Abstract
This Selected Issues paper for Australia highlights the dynamics of the Australian real exchange rate and its impact on Australia’s trade. The main findings are that the Australian real exchange rate is largely driven by world commodity prices and that it adjusts relatively rapidly to large shocks, with an estimated half-life of 16 months. The real exchange rate is a significant determinant of Australian imports, with an elasticity of one, but does not appear to have a significant impact on Australian exports.
II. Trade Liberalization and Economic Growth: Australia’s Experience8
A. Introduction and Summary
47. While the benefits of free trade have long been established by economic theory, the difficulty of measuring them has kept the policy debate alive. Numerous empirical studies have found a positive effect of trade openness on economic growth, only to be challenged by other studies questioning the robustness of the results. Measurement problems, reverse causality, and the sensitivity of the results to model specification are at the core of the controversy. More recent research has focused on addressing these problems and a consensus on the positive link between free trade and economic growth seems to be emerging.
48. The Australian experience with trade liberalization provides an interesting case study on the benefits of trade reform. Prior to the remarkable economic growth performance of the past decade, the Australian economy had performed poorly for more than two decades due to high tariffs, heavy product market regulation, centralized labor market institutions, and high inflation. The turn around in Australia’s growth performance began with the progressive opening of the economy. The lowering of barriers to foreign competition played a catalyst role for a sustained and comprehensive reform in the labor and product markets, which has significantly enhanced the competitiveness and the flexibility of the Australian economy. As a result, total factor productivity (TFP) accelerated, lifting average growth in GDP per capita (in purchasing power parity (PPP) terms) from 1.8 percent during 1970-85—the pre-reform period—to 2.4 percent during the past 8 years. Furthermore, business cycle fluctuations have been dampened, keeping actual growth close to potential.
Labor and Productivity Indicators
(Average growth rates, in percent)
GDP per hour worked.
Based on ‘harmonised’ price indices for information and communication technology capital goods.
Labor force as a percent of population aged 15–64.
Labor and Productivity Indicators
(Average growth rates, in percent)
| GDP per capita | Labor Productivity 1/ | TFP 2/ | Labor Force Participation Rate 3/ | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1970-85 | 1985-95 | 1995-02 | 1970-85 | 1985-95 | 1995-02 | 1985-95 | 1995-02 | 1970-85 | 1985-95 | 1995-02 | |
| Australia | 1.8 | 1.9 | 2.4 | 1.8 | 1.3 | 2.5 | 1.2 | 2.1 | 60.9 | 61.9 | 62.9 |
| Canada | 2.1 | 1.3 | 2.2 | 1.7 | 0.9 | 1.8 | 0.0 | 1.0 | 63.6 | 66.2 | 65.5 |
| United Kingdom | 1.7 | 2.3 | 2.2 | 2.7 | 2.0 | 2.0 | 1.2 | 0.9 | 62.0 | 63.0 | 62.8 |
| United States | 2.4 | 2.0 | 1.7 | 1.6 | 1.2 | 2.0 | 0.8 | 1.1 | 62.3 | 66.0 | 66.9 |
GDP per hour worked.
Based on ‘harmonised’ price indices for information and communication technology capital goods.
Labor force as a percent of population aged 15–64.
Labor and Productivity Indicators
(Average growth rates, in percent)
| GDP per capita | Labor Productivity 1/ | TFP 2/ | Labor Force Participation Rate 3/ | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1970-85 | 1985-95 | 1995-02 | 1970-85 | 1985-95 | 1995-02 | 1985-95 | 1995-02 | 1970-85 | 1985-95 | 1995-02 | |
| Australia | 1.8 | 1.9 | 2.4 | 1.8 | 1.3 | 2.5 | 1.2 | 2.1 | 60.9 | 61.9 | 62.9 |
| Canada | 2.1 | 1.3 | 2.2 | 1.7 | 0.9 | 1.8 | 0.0 | 1.0 | 63.6 | 66.2 | 65.5 |
| United Kingdom | 1.7 | 2.3 | 2.2 | 2.7 | 2.0 | 2.0 | 1.2 | 0.9 | 62.0 | 63.0 | 62.8 |
| United States | 2.4 | 2.0 | 1.7 | 1.6 | 1.2 | 2.0 | 0.8 | 1.1 | 62.3 | 66.0 | 66.9 |
GDP per hour worked.
Based on ‘harmonised’ price indices for information and communication technology capital goods.
Labor force as a percent of population aged 15–64.
49. This paper provides a quantitative assessment of the potential gains from structural reforms, with a focus on trade liberalization. It is divided into four sections. The second section briefly reviews the large literature on the link between trade openness and economic growth and the third section focuses on Australia’s experience with trade liberalization. The fourth section attempts to quantify the impact of trade liberalization on economic growth by estimating a reduced form growth equation using panel data with newly constructed indicators of structural reforms.
50. The main conclusion from the literature review is that policies that eliminate barriers to trade yield economic benefits that generally outweigh their short-term costs, although there remains considerable uncertainty on the magnitude of these benefits. In the case of Australia, the econometric results show that trade liberalization and the resulting increase in the degree of openness of the economy may have lifted annual GDP per capita growth by about ¾ percentage point and that there is scope for additional significant gains from further reforms. The econometric analysis also finds strong complementarities between trade liberalization and labor and product market reforms.
B. Trade Openness and Economic Growth: A Review of the Empirical Evidence
51. A considerable body of literature has highlighted the importance of trade liberalization in facilitating economic growth. The dominant view is that trade liberalization stimulates growth. A few influential studies, however, have questioned the robustness of the relationship between trade liberalization and economic growth, generating a heated debate on the benefits of trade liberalization (Table 1).
