Protectionism in a globalising world : new era, new effects?

(1)

1

PROTECTIONISM IN A GLOBALISING WORLD:

NEW ERA, NEW EFFECTS?

By Siebe de Roest

Supervisor: Sebastian Krapohl

Master Thesis

MSc. Political Science: Political Economy University of Amsterdam

Graduate School of Social Sciences Date: 18-06-2019

(2)

2

1. Introduction

This thesis aims at the effect of protectionism on the economy in the era of globalisation. This topic has gained relevance, because after a long period of economic liberalism, the world has seen a re-emergence of protectionist measures in the past two years. Mostly, protectionist measures are condemned in the scholarly debate. However, despite the widespread condemnation of protectionism, protectionist measures have gained presence in the current global economy. Since this is a major change after prevailing economic liberalism, it is yet to be analysed what the effects of this trend are. Therefore, the topic is highly relevant in the current scholarly debate on use of protectionist measures.

Furthermore, the topic of this thesis contributes to the larger debate on the power of the state. The power of the state in the scope of globalisation is a highly discussed topic.1 As protectionist measures are means by which the state can exert its power over the economy, it is one factor of the power of the state.2 The term ‘protectionism’ refers to a group of measures a state can use to influence the economy.3 In this thesis, all possible forms of protectionist measures are addressed; this includes both non-tariff measures and tariff measures.

The effect of protectionism on the economy will assessed by examining the effectiveness of the aforementioned protectionist measures. Protectionist measures are primarily implemented with the purpose of protecting a domestic economy.4 The reasoning behind this purpose is that protectionist measures restricts importing practices or makes it more expensive. Due to higher extra costs on imports, domestic substitutes will be used as these are not covered by the protectionist measures. As domestic products or services will be acquired increasingly compared to foreign ones, the domestic economy is protected. In this research project, the process of protecting a domestic economy will be pointed out by examining trade flows of imports in four different sectors. If the protectionist measures would do their job, then the trade flows for the imported goods should show a dip in trading volume from the moment of the enactment of the protectionist measures onwards. For this reasoning, the power of the state will be addressed by examining the effectiveness of the protectionist measures on the trade flows.

1

Sharma (2006); Hirst & Thompson (1995); Jessop (2003)

2 Milner (1989, pp. 3-6) 3 Strange (1985, p. 235); Cohn (2016, p. 59) 4 Cohn (2016, p. 91)

(5)

5

The context of this thesis is the current era of globalisation. In this globalising world, the global economy is increasingly becoming interconnected.5 States and markets are interdependent and their economies likewise. In the period before the widespread globalisation, states were the foremost powers for changing the economy.6 Thus, states were the most powerful actors in the economy. Gradually however, states started shifting their power monopoly to international institutions.7 Moreover, the global economy has changed as it has become more interconnected. Companies have outsourced their supply chains, digital and physical improved infrastructures decreased relative distances, and markets ceased to be restricted by country boundaries.8 These factors, among others, have created a globalised world where the power of the state is more of a question mark than a given phenomenon; the interdependent character of states as of today, questions whether they still are these foremost powers in the economy. This thesis will assess whether the state can still influence the market if the state would like to do so, by assessing the effectiveness of protectionism in the era of globalisation. The question is whether the global economy has become so strong and rigid, that it could be too rigid for states to change it. This puzzle is the core puzzle of this research project.

Given economic logic, the enactment of protectionist measures should decrease trade flows; which is the result of several features, among which is a higher price on imported goods, which would lead to import substitution.9 However, as the global economy is increasingly becoming interconnected, the question arises whether this logic still holds. If the global economy is indeed becoming very rigid, maybe the protectionist measures do not have their desired effect. In other words, if the trade flows are not significantly affected by the protectionist measures, the global economy has proven its rigidness. This would indicate that the state is losing its predominant power over the economy. The research question of this thesis project is:

How has the effect of protectionist measures on trade flows changed through globalisation? This research question will be answered through two hypotheses, which will be answered by assessing the effect of protectionist measures on trade flows for four different sectors and five different trading relations. The sectors involved include two sector of commodity goods and

5

This will be further explained in detail in the next chapter.

6

Keohane (2005, pp. 49-51)

7

Keohane & Martin (1995, pp. 45-46)

8

Emmerij (1992, pp. 7-8)

9

(6)

6

two sectors of manufactured goods. Both the iron and steel sector and aluminium provide for sectors of commodities. The manufactured goods sectors are both the sector of the automobile industry and the sector of electronic machinery. These sectors are analysed for the bilateral trading relations between the United States (US) and Brazil, Canada, China, Germany, and Mexico. The analysis will provide two insights. First, it will be concluded that the effect of protectionism decreased parallel with increased globalisation. Second, the analysis will show that the effect of protectionism is larger for trade in commodities compared to trade in manufactured goods. From these two findings, the answer to the research question can be derived. The effect of protectionist measures on trade flows has decreased through globalisation, with a stronger effect for manufactured goods than for commodities.

(7)

7

2. Literature review

In this chapter, the literature to which this thesis contributes is presented. This thesis touches upon a larger debate through a more detailed debate, which is the core of this thesis. The larger debate concerns the relation between the state and the market in the current era of globalisation. Within this debate, the more detailed debate concerns the effectiveness of protectionist measures enacted by the state, in the era of globalisation. The latter is the focus point in this thesis. For sketching the whole picture, both the larger debate and the more detailed debate are discussed in this literature review. The set-up of this literature review is as follows. At first, paragraph 2.1 addresses the literature on the role of the state within the 21st century. Here, new challenges for the state in relation to globalisation are at the centre of the focus. In three different sections, the three main camps are put forward in this debate. The first camp consists of the scholars who argue that the state remains at the centre of power. The second camp consists of the scholars who argue for an altered position of the state, wherein the state is argued to remain at the centre of power albeit in a different environment. The third camp consists of the scholars who argue for the end of the state-centric power division. In the section thereafter, paragraph 2.2, the literature on the relation between the state and the market is put forward. The use of protectionist measures is at the forefront. The last section, paragraph 2.3, addresses the influence of protectionist measures for the economy; and in larger detail for trade flows. Following this set-up, all features of this thesis are put into scope. §2.1. The state within the 21st century

Scholars write on the different environment in which the state positions itself now, compared to decades ago.10 The era before globalisation is different for states compared to the current era of globalisation. In the current era, the state is experiencing new challenges and new forces that shape its environment. Widespread globalisation pushes forward new actors on the global scale and current scholarly debates question whether these new actors could replace the state as central power, or whether the world is as globalised as states allow it to be. Different state interests, different relevant actors, different markets, all kind of factors contribute to a different environment for the traditional state. Literature at hand is mostly focussing on the description of this environment and on the description of the behaviour of the state within this environment. For both elements, trends are brought into scope and are used to describe

10

For example, see Krasner (1976); Sharma (2006); Jessop (2003); Cox (1997). More sources will be referred to in the following sections.

