsten, C. Fred The United States and the World Economy: Foreign Economic Policy for the Next Decade. Washington, D.C.: Institute for International Economics. Bernard, Andrew, Jonathan Eaton, J. Bradford Jensen, and Samuel Kortum Plants and Productivity in International Trade. American Economic Review 93, no. 4 (September): Bradford, Scott C., Paul L. E. Grieco, and Gary Clyde Hufbauer The Payoff to America from Global Integration. In The United States and the World Economy: Foreign Economic Policy for the Next Decade, ed. C. Fred Bergsten. Washington, D.C.: Institute for International Economics. Broda, Christian, and David Weinstein Globalization and the Gains from Variety. Quarterly Journal of Economics 121: Broda, Christian, Joshua Greenfield, and David Weinstein From Groundnuts to Globalization: A Structural Estimate of Trade and Growth. NBER Working Paper no Cambridge, Mass.: National Bureau of Economic Research. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern Developing Countries Stake in the Doha Round. Research Seminar in International Economics Discussion Paper no University of Michigan School of Public Policy. rsie/workingpapers/papers /r495.pdf. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern Multilateral, Regional, and Bilateral Trade-Policy Options for the United States and Japan. World Economy 26(6): Cheong Inkyo Economic Integration in Northeast Asia: Searching for a Feasible Approach. Working Paper Seoul: Korea Institute for International Economic Policy. December. Cheong Inkyo A Korea-Japan FTA: Economic Effects and Policy Implications. venus.icre.go.kr/metadata/10045_ikjung.doc. Cheong Inkyo Regionalism and Free Trade Agreements in East Asia. Asian Economic Papers 2, no. 2, Choi Inbom and Jeffrey J. Schott Free Trade between Korea and the United States? Washington, D.C.: Institute for International Economics.

23 48 The Korea Economic Institute Cline, William Trade Policy and Global Poverty. Washington, D.C.: Institute for International Economics. Dent, Christopher The Foreign Economic Policies of Singapore, South Korea, and Taiwan. Northampton, Mass.: E. Elgar. DeRosa, Dean A., and John P. Gilbert Quantitative Estimates of the Economic Impacts of U.S. Bilateral Free Trade Agreements. In Free Trade Agreements: U.S. Strategies and Priorities, ed. Jeffrey Schott. Washington D.C.: Institute for International Economics. DeRosa, Dean A., and John P. Gilbert Predicting Trade Expansion under FTAs and Multilateral Agreements. Working Paper no Washington, D.C. October.: Institute for International Economics. DeRosa, Dean A., and John P. Gilbert Appendix A: Technical Aspects of the Gravity and Computable General Equilibrium Models. In Toward a U.S.-Indonesia Free Trade Agreement, ed. Gary Clyde Hufbauer and Sjamsu Rahardja. Washington D.C.: Institute for International Economics. Feenstra, Robert C New Product Varieties and the Measurement of International Prices. American Economic Review 84, no. 1 (March): Feenstra, Robert New Evidence on the Gains from Trade. Review of World Economics/ Weltwirtschaftliches Archiv. December. Feenstra, R. C., and H. L. Kee Export Variety and Country Productivity: Estimating the Monopolistic Competition Model with Endogenous Productivity. NBER Working Paper no Cambridge, Mass.: National Bureau of Economic Research. Feenstra, R. C., D. Madani, T. Yang, and C. Liang Testing Endogenous Growth in South Korea and Taiwan. Journal of Development Economics 60: Frankel, Jeffrey A Regional Free Trading Blocs in the World Economic System. Washington, D.C.: Institute for International Economics. Frankel, Jeffrey A., and David Romer Trade and Growth: An Empirical Investigation. American Economic Review 89, no. 3 (June): Hallak, Juan Carlos, and James Levinsohn Fooling Ourselves: Evaluating the Globaliization and Growth Debate. NBER Working Paper no Cambridge, Mass.: National Bureau of Economic Research. IDE/JETRO (Institute of Developing Economies/Japan External Trade Organization) Toward Closer Japan-Korea Economic Relations in the 21st Century. Tokyo: JETRO. March. IMF (International Monetary Fund) IFS. Various years. International Financial Statistics (IFS). Washington, D.C.: IMF.