Studies on Trade Openness, Growth, and Productivity
Studies on Trade Openness, Growth, and Productivity
| Studies | Methods | Key Findings |
|---|---|---|
| Sachs and Warner (1995) | OLS Panel of 79 Countries, 1970–89. | Dummy variable for trade openness is highly significant. Open economies grow, on average, by 2½ percentage points more than the closed economies. |
| Coe, and Helpman (1995) and Coe, Helpman, and Hoffmaister (1997) | WLS panel of 22 advanced economies and of 77 developing countries, 1971-90. | Total factor productivity is positively correlated with foreign R & D. Economies that are more open to imports derive larger benefit from foreign R & D. |
| Edwards (1998) | WLS and Instrumented WLS. Test 9 alternative measure of openness using panel data ranges from 38 to 87 countries and slightly different time-period. | More open economies tend to have faster productivity growth than more closed economies; results are robust to the use of different openness indicators, estimation techniques, time-periods, and functional forms. |
| Chand (1999) | OLS panel of 2 digit ANZSIC, Australian Manufacturing sectors, 1967-68 to 1995-95. | Using the most conservative estimate, a 1 percent drop in the nominal rate of protection leads, on average, to 0.2 percent rise in total factor productivity. |
| Rodriguez and Rodrik (1999) | OLS Panel of 74 Countries, 1970-89. Individual components of Sachs and Warner’s Index are used in the in a growth regression. | The black market premium and existence or absence of state monopolies in the export sector are statistically significant, whereas other components of the index—average import tariffs, tariff barriers, and whether or not the economy in question is socialist—are not. Introduction of institutional quality index reduces the value and significance of trade openness. |
| Frankel and Romer (1999) | OLS and IV cross-section of 98 countries, 1985. | IV estimates of the effect of trade on income are larger than OLS estimates. IV estimates imply that one percentage point increase in the ratio of trade to GDP raises income per capita by between one- half to two percent. |
| Rodrik, Subramanian, and Trebbi (2002) | OLS and IV cross-section of 64, 80, and 140 countries, 1995. | Once the institutional the variable is added, the estimated coefficients on geography and openness have the wrong sign. IV estimate of the effect of institutions on income is nearly three times as large as the OLS estimates. |
| Wacziarg and Welch (2003) | OLS; Panel of 116 Countries 1990-99 and fixed- effects regression 1950-98. | Sachs and Warner’s results break down completely for the 1990s. Countries that liberalize trade regime experience, on average, 1½ percentage points higher growth compared with pre- liberalization period. |
| Dollar and Kraay (2004) | OLS and IV cross-section of 187 countries, 1980s and 1990s. | IV estimate of the effect of trade on income is larger than OLS estimates. IV estimates indicate that a 100 percent increase in a country’s trade share raises incomes by 48 percent over a decade. |
| IMF (2004) | GMM panel of 15 OECD Countries, 1984-95. | Cumulative gains from structural reform in trade, product market, and labor market are positive, but they predominantly materialize in the long run. |
Studies on Trade Openness, Growth, and Productivity
| Studies | Methods | Key Findings |
|---|---|---|
| Sachs and Warner (1995) | OLS Panel of 79 Countries, 1970–89. | Dummy variable for trade openness is highly significant. Open economies grow, on average, by 2½ percentage points more than the closed economies. |
| Coe, and Helpman (1995) and Coe, Helpman, and Hoffmaister (1997) | WLS panel of 22 advanced economies and of 77 developing countries, 1971-90. | Total factor productivity is positively correlated with foreign R & D. Economies that are more open to imports derive larger benefit from foreign R & D. |
| Edwards (1998) | WLS and Instrumented WLS. Test 9 alternative measure of openness using panel data ranges from 38 to 87 countries and slightly different time-period. | More open economies tend to have faster productivity growth than more closed economies; results are robust to the use of different openness indicators, estimation techniques, time-periods, and functional forms. |
| Chand (1999) | OLS panel of 2 digit ANZSIC, Australian Manufacturing sectors, 1967-68 to 1995-95. | Using the most conservative estimate, a 1 percent drop in the nominal rate of protection leads, on average, to 0.2 percent rise in total factor productivity. |
| Rodriguez and Rodrik (1999) | OLS Panel of 74 Countries, 1970-89. Individual components of Sachs and Warner’s Index are used in the in a growth regression. | The black market premium and existence or absence of state monopolies in the export sector are statistically significant, whereas other components of the index—average import tariffs, tariff barriers, and whether or not the economy in question is socialist—are not. Introduction of institutional quality index reduces the value and significance of trade openness. |
| Frankel and Romer (1999) | OLS and IV cross-section of 98 countries, 1985. | IV estimates of the effect of trade on income are larger than OLS estimates. IV estimates imply that one percentage point increase in the ratio of trade to GDP raises income per capita by between one- half to two percent. |
| Rodrik, Subramanian, and Trebbi (2002) | OLS and IV cross-section of 64, 80, and 140 countries, 1995. | Once the institutional the variable is added, the estimated coefficients on geography and openness have the wrong sign. IV estimate of the effect of institutions on income is nearly three times as large as the OLS estimates. |
| Wacziarg and Welch (2003) | OLS; Panel of 116 Countries 1990-99 and fixed- effects regression 1950-98. | Sachs and Warner’s results break down completely for the 1990s. Countries that liberalize trade regime experience, on average, 1½ percentage points higher growth compared with pre- liberalization period. |
| Dollar and Kraay (2004) | OLS and IV cross-section of 187 countries, 1980s and 1990s. | IV estimate of the effect of trade on income is larger than OLS estimates. IV estimates indicate that a 100 percent increase in a country’s trade share raises incomes by 48 percent over a decade. |
| IMF (2004) | GMM panel of 15 OECD Countries, 1984-95. | Cumulative gains from structural reform in trade, product market, and labor market are positive, but they predominantly materialize in the long run. |
52. More recent studies have attempted to settle the controversy by addressing measurement and endogeneity bias problems in the earlier research:
A recent literature review by Berg and Krueger (2003) argues that recent refinements to the measurement of trade openness and model specification appear to have strengthened the robustness of the link between trade openness and economic growth. They note, however, that trade openness is only one of many determinants of economic growth, and that trade openness is likely to be highly correlated with other determinants, making it difficult to disentangle the effect of trade openness from the effect of policies that typically accompany trade reform, such as more stable macroeconomic policies and other domestic market reforms.
By using decadal difference regressions, Dollar and Kraay (2004) find a strong correlation between changes in trade volumes and changes in economic growth.9 The results are robust to the inclusion of variables representing the quality of institutions, suggesting that the estimated trade openness coefficient is not simply capturing the overall quality of the growth environment.
A recent econometric study by the IMF (2004) found that trade liberalization has a positive impact on GDP per capita growth in the long term, although its short-term effect could be negative. To address the issue of omitted variables, the study includes time series of structural policy indicators in five areas—the financial sector, international merchandise trade, labor market, product markets, and the tax system—to analyze the dynamic benefits of structural reforms in OECD countries. The study corrects for the potential endogeneity bias by estimating dynamic growth equations with the Generalized Method of Moments developed by Arellano and Bond (1991).
53. Some studies have focused on the channels through which trade reform may affect economic growth. Coe and Helpman (1995) found that developed countries with higher import propensities tend to have higher productivity, reflecting a significant interaction between imports and the ability to benefit from foreign R&D. Using the same methodology, Coe, Helpman, and Hoffmaister (1997) also found similar results for developing countries. Hakura and Jaumotte (1999) examined the role of inter- and intra-industry trade in technology diffusion and concluded that intra-industry trade tends to stimulate more technology transfer than inter-industry trade.
54. Unlike econometric studies, which are based on reduced form growth equations, analyses that are based on computable general equilibrium (CGE) models allow a better understanding of the channels through which trade affects economic growth. CGE analyses also address the reverse causality problem. Moreover, by capturing inter-industry linkages, disaggregated CGE models are well equipped to assess distributional effects, adjustment costs, and other important aspects of trade liberalization. Dynamic CGE models can also differentiate between short- versus long-term productivity gains associated with a particular policy. A number of researchers have used CGE models, which of course are sensitive to model specification and parameter calibration, to quantify the impact of trade liberalization under different frameworks, including unilateral arrangements, multilateral arrangements, regional arrangements, and bilateral free trade agreements.
55. While CGE-based analyses generally find a positive impact of trade liberalization on economic growth, there remains considerable uncertainty about the size of the welfare gain (Table 2). The divergence in the findings can be attributed to differences in model structure, experimental design, and trade elasticities. Dynamic models, by capturing dynamic gains in productivity, tend to predict larger gains from trade liberalization than static models.
