Effects of the 1994 Uruguay Round on
the trade-to-GDP ratios of developing countries
Bachelor's thesis June 2018
B.Sc. Economics and Business
Specialization
Economics and finance
Name
Aleena Rahman
Student number
11106964
Supervisor
Dr. Peter Foldvari
Abstract
Ever since Andrew Rose published the paper in 2003, Do we really know that the WTO increases trade?, there has been a spike of research into the effects of the WTO which were previously largely taken for granted. This paper is an addition to the empirical research done to answer the question, have traditional trade policies helped increase trade and ultimately welfare. In this paper I focus on the Uruguay Round of 1994 which was the most dramatic change of trade laws in modern history. This study uses panel data over the years 1975-2015 followed by a fixed-effects regression. I quantify the constituents of the Uruguay Round which are broadly the protection of the liberalization of agriculture, FDI, service sectors, and intellectual property to see if, depending on how important these sectors are to the economy of a country, they have had any significant effects on trade. The results indicate that agriculture positively affects trade in developing countries but these effects are mitigated by WTO membership. They also show that FDI also has a positive impact on trade in developed and developing countries alike.
Statement of originality
This document is written by Aleena Rahman who declares to take full responsibility for
the contents of this document. I declare that the text and the work presented in this
document is original and that no sources other than those mentioned in the text and its
references have been used in creating it. The Faculty of Economics and Business is
responsible solely for the supervision of completion of the work, not for the contents.
Table of contents
1. Introduction
4
2. Literature review
5
2.1. History of International trade and the formation of the WTO
5
2.2. The bigger picture: GDP
6
2.3. Developed vs. developing countries
7
2.4. Services vs. manufacturing sectors
8
2.5. Volume vs. variety of goods
8
3. Data collection and analysis
10
3.1. Methodology
10
3.1.1. Components of the Uruguay Round
10
3.1.2. The regression model
12
3.2. Results
13
3.2.1. Regression results
14
3.2.2. Time dummies
17
4. Conclusions
18
5. Limitations
19
6. References
20
7. Appendix
22
1. Introduction
When WTO was officially established in 1995, it symbolized a huge step towards the spread of globalization to all corners of the globe as well as a huge step towards better choice and consequently better standard of living for consumers by having access to all the variety offered by traders of the international market. Anderson et al. (2000) estimated the potential increase in world welfare upon joining the WTO to be well over $250 billion. A customary part of the journey up the development ladder is globalization, comprising of free trade, privatization, FDI deregulation, and protection of intellectual property rights (H.J. Chang, 2008). Basic trade theory says, remove the tariffs and trade will rise, but the empirics point in a different direction (Eicher and Henn, 2010). Rose published his famous report in 2003, indicating that WTO membership has not had a significant impact on international trade (Rose, 2003); naturally these findings sparked a massive discussion over a previously undebated topic (Dutt et al., 2013).
Plenty of detailed research has since been conducted, aimed at various aspects of the WTO and its effects on developed and developing countries. The aim of this paper is to add to that research using a different combination of variables and testing their effects on the common variable, trade, over the years. This research is centered around the Uruguay Round of 1994 which also marked the birth of the WTO, considered to be the largest drop in tariffs in modern history (Buono and Lelane, 2011). I quantify the different components of the Uruguay Round into measurable variables and test whether a significant relationship exists between each of them and the trade-to-GDP ratios of 81 countries. I go into greater detail of what these variables are and the reasons each of them has been chosen to collectively represent the Uruguay Round.
In the following sections, I start off by giving a brief overview of the history of international trade and the WTO and its purpose, followed by arguments and empirical research collected from previous papers in favor of and against international trade which will paint an up-to-date picture of the debates, the facts, and the controversies. I go into the details of the Uruguay Round, going over each component’s implications on the trade-to-GDP ratio one-by-one. Following that will be the quantitative approach of tackling the research question where the model and variables used are explained and how they relate to the theoretical concepts that are the backbone of this research.
2. Literature review
2.1. History of International Trade and the formation of the WTO
The end of World War 2 and the start of the Bretton Woods system marked a significant turning point in the international trade environment. Laws were established for the long run to limit trade barriers and quotas and shortly, 23 countries agreed to the terms of the GATT to govern trade among nations (WTO Annual Report, 2014). GATT was in place to reduce or completely eradicate barriers to trade to push international trade (Dutt et al., 2013). Talks of the Uruguay Round had started by the 1980s and by 1994 it had become five times as popular so on April of 1994, 123 countries signed the Uruguay Round. The WTO became official in 1995, marking another step in the pursuit of the free market economy. The Uruguay Round is preceded by seven rounds and followed by the Doha development agenda which started in 2001 but has since been at a standstill due mainly to two of the leaders of the developing world, Brazil and India, who argue that free trade is disadvantageous to developing countries (Anderson et al., 2000). Author Andrew Rose can possibly be credited with stalling this agenda longer than the usual.