24 Static and Dynamic Consequences of a KORUS FTA 49 IMF (International Monetary Fund) WEO World Economic Outlook (WEO) Database. Washington, D.C.: IMF. Kawasaki, Kenichi The Impact of Free Trade Agreements in Asia. RIETI Discussion Paper Series no. 03-E-018. Tokyo: Research Institute of Economy, Trade, and Industry. Kim Heung-chong Has Trade Intensity in ASEAN + 3 Really Increased? Evidence from a Gravity Analysis. Working Paper no Seoul: Korea Institute for International Economic Policy. Kim Heung-chong et al An Analysis on the Economic Effects of a Korea-EU FTA and Policy Implications on the Korean Economy. Seoul: Korea Institute for International Economic Policy. Kiyota, Kozo, and Robert M. Stern Economic Effects of Korea-U.S. Free Trade Agreement. Washington, D.C.: Korea Economic Institute. Lee Chang-jae et al The Economic Effects of a China-Japan-Korea FTA and Implications for the Korean Economy. Seoul: Korea Institute for International Economic Policy. Lee Chang-soo and Park Soon-chan An Examination of the Formation of Natural Trading Blocs in East Asia. Working Paper no Seoul: Korea Institute for International Economic Policy. Lee Hong-shik et al Economic Effects of a Korea-China FTA and Policy Implications (I). Seoul: Korea Institute for International Economic Policy. Lee Jun-yu and Lee Hong-shik Feasibility and Economic Effects of a Korea-U.S. FTA. Seoul: Korea Institute for International Economic Policy. Linnemann, Hans An Econometric Study of International Trade Flows. Amsterdam: North-Holland. Maddison, Angus The World Economy: Historical Statistics. Paris: Development Centre of the Organization for Economic Cooperation and Development. McDaniel, Christine, and Alan Fox U.S.-Korea FTA: The Economic Impact of Establishing a Free Trade Agreement (FTA) between the United States and the Republic of Korea. Investigation no ; USITC publication no Washington, D.C.: United States International Trade Commission. McKibbin, Warwick J., Lee Jong-wha, and Cheong Inkyo A Dynamic Analysis of a Korea-Japan Free Trade Area: Simulations with the G-Cubed Asia-Pacific Model. Working Paper no Seoul: Korea Institute for International Economic Policy. Nakajima, Tomoyoshi An Analysis of the Economic Effects of Japan-Korea FTA: Sectoral Aspects. ERINA Discussion Paper No.0202e. Niigata, Japan: Economic Research Institute for Northeast Asia.

25 50 The Korea Economic Institute Nam Young-sook et al Economic Effects of Korea-China FTA and Its Main Issues. Seoul: Korea Institute for International Economic Policy. Ng, Francis K. T Data on Trade and Import Barriers. Washington, D.C.: World Bank. OECD (Organization for Economic Cooperation and Development) The Sources of Economic Growth in OECD Countries. Paris: OECD. Park Soon-chan and Kang Moon-soong Korea-U.S. FTA: Trade and Investment Creation Effects and Trade Structure. Seoul: Korea Institute for International Economic Policy. Productivity Commission. Various years. Measures on Restrictions on Trade in Services Database. Melbourne: Government of Australia, Productivity Commission. Richardson, J. David Sizing Up the Micro-Data Benefits. Institute for International Economics, Washington, D.C. Unpublished manuscript. Rose, Andrew K Which International Institutions Promote International Trade? Manuscript accepted for publication in Review of International Economics. faculty.haas.berkeley.edu/arose/comparer.pdf. Rose, Andrew K Response to Subramanian and Wei. arose/swresponse.pdf. Schott, Jeffrey J Assessing U.S. FTA Policy. In Free Trade Agreements: U.S. Strategies and Priorities, ed. Jeffrey J. Schott. Washington, D.C.: Institute for International Economics. Schott, Jeffrey J., Scott C. Bradford, and Thomas Moll Negotiating the Korea United States Free Trade Agreement. Washington, D.C.: Institute for International Economics. Scollay, Robert, and John P. Gilbert New Regional Trading Arrangements in the Asia Pacific? Washington, D.C.: Institute for International Economics. Sohn Chan-Hyun and Yoon Jinna Does the Gravity Model Fit Korea s Trade Pattern? Implications for Korea s FTA Policy and North-South Korean Trade. Working Paper no Seoul: Korea Institute for International Economic Policy. UN (United Nations) Commodity Trade Statistics Database (Comtrade). New York: UN. Wang Yunjong and Cheong Inkyo The Economic Effects of US- Korea FTA [in Korean]. World Bank. Various years. World Development Indicators. Washington, D.C.: World Bank. WTO (World Trade Organization) Statistics Database: Trade Profiles. Geneva: World Trade Organization. (accessed July 2007).

26