CGE Studies on the Impact of Tariff Reduction
CGE Studies on the Impact of Tariff Reduction
| Studies | Experiments | Model and Features | Key Findings |
|---|---|---|---|
| Dixon et al (1986) | Short-run effects of 50 percent across-the- board cut in protection in Australia | ORANI: static, single country | Significant reduction in output of import-competing sectors (textile, clothing, footwear, and motor vehicles) are more than offset by increasing output of exporting sectors (Coal, iron and meat products). The short-run net positive gain is 0.2 to 0.4 percent of GDP. |
| Hartel and Martin (1999) | WTO, Compare 40 percent cut in manufacturing tariffs worldwide with the same cut in agriculture tariffs | GTAP: Version 4, static, multi- countries | Total gain from tariff cuts in manufacturing is US$380 billion, with three-quarters accruing to developing countries. Total gain of the same cut in agricultural protection is US$69 billion, with less than one-fourth accruing to developing countries. |
| Productivity Commission (2000) | Removing all tariffs that are under 5 percent | MONASH: recursive dynamic, single country | Initially, the gain is A$240 million (measured in current dollar), 0.04 percent of GDP. By 2010, the gain with productivity improvement in the affected industries is A$480 million, 0.08 percent of GDP. Without productivity improvement, GDP gain reduces to A$120 million. |
| Scollay and Gilbert (2000) | APEC liberalization in agriculture and food products. | GTAP: Version 4, static, multi- countries | Total gain ranges from US$56 to US$118 billion depending on liberalization scenarios. Big countries (Japan, China, and USA) receive the largest welfare gains in absolute terms. Canada experiences welfare losses in all scenarios, while Mexico, Indonesia, and Malaysia also experience losses if non-APEC members do not reciprocate. |
| McKibbin et al (2000) | Free Trade Area between AFTA (ASEAN Free Trade Area) and CER (The Australian and New Zealand Closer Economic Relations) | APG-Cubed: dynamic, multi- countries | Total gain of AFTA-CER FTA is US$48 billion over the period of 2000-20, US$26 billion for AFTA and US22 billion for CER. All member countries gain from the arrangements. Indonesia, the Philippines, and Thailand gain more due the size of the initial barrier and direction of trade. |
| Center for International Economics (2004) | Free Trade Area between Australia and the United States (AUSFTA) | APG-Cubed and GTAP | Australia’s GDP increases by US$6.1 billion per year. Investment liberalization makes the biggest contribution to overall gain. |
| Hartel et al (2004) | Free Trade Arrangement of the Americas (FTAA) | GTAP: Version 5, static with sensitivity analysis | Welfare effects are positive for ten of the thirteen FTAA members, but are negative for the rest of the world. In absolute terms, the United States receives the largest gain. |
CGE Studies on the Impact of Tariff Reduction
| Studies | Experiments | Model and Features | Key Findings |
|---|---|---|---|
| Dixon et al (1986) | Short-run effects of 50 percent across-the- board cut in protection in Australia | ORANI: static, single country | Significant reduction in output of import-competing sectors (textile, clothing, footwear, and motor vehicles) are more than offset by increasing output of exporting sectors (Coal, iron and meat products). The short-run net positive gain is 0.2 to 0.4 percent of GDP. |
| Hartel and Martin (1999) | WTO, Compare 40 percent cut in manufacturing tariffs worldwide with the same cut in agriculture tariffs | GTAP: Version 4, static, multi- countries | Total gain from tariff cuts in manufacturing is US$380 billion, with three-quarters accruing to developing countries. Total gain of the same cut in agricultural protection is US$69 billion, with less than one-fourth accruing to developing countries. |
| Productivity Commission (2000) | Removing all tariffs that are under 5 percent | MONASH: recursive dynamic, single country | Initially, the gain is A$240 million (measured in current dollar), 0.04 percent of GDP. By 2010, the gain with productivity improvement in the affected industries is A$480 million, 0.08 percent of GDP. Without productivity improvement, GDP gain reduces to A$120 million. |
| Scollay and Gilbert (2000) | APEC liberalization in agriculture and food products. | GTAP: Version 4, static, multi- countries | Total gain ranges from US$56 to US$118 billion depending on liberalization scenarios. Big countries (Japan, China, and USA) receive the largest welfare gains in absolute terms. Canada experiences welfare losses in all scenarios, while Mexico, Indonesia, and Malaysia also experience losses if non-APEC members do not reciprocate. |
| McKibbin et al (2000) | Free Trade Area between AFTA (ASEAN Free Trade Area) and CER (The Australian and New Zealand Closer Economic Relations) | APG-Cubed: dynamic, multi- countries | Total gain of AFTA-CER FTA is US$48 billion over the period of 2000-20, US$26 billion for AFTA and US22 billion for CER. All member countries gain from the arrangements. Indonesia, the Philippines, and Thailand gain more due the size of the initial barrier and direction of trade. |
| Center for International Economics (2004) | Free Trade Area between Australia and the United States (AUSFTA) | APG-Cubed and GTAP | Australia’s GDP increases by US$6.1 billion per year. Investment liberalization makes the biggest contribution to overall gain. |
| Hartel et al (2004) | Free Trade Arrangement of the Americas (FTAA) | GTAP: Version 5, static with sensitivity analysis | Welfare effects are positive for ten of the thirteen FTAA members, but are negative for the rest of the world. In absolute terms, the United States receives the largest gain. |
56. Overall, the empirical evidence supports a positive relationship between trade openness and economic growth. Although regression analyses suffer from some methodological limitations, including the difficulty of capturing the complex channels through which reforms affect growth, the bulk of the evidence suggests that trade openness is positively correlated with growth. Similarly, CGE-based studies generally find positive welfare effects from trade liberalization, although the size of the gain may be quite sensitive to model specification and calibration. These studies also find that other supporting policies, such as product and labor market reforms, enhance the benefits of trade liberalization.
C. Has Australia Gained from Opening Up Its Economy?
57. Prior to the recent episode of remarkable economic growth, the Australian economy had performed poorly for more than two decades. From the 1970s through the 1980s, economic growth rates were significantly lower than those of other OECD countries, causing Australia’s income rank to slip steadily. According to the Parham (2002), Australia’s GDP per capita fell from the 7th highest among OECD countries in 1960 to the 15th highest in 1990.
58. High tariffs, extensive product market regulation, centralized labor market institutions, and high inflation contributed to Australia’s poor growth performance. A study by the Economic Planning Advisory Council (1996) found that much of Australia’s below-average growth during the 1970s and 1980s was due to poor productivity performance generated by the high-tariff regime. Historically, Australia was a less open economy than many other OECD countries.10 Together with the extensive product market regulations and the centralized wage-bargaining system, trade protection supported the growth of high-cost, domestic-oriented industries. Successive rounds of tariff increases proved necessary to maintain the competitive position of many of these inefficient industries. At the same time, the interaction between the centralized wage-bargaining system and the commodity price cycle complicated the task of managing inflation.
59. The improvement in Australia’s growth performance began with the progressive opening of the economy. It is not surprising that Australia’s economic performance improved once the failed policies of the past were reversed. The first wave of reforms began with a gradual trade liberalization in the mid 1980s, reducing almost all tariffs to a maximum of 15 percent.11 Subsequently, the maximum tariff was reduced further to 5 percent by 1996, with an exception made for passenger motor vehicles and textile, clothing, and footwear. As a result, average tariff on imports declined from 10 percent in 1985 to about 3½ percent in 2001.