In 2003, Andrew Rose published a paper that shook the universal faith put into free trade policies (Dutt et al., 2013). His paper estimated the effects of the WTO and GATT on international trade. Rose did an extensive research using a gravity model. His model estimated the level of trade between two countries at a certain point in time using a wide variety of variables that included generic measures such as GDP, population, and area, and also other more tailored ones like dummy variables for whether the two countries had a common language, were ever colonies under the same colonizer, and if they were WTO/GATT members at certain points in time. The results of his regressions showed that neither WTO nor GATT membership have had a significant impact on the trade patterns of a country, however, his conclusions were open in that he said that perhaps WTO membership still has influenced global welfare through channels other than trade.
Although there is debate over the effectiveness of the WTO, one thing there is agreement over is that it has indirect but strong links to some aspects of economic development. Acemoglu and Zilibotti found that development goes hand in hand with diversification opportunities, Hausmann et al. have shown that the types of goods that a country exports matters i.e. goods or services, Broda et al. found that a growth in the number of countries that are traded with can contribute to the country’s own productivity growth (Dutt et al., 2013), and Siddiqui and Iqbal show that trade growth has a positive relationship with productivity growth (Siddiqui and Iqbal, 2005) while Chang (2008) explains its adverse effects. An abundance of research has been conducted on various aspects of international trade and the different ways it is conventionally theorized to effect global
welfare. Below I present some prior research grouped by aspects of the economy that are affected by the WTO and how. Some striking findings are as follows.
2.2. The bigger picture: GDP
GDP growth is the reason economic policy makers get out of bed each day. It is a reflection of a country's economic, political, and social performance, so let us agree that it is a pretty important number. Siddiqui and Iqbal (2005) analyze the impact of trade liberalization on GDP growth for Pakistan over the years 1972 to 2002. They make use of two models where GDP growth is the dependent variable and explanatory variables include trade growth, investment growth, and population growth; in the second variation of their model they split the trade variable into exports and imports and introduce a dummy variable for when Pakistan joined the WTO. They found from their first model a significant negative relationship between trade growth and GDP growth but found an insignificant positive relationship between both import and export growths and GDP growth.
It is a curious result that trade growth has had a negative effect on GDP growth; however, Chang (2008) attributes the success of advanced nations to the very use of protectionist policies such as tariffs and quotas, FDI regulation, and subsidizing. Much of the success of developed countries is associated with their liberal trade policies; however the reality holds that throughout the course of history, these same countries have broken the rules of free trade during their fastest development stages that they sell to developing countries today (Chang, 2008). In fact, according to Anderson et al. (2000) many farmers in developed countries received considerable subsidies years after joining the WTO, despite cutting down on agricultural subsidies being on top of the Uruguay Round agenda. They explain that the reason why trade growth seems to have had ambiguous effects on GDP growth is that new trade policies have not had the chance to have an impact. Anderson et al. (2000) explain that traditional market policies of the WTO still have much to contribute to welfare in developed and developing countries alike due to the fact that trade barriers are still in full swing; many countries today have not fully implemented the components of the Uruguay Round enough for them to have material effect.
An IFPRI study says that these continued subsidies of developed countries cost developing countries around $24 billion in annual agricultural income as developed countries make their exports more competitive by driving prices down; furthermore, in 2001 the United States subsidized one commodity, cotton, with $3.4 billion (Diao et al., 2003), apart from subsidizing other goods. During the same year, subsidies in sugar for each OECD country were on average $6.35 billion, while the total collective amount of sugar subsidies for developing countries was $6.5 billion; it comes as no surprise that the share of developing countries sugar exports decreased from 71% to 54% from the late 1980s to the early 2000s while that for developed countries increased (Faeth et al., 2007). The following chart shows how little has changed in terms of agricultural subsidies for many countries from 1995 to 2015.
Changes in agricultural subsidies from 1995 to 2015. Source: OECD
Chang questions why free trade is promoted as a tool for development, using the examples of China and India having had rapid growth over recent years regardless of their use of protectionist policies over that time. He also asserts that there is still much to learn for developing countries from history, particularly the time periods over which European countries such as the Netherlands, Denmark, Belgium, Switzerland, among others, were growing fast. Over those time periods they were aided by government investments in farming education and equipment that placed these countries on the top of the global agriculture sector regardless of their small size.
2.3. Developed vs developing
Chang (2008) makes it evident that developed and developing countries react differently to trade policies. Eicher and Henn (2011) build on research done by Rose, Tomz et al., and Subramanian and Wei, who showed that the WTO has been more effective for advanced nations than for developing ones. Although they concluded that WTO effects on trade are not significant among the latter group of countries, their findings span more than that; they also found that WTO membership promotes trade among adjacent developing countries at the expense of distant trade. Additionally, they concluded that the WTO had significant positive effects on larger countries with high market power before joining the WTO. One explanation for this follows from the terms-of-trade theory which implies that countries with high market power can influence international terms-of-trade prices and do this by implementing trade barriers that mold international trade laws to their advantage.