(8)

8

possible future outlooks. In the academic debate on this issue, there are three main positions. The first position argues that the state remains at the centre of power. The second positions argues that the ‘nature’ of the state changes in order to be able to adapt to the new environment. The third positions argues that the state loses its centrality of power.

§2.1.1. The state remains at the centre of power

On the one side of the spectrum are the scholars who argue that the state remains at the centre of power. In 1976 already, Stephen Krasner discussed the trend within the political science literature, which pushes the state out of the centre of power.

“In recent years, students of international relations have multinationalized, transnationalized, bureaucratized, and transgovernmentalized the state until it has virtually

ceased to exist as an analytic construct.”11

Krasner explicitly rejects the proposition that the state is not relevant anymore. At the core of his argument, is that states themselves have chosen for interdependence, instead of facing a non-governable power called ‘interdependence’. However, it should be noted that this argument dates back to the 1970s, which is years before widespread globalisation took its pace.

More recently, Krasner’s view is supported again. Among the many ways by which the relevance of the state could be determined, Sharma addresses developmental issues as key factor. Sharma refers to the realm of development politics to highlight the importance of the state.12 All reports on development issues refer solely to states as entities, despite the big influence of companies and NGOs on the degree of development of a state. Yet Sharma acknowledges the disputable situation of the nation-state. At the core of this question, Sharma places two key elements of the state: territory and sovereignty.13 Concerning territory, Sharma claims that the territory of a state loses relevance given that people, goods, services and money cross borders more easily than ever before. Concerning sovereignty, Sharma mentions the World Trade Organisation as an international institution that dictates the policy freedom of states; though still subordinate to the voting behaviour of the member states. So yes, territory and sovereignty now have a different meaning, but the state remains relevant.

11 Krasner (1976, p. 317) 12 Sharma (2006, pp. 20-21) 13 Ibid (p. 6)

(9)

9 §2.1.2. The altered nature of the state

Within the spectrum of positions in this scholarly debate, a second position is a middle position in between scholars who argue for the centrality of the state against scholars who argue for the diminished relevance of the state. In this middle position, scholars argue that the state has to alter its nature. These scholars argue that the state shares its power with international and supranational institutions and that the global economy becomes stronger than the state. Among the literature supporting a state-centric view combined the rapid globalisation processes, are Hirst and Thompson. They argue that the national-level economic processes remain at the centre of importance and that the international economy is far from ungovernable.14 Their claims are based on the power division in international institutions. These institutions are relevant to the discussion as current global economic affairs are substantially guided through international institutions. For scholars who argue for the depletion of the state, these institutions are among the main reasons for pointing towards the end of the dominance of the state. Hirst and Thompson acknowledge that the international economic system has seen a change in its structure, where institutions are increasingly important for the shaping of the system. Contrarily to globalisation theorists however, they highlight that these institutions remain subordinate to states. It is the state who decides, by means of the policies of the institutions, how the power is divided and what is allowed to happen. Therefore, the ‘powerful’ international institutions are as powerful as states allow them to be. Altogether, these institutions remain subordinate to the states who are member or participant of them.

Furthermore, Hirst and Thompson argue that the world described as a world in the midst of widespread ‘globalisation’, is not a world in which states are small actors amongst a system of overlapping and competing agencies of governance.15 For the same logic as the former argument, they acknowledge the fragmented power relations in the global playing field, whilst stressing that merely the states are at the centre of that global playing field. However, they do acknowledge that this world might happen to develop itself in the future. This process could establish a globalised world system comparable to the Middle Ages. In such a system, there exist many authorities which are all competing for governance. Perhaps, that is happening

14

Hirst & Thompson (1995, abstract). It should be noted that this article is written in 1995, which is just at the very start of the rapid development of modern communication technologies. This means that the world changed significantly after their publication.

15

(10)

10

right now, 24 years after their publication. In essence, Hirst and Thompson argue that the function of the state has not changed dramatically due to globalisation.

Jessop adds to the scholarly discussion in the same ‘scholarly camp’ as Hirst and Thompson by addressing the new environment combined with the dominant position of states. Jessop does so by addressing the dependency of globalisation on state institutions and practices at many different territorial scales.16 Though, Jessop also acknowledges that the absolute separation between the political domain and the economic domain is becoming vague. In this new situation, the state promotes a continuous innovation and flexibility in relatively open economies; it is the task of the state to strengthen the structural and systemic competitiveness.17 The state then acts to transfer some power upwards towards international bodies and institutions, and to transfer some power downwards to lower-level governmental institutions. So yes, the state remains in the centre of power, but has different tasks now, according to Jessop.

Former theories related to a dualistic vision on the state-world relation. Robinson however, suggests a vastly different starting point for this analysis. According to Robinson, the future of the classic state is to become a transnational state, instead of a nation-state. This transnational state has both an economic element, which is not restricted by national boundaries, and a political element, which is restricted by its national boundaries. Consequently, a state emerges that is economically integrated in a global system and politically a ‘classic’ state. To this theoretical idea, Robinson has added to the scholarly debate.18 In essence, this view underscores the centrality of power at the side of the state, whilst acknowledging that the state has to change for fitting within the new environment. In this case, the new nature of the state contains an economic adaptation to the new environment combined with the preserved political power of the state.

§2.1.3. The state ceases to be at the centre of power

While the latter two visions hold on to the centrality of power at the state, with or without adaptation to the new environment, a third ‘scholarly camp’ introduces the vision that argues that the state loses its centrality of power. These scholars suggest an opposite side and acknowledge that a change is actually vivid. Given that both the environment and the behaviour of states within the new environment are changing, the question is what the 16 Jessop (2003, webpage). 17 Jessop (2003, conclusion) 18 Robinson (2001, pp. 158-167)

(11)

11

consequences would be for the power of the state. Among these scholars, it is believed that globalisation processes are more powerful than states, which creates a vastly different environment; an environment in which the state is not at the centre of power.