FALLING RATES OF ASSISTANCE TO MANUFACTURING 1/
1986-96 -2000-01
Citation: IMF Staff Country Reports 2004, 354; 10.5089/9781451802061.002.A002
Source: Productivity Commission-2000.1/ Breaks in the series reflect periodic revisions to industry input and output tables used in these estimates.FALLING RATES OF ASSISTANCE TO MANUFACTURING 1/
1986-96 -2000-01
Citation: IMF Staff Country Reports 2004, 354; 10.5089/9781451802061.002.A002
Source: Productivity Commission-2000.1/ Breaks in the series reflect periodic revisions to industry input and output tables used in these estimates.FALLING RATES OF ASSISTANCE TO MANUFACTURING 1/
1986-96 -2000-01
Citation: IMF Staff Country Reports 2004, 354; 10.5089/9781451802061.002.A002
Source: Productivity Commission-2000.1/ Breaks in the series reflect periodic revisions to industry input and output tables used in these estimates.60. The lowering of barriers to foreign competition played a catalyst role for a sustained and comprehensive domestic reform program. The reform of anti-competitive business regulations and industrial relations, culminating in the adoption of the National Competition Policy in 199512 , can be seen as a logical outcome of trade liberalization (Banks, 2003). Increased competition from overseas helped ease old antagonisms between management and labor; and, by making it more difficult for businesses to pass higher costs on to consumers, generated powerful political pressures for domestic reforms. Public enterprise reforms took place through privatization, corporatization, and changes in the regulatory framework in late 1980s. Further reforms in the product market and the state-owned enterprises were implemented in 1995 within the framework of the National Competition Policy. In the labor market, a two-tier wage setting process was introduced in 1987 under which general wage increases were determined nationally, but productivity-based increases could be negotiated at the enterprise level. To advance the move toward enterprise bargaining, the Industrial Relations Act was enacted in 1993. Further progress on the labor market reform was achieved with the passage of the Work Relation Act in 1996 and its subsequent amendments.13
61. Australia also laid the foundation for sound macroeconomic policy in the early 1990s by adopting an inflation-targeting framework for monetary policy and beginning the process of fiscal consolidation. Significant progress in product and labor market reforms, along with a prudent fiscal policy, contributed to the successful implementation of the inflation-targeting framework, leading to a significant reduction in inflation. Inflation declined rapidly from double-digit levels in the 1970s and 1980s to below 3 percent by the early 1990s. The credibility of Australia’s fiscal policy has been enhanced through the 1998 Charter of Budget Honesty, which provided a framework for the conduct of government fiscal policy by requiring that it be based on principles of sound fiscal management and by facilitating public scrutiny of fiscal policy and performance.
62. Trade liberalization and other domestic reforms have significantly enhanced the competitiveness and the flexibility of the Australian economy. As a result, Australia’s trade increased markedly. After the import substitution strategies were abandoned, domestic manufacturing has become significantly more competitive, with exports of manufacturing increasing from 15 percent to 24 percent during the 1990s (Productivity Commission, 2003).14 Improved productivity and robust employment growth boosted average growth in GDP per capita to 2.7 percent during 1993-2002, compared with the OECD average of 1.7 percent during the same period. Moreover, the current expansion, now in its 13th year, has been remarkably resilient to severe negative shocks, including the Asian crisis and the global slowdown after the bursting of the tech bubble, supporting the view that the reforms have also enhanced the flexibility of the economy.
D. Trade Liberalization and Economic Growth: A Panel Data Analysis
63. This section tests econometrically whether trade liberalization has had an impact on economic growth in Australia and other industrial counties; and, if so, whether the benefits of trade liberalization are larger when implemented in conjunction with other structural reforms such as labor and product market reforms. To assess the significance of trade liberalization in the growth performance of industrial countries, a reduced form growth equation was estimated using panel data covering 15 OECD countries for the period 1975 to 1998.15
64. The reduced form equation, which can be derived from a neoclassical production function expressed in growth rates, relates output growth to a linear function of the growth of factor inputs—the stock of physical capital and total employment—and the growth of total factor productivity (TFP). The estimated growth equation is obtained by, in turn, expressing TFP growth as a function of a set of explanatory variables.16 Given the focus in this paper on the effect of trade liberalization on growth, total factor productivity growth is assumed to be determined by a trade liberalization index, its interaction with other structural reform indexes to test whether a comprehensive approach to structural reforms yields a larger growth dividend than a piecemeal approach. Another trade-related variable that has been found to be positively related to growth is the degree of openness of the economy generally measured by the ratio of exports of goods and services to GDP.17
65. In addition to a comprehensive structural reform program, Australia also implemented a wide-ranging reform of its macroeconomic policy framework centered on bringing inflation down from a two-digit to a low single-digit level, and consolidating its fiscal accounts. Consequently, inflation and the size of the government, which is measured by public consumption as a share of GDP, are additional explanatory variables in the estimated growth equation.