It is not all up to the developed countries though. Konan and Maskus (2005) find that services liberalization has a much greater impact on welfare than goods liberalization (more on this in the next sub-section) and that developing countries could be reaping the benefits of international trade far more than they currently are if they did not place heavy overprotection policies on their own domestic service industries; according to Anderson et al., these restrictions on trade have had negative effects on many other industries in developing countries.
Anderson et al. (2000) study how many import restrictions still remain today after implementation of the Uruguay Round and how they affect the economic welfare of developing and developed countries alike. They argue that trade liberalization will result in increasing income, and ultimately better standard of living, for unskilled workers in developing countries when they sell to consumers in developed countries and this may offset the effects of the rise in prices.
2.4. Services vs goods
Swiecki (2017) analyzes which of the debated factors that determine structural change are most effective. Using a sample of 45 countries of varying levels of economic development, he tests international trade, sector-biased technological progress, non-homothetic tastes, and changing wedges between factor costs across sectors. His results show that sector-biased technological progress is the key factor to determine structural change, and, more interestingly for this research, is fundamental in understanding the shift from manufacturing to service sectors in developed countries.
Konan and Maskus's (2005) research shows that reducing trade barriers in manufacturing industries results in a small increase in welfare, but that in service industries results in a more dramatic increase in welfare that is reflected in all of a country's sectors. Their findings bring to light the notion that industry type matters in international trade. They conclude that one of the reasons developing countries have not realized the gains of trade liberalization is because of overprotection of their own domestic service industries. Indeed, going back to Siddiqui and Iqbal's (2005) research, Granger Causality tests on their results show that there is a significant positive relationship between GDP growth and investment, backing Konan and Maskus that FDI has a critical position in contributing to a country's welfare and possibly even trade. They estimate that around 75% of trade liberalization gains can be made from lifting the barriers on FDI alone. Interestingly they simulate the absence of trade restrictions in Tunisia and find that potential gains are three times as high from reducing tariffs on service sectors than on manufacturing sectors.
2.5. Volume vs variety
Buono and Lelane (2011) present another perspective to guage the effects of the WTO on international trade. They look at the effects of tariff reductions on two measures of trade: extensive margin, which is the number of exporters, and intensive margin of trade, which is the average volume exported by each exporter. Their research tracks the performances of a sample of French firms post the tariff reductions of the Uruguay Round spanning across 57 sectors and 147 countries.
Eaton et al. did a similar study on Colombian exporters and found that new entrants started with low volumes but as competition filered weaker players out, survivors grew fast and only a few years later,
dominated half the exports market of the country they were exporting to. This suggests that if decreasing tariffs leads to more firms exporting (i.e. affects trade via the extensive margin), those with competitive advantage, even new players, will be able to rise to dominate the entire market. If, on the other hand, decreasing tariffs leads to the same firms exporting more (i.e. affects trade via the intensive margin exclusively), the welfare gains are shifted from the entire economy to the exporters. Buono and Lenane's results show that reducing tariffs did not encourage more firms to export but only increased the volumes exported by already existing ones. This falls in line with Eicher and Henn's (2011) conclusions that often only countries that already had high market power prior to WTO showed to have been positively affected by the WTO.
Intensive and extensive margins of trade. Source: Dutt et al. (2013)
Building on Broda et al.'s (2006) results that trade in new varieties of products is the number one reason for rising growth in GDP, Dutt et al. (2013) investigate the effect of WTO membership on extensive product (trade in variety of goods) and intensive product (volume of trade in each good) margins of trade. Their gravity model concludes that WTO membership increases extensive product margin of trade by 25% but also has a negative impact on the intensive margin of 7%. They explain that these results suggest that the WTO works as a reduction in the fixed costs of trade rather than variable costs which implies trade barrier reductions lead to more entry and more competition; this ultimately means dilution of market shares and pulls down the average trade per firm. Dutt et al. (2013) also highlight that WTO membership encompasses more than just tariff rules
but also assure access to the international trade market which presents a whole new set of implications on the variables discussed around tariff reductions.
3. Data collection and analysis
3.1. Methodology
3.1.1. Components of the Uruguay Round
My regression model is comprised of variables inspired directly from the components of the Uruguay Round which are as follows:
- decrease agricultural subsidies - decrease restrictions on FDI - protect intellectual property
- increase trade in banking and insurance.