In this third ‘camp’, Susan Strange adds to the theoretical debate.19

Cox neatly defines the core proposition of Strange in his review:

“The ontology of Strange’s new realism includes a decline in the authority of states, an increase in the authority of big transnational firms, a parcelling of authority downwards from states to smaller territorial entities, along with a general erosion of power based on territory and a rise in non-territorial power in economy, technology

and communications.”20

The reason for the power shift towards other entities than merely states, is explained by Strange as a search for better ways of managing society.21 The outcome of such a search then, is global governance. While this term remains vague, it is clear that the state could not remain at the centre of power, according to Strange. It is argued that global governance includes different actors, which enlarges the playing field. Therefore, politics is not merely assigned to politicians, but also to big corporations and institutions. Foremost, transnational corporations are active in expressing their interests at the political stage; mostly with their own long-term survival as underlying motivation. As it is explained how all actors are pursuing their own objectives, Strange observes that, altogether, they form the foundation for the policies formulated by the governments of states. Hereby, the states become the means by which policy is formulated, rather than being the power that decides which policy is formulated. As the whole political process is scattered over many actors, the predominant position of the state is challenged by Strange.

Regardless of the camp chosen in the discussion on the relation between the state and the globalised world, it is undisputable that the world has changed in the realm of state-world relations. The effect of this new state-world relation is described in the work of Iversen and Wren.22 As is described earlier, the process of globalisation has created a new environment in which states have to position themselves. Among the challenges that this creates, is the increased unemployment rates as highlighted by Iversen and Wren. They argue that this is the 19 Strange (1996) 20 Cox (1997, pp. 367-368) 21 Strange (1996, p. 183) 22

(12)

12

result of the internationalisation of capital markets, and of growing levels of transnational investment, and of greater exposure to low-wage competition from newly industrialising countries.23 Iversen and Wren take growing unemployment rates as proxy for addressing the incapacity for states to deal with this changing nature. What this means, is that it is argued that the state cannot cope with the new environment. The relevance of this statement for this thesis, is that the new environment poses new challenges to states. Clearly, states might not be able to cope with domestic problems due to globalisation.

The new relation between the world and the state will be narrowed down towards the economic side of this debate. The relation between the state and the market will be the focus of this thesis. More specifically, the issue of protectionism will be the problem that this thesis addresses. The question is whether the state is still in charge of the economy with regard to protectionist measures. The analysis of this thesis will address this factor of the current state. §2.2. State power in the economy

In the past century, the global economy has seen waves of protectionism.24 The foremost period of large-scale protectionist measures is the period of the 1930s.25 After the depression, Reed Smoot and Willis Hawley promoted a bill with the purpose to keep out foreign goods from the United States. President Hoover signed the bill despite the advice of many academics to veto the legislation.26 The reasoning of the academics was mainly fear for a full-out trade war. After signing the bill, their fear became reality. The Smoot-Hawley act came to be known as the foremost example of protectionism and, moreover, the foremost example of the economic devastating impact protectionism can have. After the Second World War, Europe had to be rebuild and protectionist measures were mostly relieved and the US Marshall Plan was enacted. This time period became known for its economic growth; and the use of protectionist measures was proved to be harmful rather than helpful. The following two sections deal with literature on the impact protectionism in times where globalisation was at its start and in times of the current era of widespread globalisation.

23

Ibid (p. 507)

24

Thompson & Vescera (1992, pp. 514-515); Eichengreen (2002)

25

Irwin (1998, p. 326)

26

(13)

13 §2.2.1. Protectionism in the era of the beginning of globalisation

Mostly by the end of the second oil crisis, protectionism became vivid. This fits within the expected picture; after a crisis, protectionism becomes attractive.27 This trend accounts for the 1930s with the Smoot-Hawley Act that was enacted after the Great Depression, and again for the protectionist measures after the second oil crisis in 1979. For the latter, the period after the second oil crisis at the end of the 1970s, it was foremost the American automobile industry that was under pressure. The economy took a hit and cheaper imports from Japan were more attractive than domestic production, posing a major thread to the US car industry. For protecting the domestic production, the United States and Japan agreed on a voluntary export restraint.28 However, several years after this agreement, many trade barriers between the US and Japan, and between the EU and Japan had been put in place. 29 The question arises, is this inevitable? Does a country actually need protectionist measures causing a trade war, all for the sake of ‘saving’ its own domestic economy?

Lovins and Lovins formulate a negative answer to this question.30 They focus on the question of tariffs in the oil sector at the end of the 1980s. Their article describes the situation of the US oil market. Foreign oil became cheaper than US oil and that caused domestic trouble. While protectionism is the foremost answer to such an issue, Lovins and Lovins suggest three options. The first being protectionism, either including tariff walls for making foreign oil appear to be more expensive (while this is merely the tariff wall), or increasing taxes on oil (direct taxes on oil or taxes on gasoline).31 The second option is pursuing free trade and only purchasing the cheapest oil, which in this case is foreign oil.32 Ricardo would argue that the US should import oil, as it is not capable of producing it at low cost and should focus on the products it could produce for comparatively lower costs.33 The third option, and this would be the preferable option according to Lovins and Lovins, is the substitution option.34 Instead of protectionism, the US should push for avoiding the use of oil. Thus, at the core of their argument, is the fear for the consequences of protectionist measures. Lovins and Lovins stress the impact of protectionist measures for households, as these would incur the higher prices on

27

Erixon & Sally (2010, pp. 7-8)

28

Berry et al. (1999, p. 400)

29

Sassoon (1990, pp. 1-2)

30

Lovins & Lovins (1987)

31 Ibid (p. 1) 32 Ibid (pp. 1-4) 33 Ruffin (2002, pp. 729-731) 34

(14)

14

oil. It is also emphasized that the protectionist measures in one sector could have a spill over effect to other sectors. Altogether, by the end of the 1980s, the fear for protectionism is vivid.

§2.2.2. Protectionism in the era of widespread globalisation

In the former section, protectionism in the era before the widespread globalisation is discussed. It is noted that protectionist measures were feared for and that it caused several retaliatory measures in at the end of the 1970s and the beginning of the 1980s. Ever since, the world has seen a large flow of liberalisation policies; mainly through the WTO.

