Trade Policy Decision-Making in Majoritarian Systems;

(1)

Trade Policy Decision-Making in Majoritarian Systems

An analysis of the causes and consequences of the protectionist bias in

majoritarian electoral systems

July 5, 2019 – Leiden University

Name:

Student Number:

Program: MA International Relations, Global Political Economy Thesis Supervisor: Dr. J. Fynn-Paul

(2)

1. Introduction

Economists have long argued that free trade, without barriers, is the most beneficial trade policy. For equally as long, they have wondered why politicians do not seem to follow their advice. After all, quantitative research has long established that trade wars have an overall negative effect on countries involved in it. The positive effects a trade war might have for specific sectors in a state cannot outweigh the negative overall effects for its economy. The dynamic seems apparent: once states start applying increasingly high tariffs to an increasing amount of imports, these actions are often reciprocal. Consequently, their own exports are also increasingly taxed and thus decreasingly competitive. A less competitive export sector in turn has negative effects on domestic economic growth, as exporting companies go out of business and workers lose their jobs.

In literature, a number of possible explanations have been proposed for why protectionism continues to be a prevalent force. All of these possible explanations, however, are dependent on a certain political mechanism – and as such, the policy outcomes under this mechanism are too. In this thesis, the relationship between a state’s electoral system, the manner in which it selects its legislators and decision makers, and a state’s trade policy will be further explored. This observation is clearly too broad to serve as a thesis topic, but the relationship between electoral systems and trade policy is nonetheless an interesting one that I wish to research further. I intend to do so by looking at the policy-making process behind foreign trade policy, through an analysis of two types of protectionist measures: both tariff as well as non-tariff barriers. Conventional wisdom in academic literature on the policy-making process behind protectionist trade policy suggests that majoritarian democracies, most often associated with a district or first-past-the-post electoral systems, have a distinct protectionist bias. However, most research and academic literature on this subject either covers only protectionist measures consisting of tariff measures, and not protectionist measures other than tariffs. By only considering tariffs, these studies fail to address the possibility that electoral systems have an ambiguous effect on protectionist bias. Furthermore, many of these sources make a convincing argument demonstrating the protectionist bias in majoritarian systems, but often lack an analysis of possible explanations behind this tendency.

Building on this observation, I intend to test the hypothesis that states with a majoritarian electoral system have a protectionist bias and analyze the reasons behind this protectionist

(4)

4

bias by looking at the policy-making process in one of the largest majoritarian democracies and trading nations in the world: the United States. Firstly, I will review the various characteristics that are central to the two main topics in this thesis: electoral systems and protectionist policies. Thereafter, I will look at the relationship between these two topics, by analyzing the effects the characteristics of the different electoral systems have on the two main means of protectionism: tariffs and non-tariff barriers. Subsequently, I will incorporate this analysis into the development of a theoretical model of the different actors and variables that influence trade policy-making. Finally, I will test the hypothesis by reviewing the trade policy of one of the largest trading countries and majoritarian democracies of the world, the United States, before reaching a conclusion.

(5)

5

2. Electoral systems and policy outcomes

The question of the relationship between electoral systems and protectionism touches upon a number of matters that are not easily defined. What kind of policies can be characterized as protectionism is not always immediately evident. Some policies might have a de jure objective of protecting the environment or consumers, whilst de facto posing a barrier to international trade. Furthermore, within the context of the relationship between protectionism and electoral systems, it is helpful to understand how different electoral systems can be distinguished, what their characteristics are and what the literature says about the potential policy effects of different kinds of electoral systems.

2.1. Types of electoral systems

In a representative democracy, the electorate chooses a group of people to represent them and to make laws and decisions on their behalf. However, the manner in which these representatives are chosen, known as the electoral system, varies from country to country. Broadly speaking, one can identify three different types of electoral system. In a plurality electoral system, also known as a “first-past-the-post” system, seats are awarded to the candidate who has received most votes. In most cases, such an electoral system is dependent on single-member constituencies. Electoral outcomes in plurality systems are affected by a phenomenon known as “Duverger’s Law”, introduced by Maurice Duverger1

, a theory “driven by the idea that in the long run rational politicians and voters will realize that it is hopeless to have more than two parties competing at national level”.2

Douglas Rae3 subsequently quantifiably demonstrated this effect by demonstrating that plurality systems tend to support the growth of two major parties4, and tend to produce electoral outcomes with either one of these parties controlling a majority of seats.5 Examples of plurality electoral systems include the United States and the United Kingdom.6

A majority electoral system has many of the same characteristics as the plurality system, both in terms of the manner in which the elections are conducted (using single-member constituencies) as in terms of the electoral outcome. The difference with the plurality system

1

Duverger 1954.