With these updates of the laws that govern international trade, the WTO aimed to increase global welfare by encouraging trade (Hanrahan, 2005) and simultaneously implies that these factors will have a significant impact on trade. I test whether each of these new additions of the Uruguay Round has had a significant impact, but first they must be quantified into measurable variables that can accurately reflect their effects.
Agricultural Subsidies
Cutting down subsidies in agriculture sectors is high on the priority list of the WTO. When India and Brazil argue against reductions of barriers to international trade for the upcoming Doha Round, one of the main points presented is the fact that in the case of many developing countries, agriculture contributes the greatest deal to their production and ultimately GDP, and rules like this put these countries at a disadvantage; the agriculture sectors of developing countries are also contended as still premature and in need of government investment. At the same time, agricultural subsidies in developed countries drive prices down which leads to income losses in developing countries (Faeth et al., 2007).
Ungor (2017) does comparisons of productivity patterns of Latin American countries and East Asian countries using a 9-sector general equilibrium model and finds that shifting labor away from the agriculture to other sectors can boost overall productivity growth. Chang (2009) on the other hand shows how heavy investment in farm education of Dutch farmers has over the years put the Netherlands on top of the global agriculture game. There were many variables to choose from to represent the position of a country in the agriculture sector; I found that agricultural land as a percentage of total land to be the most spot-on for this research. It is comparable across all countries of the world whatever differences in other characteristics they may have such as size, type and amount of crop grown. I collect this data from the World Bank data bank.
FDI
The notion of diminishing returns to capital point to the idea that investment capital flows from developed to developing countries. Because the ratio between capital and labor is much smaller in developing countries, capital invested will have higher relative returns than in a developed country where capital and labor are more in tandem (Lucas, 1990). The only problem is, this makes perfect sense but we never see it happening. Capital in the form of investment hardly, if ever, reaches developing countries. Robert Lucas (1990) provided an explanation which became known as the Lucas paradox. Simply put, the Lucas paradox explains that developing countries receive little capital investment due to international capital market imperfections, risks associated with investing in their industries, and asymmetric information.
Reducing the restrictions on FDI opens possibilities to invest abroad but risk aversion among investors seems to be a more common economic phenomenon than capital flowing into developing countries. Regardless, the WTO has deemed it a step towards higher trade and Siddiqui and Iqbal (2005) have found strong links between FDI and productivity growth.
I have used the natural logarithm of FDI net inflows collected from the World Bank to depict this component of the Uruguay Round; using the natural logarithm makes the FDI comparable across countries of different sizes and economies. Data on net FDI inflows was also collected from the World Bank.
Intellectual Property
Stricter protection of intellectual property rights is claimed to increase trade, but how much does intellectual property contribute to trade to begin with? Conceptualizing this into a variable was tricky at first; however, I believe using the natural logarithm of the number of patent applications is a good place to start. It is a better alternative to using the number of patents granted per country as this eliminates the effects of patent granting procedures which can differ greatly across countries. I found data on patent applications by residents and non-residents on the World Bank and summed them together for the total number of patent applications per country in a given year.
Insurance
Insurance is a major component in the service sector of any economy. Konan and Maskus (2008) found that reductions in barriers in services trade lead to more striking increases in welfare than those in goods trade, thus fueling the relevance of these industries in the international trade market. I use market share of insurance collected from the OECD website as the representative variable for the importance of the insurance sector in a country.
Other variables I add are rule of law, specifically freedom to trade internationally and legal systems and
property rights, based off the economic freedom index from the Fraser Institute, which gives a score between 0
Secondly, I add a dummy variable for the years that the country is a WTO member; 1 is for the years it is a member, 0 is for the years it is not. Rose (2003) used a binary variable for WTO/GATT membership, and Siddiqui and Iqbal (2005) used a dummy variable to observe WTO membership effects. I also add two interaction variables: one between the WTO dummy variable and the agriculture variable, and one between the WTO dummy and the insurance variable to check whether WTO membership causes the effects of agriculture or insurance to differ.
I use panel data to run a fixed-effects regression covering 81 countries over the years 1975-2015. My sample narrowed to the 81 countries for which data on the variables chosen was most available and this generally amounted to between 1,500 and 3,000 observations which set us off with a good start. The decision to use 20 years before and 20 years after the establishment of the WTO was to hopefully see the long term effects of the WTO on trade as the effects of economic policies are pronounced well into the future.
Below is a table with summary statistics on each variable used in the regression.
3.1.2. The regression model
The fruit of these ideas is a regression formula that looks like this,
TtoGDP = Β0 + Β1Ag + Β2In + Β3RL + Β4LnFDI + Β5WTO + Β6LnIp + Β7WTOAg + Β8WTOIn + e,
with the abbreviations standing for, TtoGDP → trade-to-GDP ratio
Ag → agricultural land as a percentage of total land In → market share of insurance companies
RL → rule of law
WTO → dummy variable
LnIp → natural logarithmof the number of patent applications. WTOAg and WTOIn are the interaction variables.