“The past decades have seen an unprecedented wave of liberalisation in trade and financial flows, which has enabled a rapid increase in the volume of these flows across countries. This evolution, which is part of the wider globalisation process, has undeniably brought prosperity to the world economy as a whole

(…).”35

However, the crisis of 2008 seems to have triggered a revival of protectionism. Brussière et al. write on the growing protectionist pressures; note that this article is written in 2011.36 In hindsight, protectionist pressures developed eventually into protectionist measures. It is argued that the protectionist measures that were already enacted by that time, had a moderate impact on the economy. The thread of protectionism on the other hand, is real according to Brussière et al. This is the result of the global economic flow of the time in which the article is written: in 2011, the global economy was recovering from the crisis. In such an instable period in time, protectionism can have devastating effects on economic growth worldwide.37 Furthermore, Viju and Kerr emphasize that political leaders from countries all over the world have acknowledged the destructive effects of protectionist measures in the wake of a crisis, and therefore condemn protectionist measures.38 They refer to the annual WTO report of 2009, where the Director General stated that countries seem to have learned from the 1930s, and acknowledge the common interest in an open global trade order. What is remarkable in this claim, is that political populism is strong in the same years. Economic mercantilism has strong support where the following logic is strong: ‘exports are good, and imports are bad’.39

35 Bussière et al. (2011, p. 5) 36 Ibid. 37

This effect is further explained in Brussière et al. (2011, pp. 7-9)

38

Viju & Kerr (2011)

39

(15)

15

Yet the protectionist measures to ensure such an economic policy, are not enacted by the time the article was written.

Taken to more recent events, the ‘Trump Tariffs’, it is clear that protectionism has risen eventually.40 In 2018, Trump has announced several import tariff on goods originating from all over the world.41 Calling on national security to cover the tariffs under WTO rulings, Trump claims to impose the tariffs; and did so accordingly. This bold economic strategy could be seen as surprising, especially since for many economists, and according to Viju and Kerr also political leaders, it is common sense that protectionism merely creates losers. Clearly, protectionism was seen as a historic concept in the globalisation era, up until Trump took office in the White House. The results of this bold economic strategy however, have yet to be explored.

40

Swanson (2018a); Swanson (2018b); Malawer (2018)

41

(16)

16

3. Theoretical framework

In the former section, the literature review, the scholarly debates concerning the topic of this thesis are put forward. Consequently, the different views on the power of the state and the different views on the effects of protectionism are explained. Moving on from this point, a theoretical framework is needed to function as fundament for constructing the hypotheses that drive this thesis. The theoretical framework is divided into three parts. The first part addresses theories that discuss the use- and existence of protectionism in the global economy. The major theoretical contributions stem from the two big schools of thought, realism and liberalism. Their contribution to the topic is put forward firstly. Moving from political science related theory towards the economic side of this topic, the fluctuation of trade flows will be discussed. Here, the view of Baldwin and Evenett will be taken into account, and the effect of protectionism will be discussed by referring to Baier and Bergstrand.42 The second part of the theoretical framework addresses theories that discuss growing globalisation. In this section, a theoretical distinction will be made between the internationalists and the hyper-globalists, and liberal institutionalists who are in between these two groups of scholars. The third part addresses more detailed, midrange theories that deal explicitly with the operationalisation of this thesis. This last part focusses on trade flows and protectionism at large in the globalised world. Here, the causal mechanism is explained from which the hypotheses are derived. §3.1. Protectionism and the global economy

When addressing the issue of protectionism in the global economy, the two major schools of thought in political science are necessary to include. Both realism and liberalism are relevant. Following claims of Joseph Nye, the liberal school of thought at large focusses on “the international society, interdependence and international institutions.”43

These three features all include cross-national factors and following that logic, liberals question the predominant power of the state. Moreover, the market is described as a fact from which states can benefit. According to liberals, through cooperation and through open trade, states can bear the fruit of the market.44 The market in itself is described as integrated in the global system and interconnected; the economy is the shaping factor in the globalised world. Therefore,

42

Baldwin & Evenett (2009); Baier & Bergstrand (1999)

43

Nye (1988, p. 238)

44

(17)

17

following the perspective of the liberals, the power of the state is small relative to the market.45

Bringing liberalism to a more detailed framework, towards the use and consequences of protectionism, a distinction is necessary between orthodox liberals and more modest liberals, i.e. Keynesian liberals. The former mainly criticises protectionist measures (and any kind of state intervention in the market) whereas the latter favours state intervention to a certain extend.4647 For orthodox liberals, it is argued that protectionist measures merely creates losers. Protectionism it is argued, is only beneficial on a short term for a single country. After this short term, retaliatory measures lure and a trade war could develop. This would pose a setback to all countries included. For Keynesian liberals, the argument is slightly ‘softer’. As described by Willoughby, Keynesian liberalism calls for state intervention in the market.48 Keynes questions the pure laissez-faire liberal view on the market; Willoughby even mentions that Keynes vision “reveals a profound distrust to the market mechanism”49

In short, the Keynesian liberals stress the importance of the market, though acknowledging that some state intervention is needed in certain circumstances.50

Shifting the focus to the other side of the spectrum, to realists, the stress is on a completely different factor compared to liberals. Realists put a high stress on the power of the state, instead of the market. Following this logic, the state remains at the centre of power, not the market. As the final call will always be at the side of the states, following realists it can be stated that the world is as globalised as the states allow it to be. While realists mainly acknowledge the importance of states in the global economic system, liberals also acknowledge the importance of states, but put a stress on international institutions through which cooperation between states is ensured. The other way around, realists would argue that the integration of international institutions is the result of state choices, not the result of an ‘uncontrollable’ power.51

This is where the two schools of thought differ. Following the realist perspective, the power of the state is undisputable. Furthermore, Nye states that the 45 Nye (1998, pp. 238-240) 46 Cohn (2016, p. 5) 47 Polanyi (1940, pp. 58-9) in Dale (2008, p. 16) 48 Willoughby (1982, p. 555) 49 Ibid (pp. 558-559) 50 Lacher (1999, p. 348) 51

Krasner (1976, p. 317). In the early days of globalisation, Krasner wrote on the schools of thought within international relations, who argue that the state ceases to have its importance. Although this period is vastly different from the period of the 2010s, his writings explain the realist perspective on the relation between the state and globalisation.