2

McLean & McMillan 2009.

3

Rae 1967.

4

Carey & Hix 2013, p. 821-822.

5

Although there are various counterexamples of this phenomenon, such as Canada.

6

In the United States, electoral rules fall within the domain of the states. Most states use plurality voting, although there are certain exceptions, such as California and Washington, which use a majority system.

(6)

6

lies within the requirement that in order to be awarded a seat in a majority electoral system, a candidate or party must have received at least half of the votes plus one. An example of a system like this is France. Although a subtle difference thus does exist between plurality and majority systems, the difference between the effects on electoral and political consequences is negligible.

Finally, proportional representation is an electoral system which attempts to make the percentage of seats awarded to each party reflect the percentage of votes each party has received. There are different ways of calculating the amount of seats under a proportional system, and the use of vote thresholds mean that this reflection isn’t always entirely proportional. Whereas plurality and majority electoral systems tend to encourage the development of two major parties, proportional representation tends to promote the development of multi-party systems, wherein coalition agreements are necessary for a government to attain a majority of seats.

2.2. Policy outcomes

Thus, in a representative democracy, the electorate chooses a group of politicians to make decisions on their behalf. There is a general agreement that these electoral rules shape the incentives of these politicians to favor certain constituencies and groups of people.7 Research shows that electoral rules may consequently have an influence on policy outcomes.8 However, what remains debatable is which electoral system causes politicians to be most responsive to which groups and constituencies, and what differences in policy outcomes this could cause. Electoral systems do not have the same effect in all cases: the country-specific socio-economic and political context wherein they operate influences their effects.9 However, a number of broad observations can be made when looking at the relationship between electoral systems and policy.

One of the main differences between the policy outcomes in majoritarian and proportional electoral systems is the group of people that are intended to benefit from government policy. The seminal work in this field is the Grossman and Helpman model, which has demonstrated “a protectionist bias in the equilibrium outcome of majoritarian systems”. 10

This model posits that a country consists of a number of electoral districts. Upon forming the government, the

7 Rickard 2012-I, p. 855. 8 Rickard 2012-II, p. 778. 9

Bingham Powell, Ferree & Scheiner 2013, p. 811.

10

(7)

7

party in power will seek to maximize the welfare of its constituents. When the number of electoral districts that are represented by the governing parties increases, the amount of total constituents whose welfare the government seeks to maximize increases as well. In majoritarian systems, the government will inevitably consist of a number of parties that represent a majority of electoral districts. However, in proportional systems, it is more likely that the government will represent more if not all electoral districts relative to majoritarian systems.11 Governments in proportional systems will thus seek to maximize the welfare of a larger amount of constituents, if not all of them.

In relation to the composition of government spending and the degree of redistribution of income and wealth, electoral rules also have predictive value. Proportional elections stimulate government policy which benefits larger groups of the population, such as welfare programs. Majoritarian elections on the other hand give representatives a greater incentive to promote and implement programs that benefit smaller, geographically concentrated groups.12 In addition, governments with coalition support of two or more political parties tend to tax and spend more than single-party governments. An alternative explanation that has been proposed for this last phenomenon centers around the location of coalition governments on the socio-economic political spectrum: countries with a proportional system are ruled relative more by center-left coalitions in relation to majoritarian governments.13

Research shows that electoral systems can influence the “diversity of perspectives included at various stages in the representative process”.14

Research furthermore demonstrates that there is a direct relationship between the number of parties represented in the legislature and the number of issues considered within a political system.15 In a proportional system, smaller parties tend to represent views on the fringes of the political spectrum, and this allows them to introduce policies and ideas in the legislature that would not have been introduced by larger parties in a majoritarian system, because they often view these issues as too risky to address.16 Even in proportional systems, ideas that are introduced by smaller, more extremist parties are often dismissed by larger, more mainstream parties in the center. However, the discussion of such ideas in the legislature leads to their introduction in the public and media discourse. As a

11

Rickard 2012-II, p. 778.

12

Olper & Raimondi 2012, p. 142.

13

Ticchi & Vindigni 2010, p. 32.