I have included developed countries to serve the purpose of a control variable as they, since the starting point of the study, 1975, generally have had relatively high and stable levels of the variables of the regression model. Below is a list of the countries selected for this study.
Developed Developing
Canada Belize Albania Bangladesh
USA Costa Rica Bulgaria India
Finland Dominican Republic Greece Iraq
France El Salvador Turkey Jordan
Germany Guatemala Algeria Nepal
Iceland Honduras Benin Oman
Italy Jamaica Cameroon Pakistan
Netherlands Mexico Chad Saudi Arabia
Norway Nicaragua Congo, Democratic Republic Syria
Portugal Panama Congo, Republic China
Spain Argentina Egypt Indonesia
Sweden Bolivia Gabon Malaysia
UK Brazil Ghana Mongolia
Bahrain Chile Kenya Papua New Guinea
Israel Colombia Madagascar Philippines
Kuwait Ecuador Malawi Thailand
Australia Peru Mali
Japan Uruguay Morocco
New Zealand Niger
South Korea Senegal
Sierra Leone
3.2. Results
This sub-section tackles the results of the study in three steps. I first discuss the coefficient estimates resulting from the fixed-effects regression, followed by results of yearly comparisons to observe time-specific effects over against a 1995 benchmark. More details of all the regressions, tests, and other commands can be found in the appendix section of the paper. I perform a total of 5 fixed-effects regressions on panel data in Stata. The regression equations are the same across all regressions but the differences lie in the data used which are as follows:
Regression 1: developed countries over the years 1975-1995 Regression 2: developed countries over the years 1996-2015 Regression 3: developing countries over the years 1975-1995
Regression 4: developing countries over the years 1996-2015 Regression 5: all countries over the years 1975-2015.
3.2.1. Regression results
The following estimates are obtained using the above data for the coefficients of the variables discussed earlier.
Regression 1 2 3 4 5 Ag –0.04410 (0.980) –0.32309(0.671) 1.21911(0.190) 11.68075(0.000) –0.22310(0.451) In –0.57453 (0.071) 0.08819(0.821) 0.09658(0.742) 0.12247(0.193) –0.17948(0.614) RL 1.51969 (0.000) –2.89057(0.099) 0.10113(0.914) –0.14949(0.920) –0.82854(0.134) LnFDI –0.96381 (0.324) 0.73034(0.163) 0.80801(0.306) 1.83990(0.046) 2.23496(0.000) WTO 0.00323 (0.999) 0 14.10407(0.339) 739.41530(0.000) 6.83796(0.235) LnIp –2.83385 (0.144) 3.76653(0.144) 7.61172(0.003) –0.31391(0.850) 1.22304(0.323 WTOAg 0.01223 (0.942) 0 –0.12439 (0.589) –13.15285 (0.000) 0.09846 (0.245) WTOIn 0.46568 (0.425) 0 –0.94958(0.223) 0.41361(0.492) –0.05063(0.904) Constant 80.31515 (0.374) 72.71960(0.053) –57.09961(0.194) –635.78070(0.000) 20.78431(0.221) F (8, 19) = 7.93 (5, 20) = 2.67 (8, 37) = 3.74 (6, 42) = 4.12 (8, 67) = 11.84 P-value 0.0001 0.0524 0.0027 0.0001 0.0000
Note: P-values in parentheses.