(18)

18

Realist view on states comprehends a learning factor.52 States learn from the outcomes of their policy and that is how policy changes emerge. What follows is a strong state that is able to adapt to what is needed.53 Therefore, the state not only remains at the centre of power, but also is also able to be there effectively.

In conclusion on these big schools of thought, it can be stated that the two streams both would acknowledge that protectionist measures have an effect on trade flows. They differ to which realists would argue that the environment in which this happening takes place favours the state over other powers, whereas liberals would argue that the state is not in the centre of power.

The origin of a downward movement of trade flows is a debated topic, wherein several factors are mentioned as leading explanations. Baldwin and Evenett foremost consider international supply chains as a major force behind decreasing trade flows.54 At the core of the argument, is that supply chains scattered over different countries magnify the effects of a change in demand for a certain good. If the demand in a country decreases, not only the manufactured good considered will experience a downturn in demand. All the components of which it consists, but also all the parts of which those components exist, and all the resources that form the base of those parts of which the components exist, will face a lower demand. Having all factors spread over different countries, would create a lower trade flow for all those countries, when demand is decreasing. This implicit effect of international supply chains, is what Baldwin and Evenett mention as the first factor of a decrease in trade flows.

The aforementioned decrease in demand is another point of discussion. A widely accepted view on this part, is that demand can decrease due to protectionist measures. Crean argues that protectionist economic policies have the power to establish an economic global downturn.55 This same notion is projected by Baier and Bergstrand. They build upon the article of Krugman who assessed why world trade has grown. 56 Krugman maps out differing views on the growth of world trade creating two camps. In the one camp, Krugman situates ‘the journalistic view’; this group views the growth of world trade as “driven by a technological

52 Nye (1998, p. 238) 53 Ibid. 54

Baldwin & Evenett (2009, pp. 2-3)

55

Crean (2009, p. 13)

56

(19)

19

imperative […], an irresistible force dissolving national boundaries”.57

In the other camp, Krugman places international economists. “[International economists view] the growth of trade as having essentially political causes [which caused] the removal of protectionist measures”.58

Within these two camps, Baier and Bergstrand build upon the second position. In their paper, it is described to what extend in relative terms income growth, income convergence, tariff reductions, and transport-cost declines contributed to the growth of world trade. Among the four factors, it is found that tariff-rate reductions explain for 23-26% the growth in world trade.59 In this sense, they have proved that the international economists are correct to a certain extent, given that the state has the power to have influence on the market. §3.2. Globalisation and the global economy

This section covers more detailed, midrange theories. The vision of these theories on globalization with regard to the global economy are put forward. The impact of globalisation seems not to be a clear-cut and one-sided fact. At the one end of the spectrum are extreme or hyper-globalists, who believe that globalization involves the creation of a “borderless world” in which multinational corporations lose their national identities, and regional and global markets replace national economies. This theoretical group stresses the cross-border element of globalisation. For example, Kenichi Ohmae asserts in The End of the Nation-State that the relevance of traditional nation states has decreased; nation states are unnatural and impossible business units in a global economy, according to Ohmae.60 When there is no longer state interference, Ohmae argues, MNC’s decisions and consumer choices will result in the rational allocation of global resources. According to this logic, states cease to be relevant, let alone effective, in managing the economy.61 Ohmae puts it bluntly:

“Traditional nation states have become unnatural, even impossible, business units in

a global economy.”62

These scholars, especially, question what role the state has that cannot be fulfilled by transnational forces. If the power of the state is indeed downscaled to providing basic needs, what then remains at its power centrality? One core function that could still be mentioned as a core practice of the state, is providing the rule of law. However, Drucker argues that in the 57 Krugman (1995, p. 328) 58 Ibid. 59

Baier & Bergstrand (1999, p. 23)

60 Ohmae (1995, p. 5) 61 Ohmae (1995, p. viii) 62 Ibid (p. 5)

(20)

20

current scope of emerging transnational companies, national law is also losing relevance given that organisations are increasingly transnational instead of multinational. In other words, these organisations are not bound to countries anymore and operate on a global level instead of a multinational one. Therefore, Drucker suggests, there is a challenge in finding ways to develop international law and supranational organisations that can enforce that law.63 According to hyper-globalists, the role of the state concerning any task is losing relevance due to globalisation. There is more attention devoted to the views of internationalists and moderate globalists than to hyper-globalists.

The internationalists are at the other end of the spectrum of globalization theories. This theoretical group acknowledges that globalisation causes more frequent cross-border interaction, but stresses these interactions remain between states. Although they recognize that interdependence is increasing and that non-state actors have a role in IPE, they believe that the world is no more “global” than it was in the nineteenth century. The international economy in the view of internationalists remains to be dominated by exchanges between national economies.64 Although some internationalists acknowledge that globalization today may be different because of the greater speed and volume of transactions, they see globalisation mainly as an economic phenomenon. In addition, they argue that in most areas of world politics, states are still the principal authorities.65

Thus, globalisation keeps the states at the centre, internationalists will argue. For Cohn, globalisation brings states –and people- even closer together. Though, Cohn acknowledges that globalisation changes the rules of the game. Cohn stated that:

“Globalisation involves the broadening and deepening of interdependence among peoples

and states”.66

Moving from the internationalist perspective towards hyper-globalists, liberal institutionalists are in the middle. By and large, this school of thought would argue that the state is constrained due to globalisation.67 Following this logic, the power of the state over the market is low in the era of globalisation. Altering the market by, for instance, imposing protectionist

63 Drucker (1997, pp. 167-169) 64 Cohn (2016, p. 7) 65 Ibid. (p. 7) 66 Ibid. (p. 6) 67 Weiss (2003, pp. 2-3)

(21)

21

measures would not have a substantial effect due to the minor power of the state. As Weiss puts it:

“Managing the national economy […] [is] nowadays seen to be increasingly beyond the

capacities of territorially centred states”.68

Weiss claims that modern countries are moving away from coordinated market economies, and towards a liberal market system. In this system, the role of the state is constraint and mainly focussed on providing basic needs rather than controlling the market. To this modest vision on the future of the state, liberal institutionalists add to the scholarly debate. In the realm of the power of the state, liberal institutionalists argue that international institutions increasingly replace the role of the state. In this school of thought, it is argued that the role of the state has changed dramatically ever since the growing power of international institutions. Institutions like the United Nations, the World Trade Organisation, the IMF, and the World Bank prescribe norms and rules for the global system and more precisely, for the global economy. States are at the centre of these institutions, albeit still subordinate to the majority vote and the policies of that institution.69 Thereby, the formal structure of these institutions where states are still at the decision making table, is underscored. The outcomes of the majority vote in the international institutions decide for the policies that are put forward in individual states.