66. The dependent variable is the annual growth rate of GDP per capita in PPP terms (GDPPC, hereafter). Generalized Least Squares (GLS) was used to estimate the growth equation in order to take into account the significant heteroskedasticity of the error term. The results are given in Table 4. As discussed above, by including factor inputs in the growth equation, the coefficients on the other explanatory variables can be interpreted as their impact on growth through TFP. All variables have the expected sign and are statistically significant at least at the 5 percent level. In particular, the trade reform index is positively correlated with GDPPC growth, suggesting that freer trade tends to promote growth through higher TFP growth.18
Australia: Structural Reform Indicators for 15 OECD Countries
Australia: Structural Reform Indicators for 15 OECD Countries
| International Trade | Product Market | Labor Market | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1975 | 1985 | 2001 | 1975 | 1985 | 1998 | 1975 | 1985 | 1998 | ||||||||||
| Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | |
| Australia | 0.11 | 15 | 0.32 | 15 | 0.69 | 15 | 0.30 | 5 | 0.30 | 6 | 0.56 | 6 | 0.50 | 9 | 0.50 | 6 | 0.50 | 6 |
| Austria | 0.73 | 10 | 0.89 | 7 | 1.00 | 1 | 0.15 | 11 | 0.15 | 11 | 0.27 | 13 | 0.53 | 8 | 0.41 | 10 | 0.40 | 10 |
| Belgium | 0.90 | 2 | 0.92 | 3 | 1.00 | 2 | 0.21 | 8 | 0.21 | 10 | 0.49 | 8 | 0.28 | 15 | 0.30 | 14 | 0.38 | 12 |
| Canada | 0.62 | 14 | 0.74 | 12 | 0.94 | 11 | 0.61 | 2 | 0.61 | 2 | 0.73 | 4 | 0.71 | 2 | 0.70 | 2 | 0.73 | 2 |
| Finland | 0.79 | 8 | 0.93 | 2 | 1.00 | 2 | 0.08 | 12 | 0.14 | 12 | 0.19 | 15 | 0.45 | 10 | 0.45 | 7 | 0.45 | 7 |
| France | 0.89 | 3 | 0.93 | 1 | 0.99 | 8 | 0.00 | 15 | 0.00 | 15 | 0.19 | 14 | 0.53 | 7 | 0.42 | 8 | 0.37 | 13 |
| Germany | 0.83 | 5 | 0.91 | 6 | 0.99 | 7 | 0.24 | 6 | 0.24 | 7 | 0.52 | 7 | 0.38 | 13 | 0.39 | 12 | 0.42 | 8 |
| Italy | 0.97 | 1 | 0.92 | 4 | 1.00 | 5 | 0.03 | 14 | 0.03 | 14 | 0.36 | 11 | 0.65 | 3 | 0.66 | 3 | 0.54 | 5 |
| Japan | 0.76 | 9 | 0.81 | 10 | 0.83 | 14 | 0.44 | 3 | 0.44 | 3 | 0.71 | 5 | 0.64 | 5 | 0.65 | 4 | 0.64 | 3 |
| Netherlands | 0.87 | 4 | 0.92 | 5 | 1.00 | 2 | 0.07 | 13 | 0.07 | 13 | 0.40 | 10 | 0.37 | 14 | 0.29 | 15 | 0.34 | 14 |
| New Zealand | 0.65 | 12 | 0.67 | 14 | 0.84 | 13 | 0.41 | 4 | 0.42 | 4 | 0.76 | 3 | 0.44 | 11 | 0.42 | 9 | 0.42 | 9 |
| Spain | 0.63 | 13 | 0.70 | 13 | 0.96 | 10 | 0.19 | 9 | 0.23 | 8 | 0.48 | 9 | 0.42 | 12 | 0.33 | 13 | 0.40 | 11 |
| Sweden | 0.81 | 7 | 0.82 | 9 | 0.99 | 6 | 0.23 | 7 | 0.23 | 9 | 0.31 | 12 | 0.55 | 6 | 0.41 | 11 | 0.00 | 15 |
| United Kingdom | 0.83 | 6 | 0.88 | 8 | 0.99 | 9 | 0.18 | 10 | 0.30 | 5 | 1.00 | 1 | 0.65 | 4 | 0.62 | 5 | 0.63 | 4 |
| United States | 0.69 | 11 | 0.75 | 11 | 0.88 | 12 | 0.73 | 1 | 0.73 | 1 | 0.77 | 2 | 0.81 | 1 | 0.82 | 1 | 0.81 | 1 |
| Mean | 0.74 | 0.81 | 0.94 | 0.26 | 0.27 | 0.52 | 0.53 | 0.49 | 0.47 | |||||||||
| Top 5 | 0.85 | 0.92 | 1.00 | 0.47 | 0.50 | 0.79 | 0.69 | 0.69 | 0.67 | |||||||||
| S.D. (whole sample) | 0.20 | 0.19 | 0.14 | |||||||||||||||
Australia: Structural Reform Indicators for 15 OECD Countries
| International Trade | Product Market | Labor Market | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1975 | 1985 | 2001 | 1975 | 1985 | 1998 | 1975 | 1985 | 1998 | ||||||||||
| Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | Value | Rank | |
| Australia | 0.11 | 15 | 0.32 | 15 | 0.69 | 15 | 0.30 | 5 | 0.30 | 6 | 0.56 | 6 | 0.50 | 9 | 0.50 | 6 | 0.50 | 6 |
| Austria | 0.73 | 10 | 0.89 | 7 | 1.00 | 1 | 0.15 | 11 | 0.15 | 11 | 0.27 | 13 | 0.53 | 8 | 0.41 | 10 | 0.40 | 10 |
| Belgium | 0.90 | 2 | 0.92 | 3 | 1.00 | 2 | 0.21 | 8 | 0.21 | 10 | 0.49 | 8 | 0.28 | 15 | 0.30 | 14 | 0.38 | 12 |
| Canada | 0.62 | 14 | 0.74 | 12 | 0.94 | 11 | 0.61 | 2 | 0.61 | 2 | 0.73 | 4 | 0.71 | 2 | 0.70 | 2 | 0.73 | 2 |
| Finland | 0.79 | 8 | 0.93 | 2 | 1.00 | 2 | 0.08 | 12 | 0.14 | 12 | 0.19 | 15 | 0.45 | 10 | 0.45 | 7 | 0.45 | 7 |
| France | 0.89 | 3 | 0.93 | 1 | 0.99 | 8 | 0.00 | 15 | 0.00 | 15 | 0.19 | 14 | 0.53 | 7 | 0.42 | 8 | 0.37 | 13 |
| Germany | 0.83 | 5 | 0.91 | 6 | 0.99 | 7 | 0.24 | 6 | 0.24 | 7 | 0.52 | 7 | 0.38 | 13 | 0.39 | 12 | 0.42 | 8 |
| Italy | 0.97 | 1 | 0.92 | 4 | 1.00 | 5 | 0.03 | 14 | 0.03 | 14 | 0.36 | 11 | 0.65 | 3 | 0.66 | 3 | 0.54 | 5 |
| Japan | 0.76 | 9 | 0.81 | 10 | 0.83 | 14 | 0.44 | 3 | 0.44 | 3 | 0.71 | 5 | 0.64 | 5 | 0.65 | 4 | 0.64 | 3 |
| Netherlands | 0.87 | 4 | 0.92 | 5 | 1.00 | 2 | 0.07 | 13 | 0.07 | 13 | 0.40 | 10 | 0.37 | 14 | 0.29 | 15 | 0.34 | 14 |
| New Zealand | 0.65 | 12 | 0.67 | 14 | 0.84 | 13 | 0.41 | 4 | 0.42 | 4 | 0.76 | 3 | 0.44 | 11 | 0.42 | 9 | 0.42 | 9 |
| Spain | 0.63 | 13 | 0.70 | 13 | 0.96 | 10 | 0.19 | 9 | 0.23 | 8 | 0.48 | 9 | 0.42 | 12 | 0.33 | 13 | 0.40 | 11 |
| Sweden | 0.81 | 7 | 0.82 | 9 | 0.99 | 6 | 0.23 | 7 | 0.23 | 9 | 0.31 | 12 | 0.55 | 6 | 0.41 | 11 | 0.00 | 15 |
| United Kingdom | 0.83 | 6 | 0.88 | 8 | 0.99 | 9 | 0.18 | 10 | 0.30 | 5 | 1.00 | 1 | 0.65 | 4 | 0.62 | 5 | 0.63 | 4 |
| United States | 0.69 | 11 | 0.75 | 11 | 0.88 | 12 | 0.73 | 1 | 0.73 | 1 | 0.77 | 2 | 0.81 | 1 | 0.82 | 1 | 0.81 | 1 |
| Mean | 0.74 | 0.81 | 0.94 | 0.26 | 0.27 | 0.52 | 0.53 | 0.49 | 0.47 | |||||||||
| Top 5 | 0.85 | 0.92 | 1.00 | 0.47 | 0.50 | 0.79 | 0.69 | 0.69 | 0.67 | |||||||||
| S.D. (whole sample) | 0.20 | 0.19 | 0.14 | |||||||||||||||
Trade Liberalization and GDP Per Capita Growth: Panel Regressions
Trade Liberalization and GDP Per Capita Growth: Panel Regressions
| Estimation Method | ||||
|---|---|---|---|---|
| Independent Variables | GLS | GLS | GLS-IV | GLS-IV |
| Log(GDPPC_0) | -0.0160 | -0.0156 | -0.0179 | -0.0162 |
| (-12.55)* | (-12.81)* | (-11.40)* | (-13.44)* | |
| Δlog(EMPL) | 0.4205 | 0.4179 | 0.4868 | 0.4380 |
| (10.41)* | (10.67)* | (11.06)* | (11.48)* | |
| Δlog(K) | 0.0710 | 0.0641 | 0.0471 | 0.0359 |
| (4.99)* | (4.55)* | (3.16)* | (2.64)* | |
| TRADE_lib | 0.0151 | 0.0087 | 0.0181 | 0.0091 |
| (12.34)* | (3.74)* | (20.28)* | (3.82)* | |
| LABOR_ref | 0.0059 | 0.0059 | ||
| (2.50)* | (4.36)* | |||
| PRODUCT_ref | 0.0069 | 0.0054 | ||
| (9.13)* | (5.32)* | |||
| TRADE_lib*LABOR_ref | 0.0090 | 0.0101 | ||
| (3.13)* | (3.73)* | |||
| TRADE_lib*PRODUCT_ref | 0.0066 | 0.0067 | ||
| (7.39)* | (8.29)* | |||
| OPEN | .0088 | .0094 | .0031 | .0074 |
| (3.44)* | (3.62)* | (2.24)** | (3.17)* | |
| INFL | 0.2412 | 0.2287 | 0.1464 | 0.2021 |