At first glance, one may notice from the above table is that the fourth regression, i.e. developing countries over the years 1996-2015, shows the greatest number of significant coefficients. This indicates that there possibly are differences in the expression of these variables between developing and developed countries, but generally, as regression 5 shows, the FDI variable seems to be the only one having a significant effect on the trade variable. Reversing the natural logarithm1, it can be seen that a 10% increase in the net FDI inflows results in an
increase in the trade-to-GDP ratio of 0.0971, a modest but significant result. This find is in line with previous research, i.e. Siddiqui and Iqbal (2005) found that investment growth was significant, Konan and Maskus (2008) estimated 75% of trade gains to be made by lifting barriers to FDI, and Chang (2008) attributes part of developed countries' success to their exercise of directing FDI in the sectors that require it before deregulation
laws were introduced. Although the coefficients of the remaining variables have turned out to be insignificant, they still spark discussion in the other regressions. Anderson et al. (2000) found that agriculture comprises of 7-8% of international trade in goods alone; this paired with the considerable growth in global service sectors can explain why the contribution of agriculture in the international trade environment is slowly disappearing. However, agricultural land as a percentage of GDP does have a positive and significant relationship with the trade variable for developing countries over the years 1996-2015. Additionally, the effects of the interaction variable WTOAg cannot be ignored here which show that WTO membership leads to a significant negative relationship of an even larger magnitude on trade-to-GDP ratio thus eliminating the positive effects of agriculture. This is in line with Chang's (2008) arguments that although agriculture can play a critical role in the welfare of a country, in the case of developing countries it is not often the case due to international trade laws being in favor of the developed world (Eicher and Henn, 2011). Moving on, the insurance sector does not seem to have had a significant effect on trade for developed and developing countries alike. Part of the reason for this could be that the insurance sectors for the majority of developing countries were still in their early stages over the time period chosen; for many countries an insurance market did not exist till well into the 1990s. An analysis of the effects of the insurance market would not be complete without taking into consideration the effects of the interaction variable WTOIn but that too has turned out to be insignificant throughout the regressions. RL, the rule of law variable, despite being narrowed down to specifically freedom to trade and legal systems and property rights, does not seem to have any significant effects for the large part except for a small positive effect on developed countries before joining the WTO. However, this variable is an index and possibly too abstract in its design to convey relationships in a model of this form. WTO membership on its own shows a positive impact on the trade-to-GDP ratio for developing countries from 1996-2015 indicating that it is likely that WTO membership took its toll on countries several years after being endorsed. More discussion on possible effects of WTO membership over the years follows this sub-section, in time-fixed effects as this can be better analyzed by taking a time dummy to observe time-specific effects on the trade-to-GDP ratio. Last but not least, the number of patent applications has a significant and positive impact only on developing countries over the years 1975-1995; it is worth noting that this effect disappears upon joining the WTO, i.e. over the years 1996-2015. However, there is much more to delve into here. Many steps go into completing a patent application and these can include how developed the research institutes are of a country, the level of education of the population, the number of PhD students, the composition of the economy i.e. are fields where innovation is thriving well developed, the rate of patent approval, and many more. It could very well be that the effects of this relationship were not fully conveyed due to how many smaller, but likely impactful, relationships were not taken into consideration within the limits of this study. Although that concludes the effects of the explanatory variables on the trade-to-GDP ratio within the regression, this study would not be complete without taking a closer look at a correlation matrix of all the individual variables as follows.
TtoGDP Ag In RL LnFDI WTO LnIp WTOAg WTOIn TtoGDP 1.0000 –0.2774 (0.0000) –0.0637 (0.0007) 0.1126 (0.0000) 0.0156 (0.4156) 0.2160 (0.0000) –0.2173 (0.0000) 0.0610 (0.0008) 0.0610 (0.0001) Ag –0.2774 (0.0000) 1.0000 0.0407 (0.0294) –0.0049 (0.8471) 0.0880 (0.0000) 0.0139 (0.4419) 0.0026 (0.9068) 0.4145 (0.0000) 0.0589 (0.0016) In –0.0637 (0.0007) 0.0407 (0.0294) 1.0000 0.2081 (0.0000) 0.2774 (0.0000) 0.1253 (0.0000) 0.2283 (0.0000) 0.1364 (0.0000) 0.7249 (0.0000) RL 0.1126 (0.0000) –0.0049 (0.8471) 0.2081 (0.0000) 1.0000 0.6420 (0.0000) 0.4386 (0.0000) 0.4492 (0.0000) 0.2644 (0.0000) 0.3058 (0.0000) LnFDI 0.0156 (0.4156) 0.0880 (0.0000) 0.2774 (0.0000) 0.6420 (0.0000) 1.0000 0.4594 (0.0000) 0.6979 (0.0000) 0.3943 (0.0000) 0.4022 (0.0000) WTO 0.2160 (0.0000) 0.0139 (0.4419) 0.1253 (0.0000) 0.4386 (0.0000) 0.4594 (0.0000) 1.0000 0.1601 (0.0000) 0.8002 (0.0000) 0.4881 (0.0000) LnIp –0.2173 (0.0000) 0.0026 (0.9068) 0.2283 (0.0000) 0.4492 (0.0000) 0.6979 (0.0000) 0.1601 (0.0000) 1.0000 0.1091 (0.0000) 0.3067 (0.0000) WTOAg 0.0610 (0.0008) 0.4145 (0.0000) 0.1364 (0.0000) 0.2644 (0.0000) 0.3943 (0.0000) 0.8002 (0.0000) 0.1091 (0.0000) 1.0000 0.4327 (0.0000) WTOIn 0.0610 (0.0001) 0.0589 (0.0016) 0.7249 (0.0000) 0.3058 (0.0000) 0.4022 (0.0000) 0.4881 (0.0000) 0.3067 (0.0000) 0.4327 (0.0000) 1.0000 Note: P-values in parentheses.