In short, two main theoretical groups of scholars are to be identified. At the one end of the spectrum are hyper-globalists. This group argues that the states ceases to be a relevant unit of analysis and that the state is not suited in the global economy anymore. At the other side of the spectrum are internationalists. According to this scholarly group, globalisation is changing the rules of the game rather than eliminating the state from the game. A third group of scholars, in between the internationalists and hyper-globalists, are liberal institutionalists who argue that states are increasingly replaced by international institutions. Given the more nuanced view of internationalists, this theory will provide for the core of the hypotheses that are put forward in paragraph 3.4.

68

Weiss (2003, p. 3)

69

(22)

22

§3.3. Causal mechanism: protectionism loses power due to global value chains

For now, it is argued that the power of the state might be in decline. From this statement, using deductive reasoning, the state also has declining power of the economy. Since protectionist measures are the foremost tools for states to influence the market, the effect of protectionist measures should decline in the scope of the decreasing power of the state. This section will elaborate on the causal mechanism driving this logic.

The relation between the decline of the state and the current outlook of the global economy, characterised by the emergence of global value chains, is described by Miyoshi.70 Instead of the state as the centre of power in the global system, transnational corporations in the global economy are described as the centre of power. These corporations have global value chains; overseas distribution of manufacturing, marketing and sales which indicate that their operations are not nationalised anymore.71 This phenomenon is not unique to the time period in which the Miyoshi article is written. The current world also knows many companies operating in different countries.72 The consequence of these ‘trans-nationalized’ structures of companies, is that these companies have anchored themselves in several countries. Cantwell contributes to this claim, by explaining that firms make use of country-specific comparative advantages for choosing which part of the firm is to be settled in which country.73 This has become possible since the era of globalisation, as in this time period, firms have been enabled to outsource their different business units due to advances in transport and communications technologies.74

The relevance of global value chains differs per sector. Mostly, sectors in which many manufacturers are involved know global value chains. Consequently, in sectors in which only a few manufacturers are involved in the production of a good, global value chains are less likely to emerge. This logic is put forward by Gereffi, who stresses that global value chains are a characteristic of capital- and technology-intensive industries where production is

70

Miyoshi (1993, pp. 735-744)

71

Ibid (p. 736); an example is provided by Sturgeon & Van Biesebroeck (2009, pp. 2-7) and by Baldwin & Taglioni (2009, p. 54)

72

Examples are global value chains and outsourced supply chains as described in Kenney et al. (2009) and in Hunphrey & Schmitz (2001, pp. 19-20)

73

Cantwell (2009, p. 36)

74

(23)

23

layered involving thousands of firms.75 This follows from the logic that a raw material does not need many ‘components’. Manufactured goods however, which consist of many parts, do have many components making global value chains more relevant.

The widespread emergence of global value chains as described above, has its consequences for the global economy. Altomonte and Ottaviano found that global supply chains have made the international trading system more resilient to external shocks.76 This is the result of long-lasting relationships and contracts, by which firms anchor themselves to partner countries and are thus not able to shift away the related trade flows in the short run. In addition, they refer to the substantial sunk costs, which are incurred in setting up the global value chain. This would form a motivation for not changing the value chain.77 Altomonte and Ottaviano looked at trade flows between the four biggest European economies and Central and Eastern Europe.78 They found that for countries trading more within supply chains, the effects of the financial crisis of 2008 were smaller compared to countries trading less within supply chains. From this finding, they conclude that global value chains make an economy more resilient to external shocks. Thus, their logic is based on the effects of a crisis. However, Krueger formulates an argument following the same logic for other external shocks.

It is the same relation between global value chains and economic rigidity that is put forward by Krueger. Furthermore and in more detail, Krueger stresses the effect of protectionist measures on trade flows in this new system of global value chains. It is stressed how the global economy took a hit after the financial crisis of 2008, yet still did not collapse as bad as could be expected. The following quote puts is neatly.

“Economies are more closely interconnected than at any time in the past. The supply chains for many goods and services stretch across a number of countries, and the truth of this statement has been vividly demonstrated to politicians and businessmen alike. And, while there have been a large number of protectionist measures, the volume of trade affected has, to date, been much smaller than might have been anticipated given the severity of the

downturn.”79

75

Gereffi (1999, p. 1)

76

Altomonte & Ottaviano (2009, p. 97)

77 Ibid. 78 Ibid (p. 98) 79 Krueger (2009, p. 22)

(24)

24

Thus, it is clear that the effect of protectionist measures on trade flows is negatively correlated to the presence of global value chains. This logic is at the core of this thesis as it explains the causal mechanism between protectionist measures and trade flows through globalisation. Moreover, it stresses that, in recent years, the global economy has seen value chains globalising and that this feature makes the economy more rigid in terms of the effect of external shocks to trade flows.

§3.4. Hypotheses

The theoretical framework has provided several theoretical insights, which contribute to the theoretical foundation of this thesis. Over the time period that is considered in this thesis, 1996 to 2018, the world economy has seen an increase in global value chains. It is also stressed that global value chains cause the related economy to have a higher rigidity compared to economies that engage in global value chains to a lower extend. Given this stress on increased rigidness through globalisation, my hypotheses follow from this theory. Also, it is stressed that manufactured goods have a more vertically integrated supply chain compared to commodity goods. Therefore, the effect of protectionist measures on trade flows should be lower in sectors of manufactured goods compared to sectors of commodity goods. Given these two logics, I come to the following two hypotheses:

H1. The effect of protectionist measures on trade flows decreases when globalisation increases.

H2. The effect of protectionist measures on trade flows is higher for commodity goods compared to manufactured goods.