| (5.58}* | (5.40}* | (3.09}* | (4.62}* | |
| D1*(INFL-0.03) | -0.2677 | -0.2598 | -0.1609 | -0.2313 |
| (-6.36)* | (-6.30)* | (-3.49)* | (-5.50)* | |
| Cg | -0.0111 | -0.0113 | -0.0049 | -0.0050 |
| (-2.13)** | (-2.30)** | (-0.77) | (-1.11) | |
| TREND | -0.0003 | -0.0003 | -0.0003 | -0.0003 |
| (-7.30)* | (-7.30)* | (-6.83)* | (-8.05)* | |
| Adjusted R2 | 0.60 | 0.61 | 0.60 | 0.60 |
| NT | 428 | 428 | 393 | 393 |
| Wooldridge Autocorrelation Test | 1.04 | 0.97 | 0.64 | 1.21 |
| H0: no autocorrelation (t-distribution) | ||||
Trade Liberalization and GDP Per Capita Growth: Panel Regressions
| Estimation Method | ||||
|---|---|---|---|---|
| Independent Variables | GLS | GLS | GLS-IV | GLS-IV |
| Log(GDPPC_0) | -0.0160 | -0.0156 | -0.0179 | -0.0162 |
| (-12.55)* | (-12.81)* | (-11.40)* | (-13.44)* | |
| Δlog(EMPL) | 0.4205 | 0.4179 | 0.4868 | 0.4380 |
| (10.41)* | (10.67)* | (11.06)* | (11.48)* | |
| Δlog(K) | 0.0710 | 0.0641 | 0.0471 | 0.0359 |
| (4.99)* | (4.55)* | (3.16)* | (2.64)* | |
| TRADE_lib | 0.0151 | 0.0087 | 0.0181 | 0.0091 |
| (12.34)* | (3.74)* | (20.28)* | (3.82)* | |
| LABOR_ref | 0.0059 | 0.0059 | ||
| (2.50)* | (4.36)* | |||
| PRODUCT_ref | 0.0069 | 0.0054 | ||
| (9.13)* | (5.32)* | |||
| TRADE_lib*LABOR_ref | 0.0090 | 0.0101 | ||
| (3.13)* | (3.73)* | |||
| TRADE_lib*PRODUCT_ref | 0.0066 | 0.0067 | ||
| (7.39)* | (8.29)* | |||
| OPEN | .0088 | .0094 | .0031 | .0074 |
| (3.44)* | (3.62)* | (2.24)** | (3.17)* | |
| INFL | 0.2412 | 0.2287 | 0.1464 | 0.2021 |
| (5.58}* | (5.40}* | (3.09}* | (4.62}* | |
| D1*(INFL-0.03) | -0.2677 | -0.2598 | -0.1609 | -0.2313 |
| (-6.36)* | (-6.30)* | (-3.49)* | (-5.50)* | |
| Cg | -0.0111 | -0.0113 | -0.0049 | -0.0050 |
| (-2.13)** | (-2.30)** | (-0.77) | (-1.11) | |
| TREND | -0.0003 | -0.0003 | -0.0003 | -0.0003 |
| (-7.30)* | (-7.30)* | (-6.83)* | (-8.05)* | |
| Adjusted R2 | 0.60 | 0.61 | 0.60 | 0.60 |
| NT | 428 | 428 | 393 | 393 |
| Wooldridge Autocorrelation Test | 1.04 | 0.97 | 0.64 | 1.21 |
| H0: no autocorrelation (t-distribution) | ||||
67. Table 5 illustrates the results in Table 4 with two scenarios. The first scenario assesses the impact on GDPPC growth from trade, product, and labor market reforms in Australia since 1985. The second scenario measures the impact on growth from hypothetical further reforms that would bring Australia on par with the top 5 reformers—that is, by raising end-of-period Australian levels of TRADE_ref, LABOR_ref, and PRODUCT_ref indexes to the average of the 5 highest levels in OECD countries. According to the first equation in Table 4, trade liberalization in Australia since 1985, as measured by the change in TRADE_ref from 1985 to 2001, has raised annual GDPPC growth by 0.56 percentage points. Including also product and labor market reforms, raises the contribution of reforms to annual growth to 0.74 percentage points. To test the existence of complementarities between trade reform and labor and product market reform, interactive terms were added in the second equation. Because of multicolinearity problems, the labor and product reform indexes only enter the equation as interactive terms with the trade liberalization index. The interactive terms are highly significant, suggesting that complementarities between reforms do indeed exist.19 According to equation (2), trade reform accounted for 0.56 percentage points in Australia’s growth in GDPPC, with complementarity effects accounting for 43 percent of the increase in GDPPC growth. Further reforms could bring additional, significant gains in GDPPC growth. Further trade liberalization that would raise TRADE_ref index to the average of the top 5 levels would raise GDPPC growth by more than ½ percentage point. Growth could be raised by as much as 0.9 percent if all 3 indexes were raised to average of 5 highest levels.
Impact on Australia’s Growth in GDPPC from Increase in Reform Indexes (in percent)1
Impact on Australia’s Growth in GDPPC from Increase in Reform Indexes (in percent)1
| Equation | |||||
|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | ||
| Reform since 1985 | |||||
| TRADE_ref | |||||
| Direct effect | 0.56 | 0.32 | 0.67 | 0.34 | |
| Total effect | 0.56 | 0.56 | 0.67 | 0.60 | |
| all 3 reform indexes | |||||
| Direct effect | 0.74 | 0.32 | 0.81 | 0.34 | |
| Total effect | 0.74 | 0.66 | 0.81 | 0.70 | |
| Further Reform to | |||||
| match top 5 countries in: | |||||
| TRADE_ref | |||||
| Direct effect | 0.47 | 0.27 | 0.56 | 0.28 | |
| Total effect | 0.47 | 0.52 | 0.56 | 0.56 | |
| All 3 reform indexes | |||||
| Direct effect | 0.73 | 0.27 | 0.79 | 0.28 | |
| Total effect | 0.73 | 0.83 | 0.79 | 0.88 | |
Impact on Australia’s Growth in GDPPC from Increase in Reform Indexes (in percent)1
| Equation | |||||
|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | ||
| Reform since 1985 | |||||
| TRADE_ref | |||||
| Direct effect | 0.56 | 0.32 | 0.67 | 0.34 | |
| Total effect | 0.56 | 0.56 | 0.67 | 0.60 | |
| all 3 reform indexes | |||||
| Direct effect | 0.74 | 0.32 | 0.81 | 0.34 | |
| Total effect | 0.74 | 0.66 | 0.81 | 0.70 | |
| Further Reform to | |||||
| match top 5 countries in: | |||||
| TRADE_ref | |||||
| Direct effect | 0.47 | 0.27 | 0.56 | 0.28 | |
| Total effect | 0.47 | 0.52 | 0.56 | 0.56 | |
| All 3 reform indexes | |||||
| Direct effect | 0.73 | 0.27 | 0.79 | 0.28 | |
| Total effect | 0.73 | 0.83 | 0.79 | 0.88 | |
68. The coefficient estimate for the degree of openness is also positively related to growth and is highly significant. The degree of openness of the Australian economy increased from 14 percent in the early 1980s to 22 percent in 2002. This could have added almost 0.1 percentage point to annual GDP per capita growth. This increase in the degree of openness may have been largely driven by the trade liberalization policies that have been implemented since the 1980s. Indeed, regressing the degree of openness on the trade liberalization index using the whole panel data yields a positive and statistically significant coefficient with a t-statistic of 15 and an adjusted R2 of 0.26. The total trade effect, obtained by combining the effects of the trade liberalization index and trade openness, is approximately a ½ of a percentage point.