Two broad conclusions can be drawn from the results of this matrix: one that there are strong correlations between the independent variables of the study which partly explains why many coefficients came out insignificant in the regression, and two that the individual correlations of the independent variables with the dependent variable are significant. Contrary to regression 4, the agriculture variable seems to have a negative effect on trade-to-GDP ratio; an increase of 1% of the agricultural land of a country is correlated with a decrease of 0.2774 in the trade-to-GDP ratio of the country. Again it seems counterintuitive that the WTO should push trade in agriculture as top priority in the Uruguay and future Doha Rounds when it negatively affects trade however, WTO membership mitigates this negative effect, also contrary to regression 4, and turns it into an increase of 0.061. Market share of insurance of a country show the same pattern but to a lesser degree; increase in the market share of insurance of 1% translates to a decrease in trade of 0.0637 and WTO membership alleviates it to an increase of 0.061. One of the main arguments against reducing barriers to international trade, as presented by Chang (2008) among others, is that it leads to unfair competition between the fairly developed industries of advanced economies and the premature industries of the developing economies. It explains the negative correlation coefficients and also why the insurance variable has a smaller negative correlation coefficient than agricultural land as this is a relatively less developed industry in developing countries, leaving a smaller time span of progress to study. The rule of law, i.e. degree of freedom to trade
internationally, is positively correlated with the trade variable which is also reflected in regression 1. The FDI variable is not significant but like in the regression has a small positive effect on the trade-to-GDP ratio. Number of patent applications seems to be negatively correlated with trade-to-GDP ratio here, however, the many other factors that go into reflecting a country's level of activity in R&D cannot be ignored.
3.2.2. Time dummies
In this sub-section I present the results of taking a time dummy of 1995 to observe possible time-specific effects over the time span of this study. The following chart shows how the trade variable each year compares with the year 1995. More details on their exact values, standard errors, t and p-values, and confidence intervals can be found in the appendix.
Trade-To-GDP ratios against 1995 benchmark
At a glance it can be seen that the trade-to-GDP ratios of countries have generally been increasing from 1975 to 2015; the dips are smaller and the peaks are higher. Closer examination of the numbers to look for patterns gives four pieces of information. Firstly, 1995 seems to have been a depression in the international trade environment as trade-to-GDP ratios were in no year below this selected benchmark. This pattern could have been anticipated by economists at the time which is possibly (at least, I would like to think so) why the Uruguay Round was so timed. Another dip can be observed around 2009, a likely result of the American housing market crisis which trickled worldwide as financial markets had already become very international. The sharp decrease in the values of those financial assets impacted the value of traded goods but mainly services during the years
that followed. 2007 was a year of rapid growth; trade-to-GDP ratios are much higher in 2008 in comparison to other years of this study. This could be explained as the point at which the effects of the Uruguay Round came into full swing since economic policy usually takes a few years to fully come into effect. These increases continued up till the financial crisis after which it dropped sharply, and full recovery was achieved around 2011 but trade-to-GDP ratios from then on till the end of the studied period, 2015, seem to be slowly decreasing rather than continuing on the growth path before the crisis. The graph below shows these patterns as the average trade-to-GDP ratios of all countries over the years. What is striking is that members of the WTO consistently show lower trade-to-GDP ratios compared to non-members over the years, and this difference increases after the 1990s up till 2010, possibly suggesting that the benefits of guaranteed market access outweigh the losses from giving up industry protection.
Average trade-to-GDP ratios over the years. Source: Worldbank
4. Conclusions
Overall the regression results generally imply weak correlations between the Uruguay Round components and trade. Agriculture seems to positively affect trade in developing countries but WTO membership counters these effects and this is backed by a number of arguments. Firstly, as Chang (2008) explains, developed countries have used protectionist policies and agricultural subsidizing in the past which has allowed the industries of those countries to develop enough to stand a competitive chance on the international trade market. It follows that
despite having land to grow on, without the government investment into farming education and equipment, the benefits cannot be reaped and reflected on the international trade performance and ultimately welfare of the country. Although Konan and Maskus (2005) explain that services liberalization has much to offer in trade gains, results bring us back to the notion that maybe the service sector still needs to grow in developing countries before it can fairly compete via reductions in trade barriers. However, as Dutt et al. (2013) pointed out, WTO membership comprises more than reductions in trade barriers but also an assurance of market access and security of goods transfer which is also reflected in the positive effects of the WTO variable in regression 4. These assurances however do not seem to protect against global crises as shown by the results of the time-fixed effects regression. Furthermore, average trade-to-GDP ratios of member countries of the WTO have consistently been lower than those of non-members as can be seen from the graph above. This is a very blunt statement to make as there are likely still many minute relationships to be considered in order to understand the full extent of the mechanics of international trade. Furthermore, due to the high levels of autocorrelation between explanatory variables and consequently low significance of coefficients of these variables, few conclusions can be drawn in regard to the effects of the WTO. All-in-all, due to the vast amounts of research pointing away from WTO membership as an effective means to reap the benefits of international trade, maybe it is time to move in a different direction.