The reasoning behind these two hypotheses is a combination of the internationalists and the hyper-globalists. These theories on globalisation argue that the globalisation process changes the environments in which states are; though hyper-globalists complete drive states out of the epicentre of power. To them, the global economy has become interrelated to an extent to which the state is out-powered by economic interests. Also driving these two hypotheses, is the vision from Krueger, who argues that the global value chains are so strong, that industries lose flexibility in avoiding the protectionist measures. Therefore, industries merely incur the higher costs generated by protectionism. This logic is at the core of the two hypotheses. The school of realists, however, would not agree to these two hypotheses. They would claim that the world is as globalised as states allow it to be. Following this reasoning, states not allowing for further globalisation would indeed be able to downsize this effect by imposing

(25)

25

protectionist measures. The protectionist measures would show their effect on trade flows. Be the hypotheses falsified, then the realists are correct.

4. Methodology

For this research project, a mixed method research design is chosen with a deductive research project. For the research project, the time period is set from the year after the creation of the WTO, 1996, to present. This is done as trade flows are analysed by assessing the change in trade flows compared to the preceding year. Given that the WTO was established in 1995, the year 1996 is the starting point of this analysis.80 Both quantitative and qualitative strategies will be used. The quantitative part of this thesis will provide insights using statistical data. The analysis will be developed upon a panel data analysis over which a regression is run. Chapter 4.2 will explain this method in more detail. The qualitative part of this thesis will be more in depth, building on the insights provided by the quantitative analysis. Different sources will provide for the data that is used in this qualitative analysis. Chapter 4.3 will elaborate on this part of the analysis.

Within the domain of research methods, this thesis follows deductive reasoning instead of inductive reasoning. Therefore, the theoretical framework functions as the starting point of this thesis.81 Several theories are put forward which all address the same issue, each in their own way. From the logics as formulated by the relevant theories, the hypotheses will be derived. These hypotheses form the core of the thesis and will lead to the answer on the research question. The hypotheses will be tested by making use of the data, which is quantitatively generated. It will then be decided whether the hypotheses are to be rejected or not. This style of reasoning falls within the scope of deductive reasoning.82

§4.1. Case selection

As protectionist measures are at the core of this thesis, a central country should be selected that makes use of these measures. Given the historic and present use of protectionist measures by the United States, the US is taken as central country for the analysis.83 For assessing trade relations, including all countries is not doable given the size of such an inquiry; as such an 80 Hoekman (2002, p. 41) 81 Bryman (2012, p. 69) 82 Bryman (2012, pp. 24-25) 83

See The White House (2002); Business Insider (Bryan, 2019); Al Jazeera (Murray, 2019a); Reuters (Lawder, Qiu,& Lee, 2019)

(26)

26

inquiry would include both trade flows and protectionist measures for bilateral relations of all countries worldwide. Therefore, the biggest trading partners from all relevant continents are selected. These five trading partners are all among the largest trading partners of the US. For geographic consistency, the five trading partners originate from different continents; except for the fact that the African and Australian continent are not addressed, because countries from these two continents are not in the top trading partners.84 According to US Census Bureau, from Central and North America, Mexico and Canada are the top trading partners. From South America, Brazil is the top trading partner. From the European continent, Germany is the top trading partner.85 From the Asian continent, China is the top trading partner. For this reasoning, these five countries are brought into the analysis of this thesis. Concerning the trade flows to be assessed, the different sectors that will be analysed have to be comparable. Therefore, the sectors should be relevant for the whole time period that is taken into account. Also, the theoretical framework pointed out that global value chains should lower the effect of protectionist measures. Therefore, both sectors that have widespread global value chains and sectors that are not typically characterised by global value chains have to be included for comparability. As commodities have less widespread global value chains compared to manufactured goods, both these two sectors should be analysed. Given these two criteria, this thesis includes the following four sectors. The two sectors that are typically commodity goods are both the sectors of iron and steel, and aluminium. The two sectors that contain manufactured goods are the automobiles sector, and the sector of electronic machinery. All four sectors are relevant during the whole period of the analysis, as they are not recently developed sectors.

§4.2. Operationalisation of variables

The analysis deals with trade flows as dependent variable, and both protectionist measures and globalisation as independent variables. In this section, the operationalisation of these variables is put forward. Starting with the dependent variable, it is to be noted that only trade flows for imports are included in the analysis. Including both import and export flows implicates that data has to be gathered for all bilateral trade flows. In addition, data on all protectionist measures has to be gathered for both sides of the bilateral relation. For timing issues, it is chosen to only focus on the import trade flows. This chosen over export trade flows as protectionist measures mainly aim at protecting one’s economy, i.e. creating a trade

84

US Census Bureau (2019, top trading partners)

85

(27)

27

surplus rather than a trade deficit.86 Therefore, import trade flows are more relevant to this thesis compared to export trade flows.

The independent variable on protectionist measures is set up to measure the correlation between the import trade flows and protectionist measures, through globalisation (this independent variable is explained later on). In the quantitative analysis, only non-tariff measures are included. The reason for this, is that applied tariffs have not changed significantly in the time period of this thesis. Applied tariffs have changed mostly in 2017 and in 2018.87 However, for these years, globalisation scores (explained later on) are not available. For this reason, the years 2017 and 2018 cannot be analysed quantitatively. As these years are the most interesting for the analysis, given the presence of increased tariffs, the tariff measures are not included into the quantitative analysis. For non-tariff measures, all types of non-tariff measures are included into the analysis. The United Nations Conference on Trade and Development (UNCTAD) refers to non-tariff measures as policy measures other than ordinary customs tariffs, which have an economic effect on trade.88 These types of measures are sanitary and phytosanitary measures; technical barriers to trade; pre-shipment inspection and other formalities; contingent trade protective measures (mainly antidumping measures and countervailing duties); non-automatic licensing, quotas, prohibitions and quantity control; price-control measures; export-related measures. All these types of measures are taken into account.

The independent variable on globalisation is operationalised via a globalisation score. In the literature, several indices for globalisation exist among which the KOF index is the index most widely used.89 “The 2013 KOF index cumulates 23 variables to an overall globalisation index and three sub-indices covering the economic, social and political dimensions of globalisation”.90

The economic globalisation taken solely is not relevant for this thesis project, as the purpose of the index for this thesis comprehends the overall outlook of globalisation and not merely the economic side. As the whole process of globalisation is to be described by the index, the general globalisation index will be used. The shortcoming of the globalisation

86

Feenstra (1992, pp. 159-160); Milner (1989, pp. 4-6); Zedillo (2009, p. 18)

87

Rooney (2018)

88

United Nations (2017, UNCTAD website)

89

The Kearney/Foreign Policy Magazine globalization index; the CSGR Globalization index; the GlobalIndex; the Maastricht Globalisation index; derived from Potrafke (2015, p. 2)

90

(28)

28

indices is the subjectivity in the variables from which the indices are comprised.91 The indices are based on components for which data was available and for which the data had sufficient quality. For this reason some, perhaps relevant, factors are left out. The other side of this feature, is that this way of constructing an index makes it possible to produce an index including many factors thus creating a wide overview. For this characteristic of the KOF index, and for its widely renowned status, the KOF index is used in this thesis.