69. Macroeconomic reform may have also had a substantial positive effect on growth. Australia adopted an inflation targeting framework in 1993 to reduce the persistently high inflation rates of the 1970s and 1980s.20 As noted, trade and other domestic reforms have contributed significantly to the successful implementation of the inflation targeting framework. Consequently, inflation was brought down from 11½ percent in the decade prior to reforms (1974-83) to less than 3 percent in the past decade. This reduction in inflation alone might have lifted Australian growth by about a ¼ of a percentage point. Following recent empirical research, inflation enters the growth equation in a nonlinear fashion to test if the effect of inflation on growth depends on whether the initial level of inflation is below or above a threshold level.21 The estimation results corroborate previous studies where a modest increase in inflation could stimulate growth if inflation remains below 3 percent but starts to hurt growth if inflation increases above 3 percent. The estimated coefficient implies that a one percentage point increase in inflation, when initial inflation is below 3 percent, will raise growth by almost a ¼ of a percentage point. However, an increase of the same magnitude, when inflation is already above 3 percent will reduce growth by 0.03 percentage points. Thus, a reduction of inflation from 11½ percent to 3 percent would boost growth by a ¼ of a percentage point.
70. Finally, an increase in factor inputs growth and a reduction in the size of the government promote growth in GDPPC. The negative and statistically significant coefficient on initial GDP per capita (GDPPC_0) indicates that GDP per capita across OECD ries exhibit conditional convergence.
71. The coefficient estimates obtained by GLS in Table 4 may be biased due to the potential endogeneity of the explanatory variables. Therefore, the growth equation was also estimated using GLS with instrumental variables (GLS-IV). All variables but the time trend were instrumented for. The list of instruments for each variable includes its own lag values and a time trend. This set of instruments is only valid under the hypothesis of non-autocorrelation.22 The results are given in the third and fourth columns of Table 4. Generally, the GLS-IV coefficient estimates are relatively close to ordinary GLS estimates. The Wooldridge autocorrelation test fails to reject the null of no autocorrelation at conventional significance levels, which validates the use of lagged variables as instruments.
72. While the results need to be interpreted with caution given that the relationship between the explanatory variables and GDP per capita growth in the estimated equations is not necessarily causal, the results support the view that structural and macroeconomic reforms have had a strong positive impact on Australia’s growth performance.
References
Arellano, Manuel, and Stephen Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equation,” Review of Economic Studies, Vol. 58, pp. 277– 97.
Barro, Robert Xavier Sala-i-Martin, Xavier, 1995, “Economic growth,” Advanced Series in Economics. New York; London and Montreal: McGraw-Hill
Bank, Gary, 2003, “Gaining from Trade Liberalization: Some Reflection on Australia’s Experience,” Paper presented to the IIB&L/CEDA Conference, ‘New Horizon in Trade: The WTO Round and Australia’s Free Trade Negotiations,” Adelaide, June 5.
Berg, Andrew and Anne Krueger, 2003, “Trade, Growth, and Poverty: A Selective Survey,” IMF Working Paper No. WP/03/30, (Washington: International Monetary Fund).
Caselli, F., Esquivel, G. and Lefort F. 2003, “Reopening the Convergence Debate: A New Look at Cross-Country Empirics,” Journal of Economic Growth, Vol. 1 pp. 363– 89.
Center for International Economics (2004), “Economic Impacts of an Australia-United States Free Trade Area,” Center for International Economics, Canberra.
Coe, David, and Elhanan Helpman, 1995, “International R&D Spillovers,” European Economic Review, Vol. 39, pp. 859– 87.
Elhanan Helpman, and Alexander Hoffmaister, 1997, “North-South R&D Spillovers,” Economic Journal, Vol. 107, pp. 134– 49.
Dennis Snower, 1997, “Policy Complementarities: The Case for Fundamental Labor Market Reform,” IMF Staff papers, International Monetary Fund, Vol. 44, No. 1, pp. 1– 35.
Chand, Satish, 1999, “Trade Liberalization and Productivity Growth: Time-Series Evidence from Australian Manufacturing,” The Economic Record, Vol. 75, pp. 28– 36 .
Dollar, David, and Aart Kraay 2004, “Trade, Growth and Poverty,” Economic Journal Vol. 114, pp. 22– 48.
Dixon, Peter, Brian Parmenter, and Russell Rimmer, 1986, “ORANI Projections of the Short-Run Effects of a 50% Across-the-Board Cut in Protection Using Alternative Data Bases,” in T. N. Srinivasan and John Whalley (Eds), General Equilibrium Trade Policy Modeling, MIT Press, Cambridge.
Economic Planning Advisory Council (EPAC), 1996, “Tariff Reform and Economic Growth,” EPAC Commission PaperNo. 10, Economic Planning Advisory Council, Canberra.
Edwards, Sebastian, 1998, “Openness, Productivity and Growth: What Do We Really Know?” Economic Journal, Vol. 108, pp. 383– 98.
Frankel, Jeffry, and David Romer, 1999, “Does Trade Cause Growth?” American Economic Review, Vol. 89, pp. 379– 99.
Hakura, Dalia, and Florence Jaumotte, 1999, ”The Role of Inter- and Intra-Industry Trade in Technology Diffusion, IMF Working Paper No. WP/99/58, (Washington: International Monetary Fund).
Hartel, Thomas and David Hummels, Maros Ivanic, and Roman Keeney, 2004, “How Confident can We be in CGE-Based Assessments of Free Trade Agreements?” NBER Working Paper 10477 (Cambridge: National Bureau of Economic Research).
Hartel, Thomas and Will Martin, 1999, “Would Developing Countries Gain from Inclusion of Manufacturing in the WTO Negotiation?” Paper resented at the Conference on WTO and the Millennium Round, Geneva, September 21.
International Monetary Fund, 2004, “Advancing Structural Reform” World Economic Outlook, (Washington: International Monetary Fund).
Khan, Mohsin, and Abdelhak Senhadji, 2001, “Threshold Effects in the Relationship Between Inflation and growth,” IMF Staff papers, International Monetary Fund, Vol. 48, No. 1, pp. 1– 21.