5. Limitations
The fact that the financial systems and insurance markets, generally speaking the service sectors, of the vast majority of developing countries over the time span chosen were still in their early stages or had not begun at all could have skewed the overall results. Also, although the regression was designed to analyze trade across borders, a very important variable was ignored in the regression which is exchange rates. Exchange rates or exchange rate fluctuations play a critical role in determining trade flows and their relationship throughout those same years can be an interesting topic to look into for future research but was unfortunately omitted due to lack of available data.
Furthermore, this research can certainly be expanded by going into greater depth in terms of the intellectual property variable. Little research has been done linking this aspect of the economy to international trade despite it being on the agenda of the WTO. However, this is no small topic as numerous factors go into play when one thinks about the mechanics behind the intellectual property industry of a country. These factors can include the level of education of the general population, research and development investments, number of research universities or institutions, how developed the technology sector is, government laws regarding patent applications, how easy the patent application process is, the time it takes for patents to be approved, just to name a few.
6. References
Anderson, K., Francois, J., Hertel, T., Hoekman, B., Martin, W. (2000). Potential gains from trade reform in the new millennium. Paper for Third Annual Conference on Global Economic Analysis, 1-20.
Brown, D.K., Deardorff, A.V., Fox, A.K., Stern, R.M. (2014). Computational Analysis of Goods and Services Liberalization in the Uruguay Round. Research Seminar in International Economics. Discussion Paper
379 (Researhgate publication 5174829, Conference on The Uruguay Round & Developing Countries, WB, 26-27 Jan, 1995.
Buono, I., Lalanne, G. (2012). The effect of the Uruguay round on the intensive and extensive margins of trade.
Journal of Int. Economics, 86, 269-283.
Chang, H.J. (2008). Under-explored Treasure Troves of Development Lessons – Lessons from the Histories of Small Rich European Countries (SRECs). University of Cambridge, 1-22.
Diao, X., Diaz-Bonilla, E., Robinson, S. (2003). How much does it hurt? The Impact of Agricultural Trade Policies on Developing Countries. IFPRI.
Dutt, P., Mihov, I., Zandt, T.V. (2013). The effect of WTO on the extensive and the intensive margins of trade.
Journal of International Economics, 91, 204-219.
Eicher, T.S., Henn, C. (2011). In search of WTO trade effects: Preferential Trade Agreements promote trade strongly but unevenly. Journal of International Economics, 83, 137-153.
Faeth, P., Fransen, L., Kurauchi, Y., La Vina, A. (2007). Agricultural Subsidies, Poverty and the Environment: Supporting a domestic reform agenda in developing countries. Policy Notes, World Resources Institute, No.1.
Green, K. P. (2018, April). A defining moment for Canada. Retrieved from https://www.fraserinstitute.org/tags/rule-law
Hanrahan, C.E., Schnepf, R. (2006). WTO Doha Round: The Agriculture Negotiations. CRS Report for Congress, pp. 1-40.
Hoekman, B., Mattoo, A. (2002). English P. Development, trade, and the WTO: A Handbook. Chap 6 The WTO:
Functions and Basic Principles, (41-49).
Konan, D.E., Maskus, K.E. (2006). Quantifying the impact of services liberalization in a developing country.
Journal of Development Economics, 81, 142-162.
Lucas, R.E. (1990). Why Doesn't Capital Flow from Rich to Poor Countries? The American Economic Review, Vol. 80, No.2, pp. 92-96.
Rifflart, C., Schweisguth, D. (2013). Draft Report on New Measures of International Trade. European Framework
for Measuring Progress.
Rose, A.K. (2003). Do we really know that the WTO increases trade? American Economic Review. Siddiqui, A.H., Iqbal, J. (2005). Impact of trade openness on output growth for Pakistan: an empirical
investigation. MPRA, No. 23757, 1(1), 1-8.
Subramanian, A., Wei, S.J. (2007). The WTO promotes trade, strongly but unevenly. Journal of International
Economics, 72, 151-175.
Swiecki, T. (2017). Determinants of structural change. Review of Economic Dynamics, 24, 95-131. Ungor, M. (2017). Productivity growth and labor reallocation: Latin America versus East Asia. Review of
Economic Dynamics, 24, 25-42.
United Nations Economic Commission for Europe [UNECE] Trade facilitation - principles and benefits. (2012) Retrieved from http://tfig. unece.org/details.html
WTO. A Year in Review 2013: Our Year. (2014) Retrieved from
https://www.wto.org/english/res_e/booksp_e/anrep_e/anrep14_chap2_e.pdf
WTO. Doha WTO ministerial 2001: Briefing Notes. Members and accession. Becoming a member of the WTO (2001) Retrieved from
https://www.wto.org/english/thewto_e/minist_e/min01_e/brief_e/brief19_e.htm WTO. Understanding the basics: The Uruguay Round. (2005) Retrieved from