§4.3. Method of Analysis

The quantitative analysis is performed with a panel data regression. This method is chosen as the dataset that is developed for this thesis, is a panel data set. This is a dataset which has a number of cases that are observed over a period of time.92 As this thesis works with six countries which are analysed with annual data, the dataset that follows is a panel data set. In addition, all countries are observed for all time periods, so the panel data is balanced.93 For panel data, a distinction has to be made between a model either for random effects or for fixed effects. There are two ways for assessing whether a random effects or a fixed effects model is suited. In this thesis, the independent variables, i.e. protectionist measures and globalisation scores, are specific for the countries included in the analysis. This means that for the countries analysed, the intercepts vary. As the fixed effect model examines differences in intercepts, this model is suited for this thesis.94 A second way to assess whether a fixed- or a random effects model is suited, is by performing the f-test for the data that is used. Be the f-test zero, a random effects model is suited. Be the f-test non-zero, a fixed effects model is suited. Running the f-test for this dataset, yields differing f-scores varying per dataset, though all significant, indicating a best fit with the fixed effects model.

For this fixed effect model, a linear regression is performed. The hypotheses of this thesis address sector specific features, though not country specific features. Therefore, the regression will be performed for each sector separately but not for each bilateral trade relation separately. The output, thus, is presented per sector. The effect of protectionist measures on trade flows will be analysed year by year, regressed to an interaction term consisting of the number of protectionist measures and the globalisation index. The percentage of change in trade flows is the independent variable, the Y- variable. The number of protectionist measures 91 Ibid (p. 4) 92 Hsiao (2014, p. 9) 93

Schmidheiny & Basel (2011, p. 2)

94

(29)

29

is the first independent variable, X1#Prot. The second independent variable is the globalisation score, X2Globsc. The third parameter is the interaction term, X1*2. This term is the product of terms X1 and X2. The next term in the equation, is the control variable X3. This variable checks whether the relationship between the independent variables and the dependent variables are not caused by a third variable.95

In this analysis, the dependent variable is the considered trade flow. As trade flows have a fluctuation outside the scope of protectionist measures and a globalisation score, there has to be a control variable for the fluctuations that are already in place. Therefore, a control variable is needed for overall differences in trade flows. As the US imports are reported for bilateral import flows, the total US import value for the sector considered is taken as control variable, variable γA1. The standard linear regression formula that follows has the following basic form:96

Y1 = α + βX1 + βX2 + βX3 + γA1 + ε

For inserting the variables mentioned above into the linear regression equation, the parameters are formulated in codes. As the regression will address pairs of countries, each parameter contains a dyad over two countries, which is referred to as dyad ij. Also, the regression will be performed per year, which is referred to as year k. For the parameters specifically, the formulation is as follows. The percentage of change in trade flows per pair of countries per year, is mentioned as % change trade dyad ij, year k. The number of protectionist measures per pair of countries per year, is formulated as #Prot dyad ij, year k. The globalisation score per pair of countries per year, is mentioned as Globsc dyad ij, year k. Finally, the control variable is mentioned as Totimp. Translating the variables of this analysis into a linear regression analysis, generates the following equation:

% change trade dyad ij, year k = #Prot dyad ij, year k + Globsc dyad ij, year k + (#Prot dyad ij, year k * Globsc dyad ij, year k) + γTotimp + ε

This equation will indicate whether the change of effect of protectionist measures on trade flows with respect to globalisation is significant or not. To test whether this is different for either trade in commodities or trade in manufactured goods, there will be a separate section in the analysis on this comparison. For comparing the two larger groups of sectors, a dummy variable is set up. This dummy variable makes groups per combination of country and sector

95

Healey (2012, p. 407)

96

(30)

30

–being either commodities or manufactured goods. Making these groups provides ten units. By running a panel data regression, with fixed effects, the dummy variable will show whether assessing commodities or manufactured goods makes a significant difference for the dependent variable.

The qualitative analysis finalizes the analysis. As mentioned before, a regression analysis for analysing tariff measures is not a good fit. This is the result of the fact that 2017 and 2018 are the most interesting years concerning import tariffs, whereas there is no globalisation score for these years. In addition, tariffs have not changed much during the years before 2017 and 2018. Therefore, running a regression analysis is not a good fit for this analysis. A qualitative analysis can provide more in depth insights. In this part of the analysis, descriptive statistics will provide for the focus areas. Moments in which tariffs increased rapidly are specified. These, then, are put next to the trade flows of the corresponding years. In the analysis, it will be described whether the expected happens, i.e. imposed tariffs are accompanied with decreased trade flows, or whether this is not the case. Next to these observations, government documents, scholarly articles, and news articles will provide for information explaining the descriptive statistics in further detail. In short, a comparison between the tariff measures and the trade flows, combined with insights provided by the aforementioned sources, will provide for the qualitative analysis on the effect of tariff measures on the trade flows.

The presentation of the output of the qualitative analysis is slightly different from the quantitative analysis, as the outputs of the quantitative analysis are presented per sector. This the result of performing a regression analysis per sector for all countries together in one data set. In STATA then, the countries are grouped, which makes the analysis bilateral. This is necessary as the trade flows are bilateral and the non-tariff measures are also bilateral. Still though, outputs are presented per sector and not per country as the hypotheses do not address country specifics. However, in the qualitative analysis, the data and the analysis are presented per dyad of countries, bilaterally. This means that for each bilateral trade relation, both the trade flows and tariff measures are bilaterally analysed. This is the result of the nature of the data; all trade relations are bilateral and all tariff measures are also bilateral. As this data has to be analysed qualitatively, rather than via STATA, the data has to be presented bilaterally. §4.3. Validity and reliability

By performing a panel data linear regression with trade flows as dependent variable, internal validity might be weak as trade flows can be affected by many external factors- not only

Protectionism in a globalising world : new era, new effects?