Mahadevan, Renuka, 2002, “Trade Liberalization and Productivity Growth in Australian Manufacturing Industries,” Atlantic Economic Journal, Vol. 30, pp. 170– 83.
McKibbin, Warwick, Lee Davis, and Andrew Stoeckel (2000), “Economic Benefits from AFTA-CER Free Trade Area,” Center for International Economics, Canberra.
Nickel, Stephen, and Luca Nunziata, 2001, “Labor Market Institutions Database,” London: Centre for Economic Performance, London School of Economics. Available via Internet: http://cep.lse.ac.uk/pubs/download/data0502.zip
Nicoletti, Giuseppe, and Stefano Scarpetta, 2003, “Regulation Productivity and Growth: OECD Evidence,” Economic Policy, Vol. 18, No. 36, pp. 19– 72.
Parham, Dean, 2002, “Microeconomic Reforms and the Revival in Australia’s Growth in Productivity and Living Standards, Paper presented to the Conference of Economists, Adelaide, October 2.
Productivity Commission, 2000, “Modeling the Effect of Removing General Tariffs,” Supplement to Inquiry Report: Review of Australia’s General Tariffs Arrangements, Productivity Commission, Canberra.
Productivity Commission, 2003, Trend in Australian Manufacturing, Commission Research, Productivity Commission, Canberra.
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi, 2002, “Institution Rule: The Premacy of Institutions over Geography and Integration in Economic Development,” NBER Working Paper 9305 (Cambridge: National Bureau of Economic Research).
Rodriguez, Fransisco, and Dani Rodrick, 1999, “Trade Policy and Economic Growth: A Skeptic’s Guide to the Cross-National Evidence,” NBER Working Paper 7081 (Cambridge: National Bureau of Economic Research).
Sachs, Jeffry, and Andrew Warmer, 1995, “Economic Reform and Process of Global Integration,” Brookings Papers on Economic Activity: 1, Brookings Institution.
Salgado, Ranil, 2002, “Impact of Structural Reforms on Productivity Growth in Industrial Countries,” IMF Working Paper 02/10 (Washington: International Monetary Fund).
Scollay, Robert, and John Gilbert, 2000, “Measuring the Gains from APEC Trade Liberalization: An Overview of CGE Assessments,” World Economy, Vol. 23, pp 175– 97.
Wacziarg, Romain, and Karen Welch, 2003, “Trade Liberalization and Growth: New Evidence,” NBER Working Paper 10152 (Cambridge: National Bureau of Economic Research).
ANNEX I Derivation of the Empirical Growth Equation
Following the empirical growth literature, a reduced form growth equation is derived from an aggregate production function as follows:23
where Yit is GDP per capita on a purchasing power parity (PPP) basis, Ait is total factor productivity, Kit is the stock of capital, and Lit is total employment, all for country i in year t. Differentiating equation (1) with respect to time yields GDP per capita growth as a function of growth in total factor productivity, growth in the capital stock, and growth in total employment:
where lower case variables represent the growth rate of the corresponding uppercase variable in equation (1). It is assumed that TFP growth, ait, is a function g of a set of factors
Substituting equation (3) into (2) yields the final equation:
The set of explanatory variables
ANNEX II Data Sources and Definitions
The data are annual for the period 1975 to 1998. The countries included in the sample are: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, United States. The definitions and sources for each variable are:
GDP per capita in Purchasing Power Parity (PPP) terms (GDPPC). Source: OECD.
Total employment (EMPL). Source: OECD.
Capital stock of the business sector (K). Source: OECD.
Inflation (INFL), based on CPIs. Source: OECD.
Public Consumption as a share of GDP (Cg) is used as a measure of the size of the public sector in the economy. Source: OECD.
Product market reform indicator (PRODUCT_ref). The indicator is constructed by Nicoletti and Scarpetta (2003) covering product market reforms over the 1975-98. It is based on indices for barriers to entry, public ownership, market structure, vertical integration, and price control in nonmanufacturing sector, which includes the following industries: passenger air transport, telecommunications, electricity, railways, post, road freight, and gas. The indicator is rescaled to range between 0 to 1, with increasing values indicating a less restrictive product market.
Trade reform indicator (TRADE_lib). The indicator is constructed by using average effective tariffs, calculated as the ratio of customs and import duties (from OECD, Revenue Statistics; and IMF, Government Financial Statistics) to the value of imports (from IMF, International Financial Statistics). The index is rescaled to range between 0 to 1, with increasing value indicating less restrictive trade regime.
Labor market reform indicator (LABOR_ref). It is based on simple average of indices of employment protection, benefits replacement rates, and benefit duration. Employment protection measures the restrictiveness of employment protection. Benefits replacement rates is the average of first-year unemployment benefits as a percentage of earnings before tax. Benefit duration is the ratio of the average benefit replacement rates in the second to the fifth year of an unemployment spell to the average benefit replacement rate in the first year of an unemployment spell. The indicator is rescaled to range between 0 to 1, with increasing value indicating more flexible labor market regime. Source: Labor Market Institutions Database developed by Nickel and Nunziata (2001). WEO extended the indicator using OECD data provided by Nicolleti.
Degree of openness of the economy (OPEN) is defined as exports of goods and services divided by GDP. Source: OECD.
Prepared by Abdelhak Senhadji (Ext. 38380) and Edimon Ginting (Ext. 38733).
Caselli et al (1996) propose a specification in which a growth equation is estimated in decadal differences with the lag of the right-hand side variable as an instrument. This approach has several desirable features that address the issue of colinearity and omitted variables, and presents a natural set of instruments to control for possible reverse causation between trade and growth.
Australia’s average import tariff was more than 12 percent in the mid-1970s and declined only slightly to about 10 percent by 1985. This was in stark contrast to OECD countries where the average tariff in 1985 was only 1.7 percent.
While tariffs were cut by 25 percent in 1973, a comprehensive reform program was only launched in mid-1980s.
Corresponding to the adoption of the Competition Policy Reform Act 1995.
The case of the Australian automobile industry clearly illustrates the benefits of trade liberalization. Notwithstanding significant reductions in tariff protection and a substantial increase in imports, local production is running close to record levels, with more than 30 percent of sales going overseas compared with less than 10 percent a decade ago (Banks, 2003).
The countries in the panel include Australia, Austria, Belgium, Canada, Finland, France, Germany, Italy, Japan, Netherlands, New Zealand, Spain, Sweden, United Kingdom, and United States. The time period is constrained by data availability on structural reform variables discussed below.
While trade liberalization and the degree of openness are correlated, the inclusion of both variables in the growth equation is likely to better capture the benefits from trade.
The reform indicators are constructed such that an increase in the index reflects a deepening of reform. For example, an increase in the trade reform index implies a reduction in trade barriers. See Annex 2 for the definition of the reform indexes.
Although the inflation targeting framework was adopted in 1993, the framework was not fully formalized until the 1996 Statement on the Conduct of Monetary Policy.
Khan and Senhadji (2001) estimate the threshold level of inflation above which inflation starts to reduce growth at 3 percent for industrial countries.
Note that estimators based on first-differencing of the growth equation such as the Arellano-Bond method are not valid here, given that the equation is not specified in its dynamic form and does not include fixed effects. Therefore, the equation should be interpreted as the long-un relationship between GDPPC growth and its fundamental determinants. Short-term dynamics were removed from the data by applying the HP filter.
While kit and lit represent growth rates of the capital stock and total employment, the variables