14 Orellana 2010, p. 614. 15 Orellana 2010, p. 614-615. 16 Orellana 2010, p. 616.

(8)

8

consequence, proportional systems tend to have a higher degree of ‘policy innovation’ and adopt novel policies quicker. 17

2.3. Tariffs, non-tariff barriers and protectionist bias

The impact of electoral systems on government policy is largely undisputed. Evidence shows that electoral systems may affect various kinds of government policy, from taxation and government spending to trade policy. Conventional wisdom says that in terms of trade policy, majoritarian systems tend to have a protectionist bias. Relative to proportional electoral systems, majoritarian systems thus tend to favor a narrow selection of the electorate at the expense of the welfare of the electorate as a whole. In relation to trade policy, this goal can be achieved by a number of protectionist measures, both tariff-related as well as by imposing non-tariff barriers. The effect of protectionism as a tool to benefit a narrow group at the expense of the economy as a whole is well documented in economic literature.18 In classical liberal economic theory, the economic benefits of free trade are widely recognized. The fewer limitations there are to free trade, the better its benefits can be maximized. Nonetheless, if one is to assess the degree to which majoritarian systems have an impact on protectionist bias, it is necessary to explore the reasons why protectionism is employed, and how. Furthermore, in order to determine the veracity of this claim, it is necessary to define what a protectionist bias in fact entails.

The Encyclopedia for Global Studies defines protectionism as “the desire to insulate from economic harm a producer of goods in one’s own nation who might suffer economically as a result of the actions of others elsewhere”.19

In short, protectionism is a term used to define a broad range of different policies and methods governments use to change the flows of international trade. Such policy instruments may include import or export tariffs, import quotas and government subsidies. The degree to which governments employ such policies is referred to as the ‘protectionist bias’ of its trade policy.20

Tariffs as a means to protect domestic industries are relatively straight forward. An import tariff is essentially a tax imposed on the importation of a good into a country. By imposing such a tariff the imported good becomes more expensive for domestic consumers or producers. Consequently, the competitiveness of such a good is decreased in relation to an

17

Orellana 2010, p. 623.

18

See e.g. Krugman 1993, Krugman 1995, Magee 1976, Rodrik 1998.

19

Juergensmeyer & Anheier 2012, p. 1571.

20

(9)

9

alternative that is domestically produced and not subjected to import tariffs. Three different types of import tariffs can be distinguished: most-favoured nation (MFN) tariffs, preferential tariffs and bound tariffs. The first category concerns the tariff rate imposed on imports from other members of the World Trade Organization (WTO) outside of preferential trade agreements. The second category are tariffs imposed on trade between two states who have entered in a reciprocal preferential trade agreement, in which they agree to give their trading partner lower tariffs than the standard MFN rate. The final category is the maximum tariff levels states have agreed with each other within the framework of the WTO. In industrial states and developed economies, as the states featured in the case studies are, there is often very little difference between MFN and bound tariff rates.21 Export tariffs are much less common and are typically levied against resources or semi-finished products to promote certain domestic industries.

Tariffs, however, are only one possible way to protect national industries from international trade. Examples of non-tariff barriers include protectionist policies such as import quotas or public procurement practices, assistance policies such as domestic subsidies or industry bailouts and non-protectionist policies such as sanitary or phytosanitary regulations. Non-tariff barriers can have an equally restrictive effect on international trade as Non-tariffs, and their use has been increasing in the last decades.22 Non-tariff barriers are harder to quantify than tariffs, due to the inclusion of a number of not easily quantifiable elements in this category, such as onerous regulatory regimes. Some argue that non-tariff barriers and tariffs are not substitutes for each other, but are used in combination. Evidence for that argument can be found in the observation that states with high tariffs also tend to impose a high amount of non-tariff barriers.23 However, on the other hand, there is also evidence that some sorts of non-tariff barriers such as subsidies do act as substitutes for non-tariffs.24

2.4. Sub-conclusion

One of the most consequential differences between the two main electoral systems, majoritarian and proportional, is their impact on the formation of political parties. Whereas in the former, over time a two-party system is very likely to develop, the latter favours multi-party systems. Empirical evidence demonstrates the validity of this effect known as

21

Yalcin, Felbermayr, Steiniger 2017 , p. 27.

22

Ehrlich 2007, p. 589.

23

Ehrlich 2007 , p. 590.

24

(10)

10

Duverger’s Law, and together with other differences between the two systems, it has a profound impact on policy outcomes. Different methods of selecting representatives result in different incentives for these representatives to act once they have been elected: in majoritarian systems, this means a bias towards favouring the representative’s local constituency, and narrower groups of people in general. Protectionism does exactly this: it favours a narrow group at the expense of the country as a whole. Therein lays a crucial observation in an analysis of the relationship between electoral systems and protectionist bias: the electoral system generates different incentives for representatives, which in turn results in different policy outcomes.

(11)

11

3. Electoral systems and protectionism

Different electoral systems offer different kinds of incentives to representatives, and this can have a significant impact on policy outcomes. The question is: how does this difference impact trade policy? Representatives have a wide range of choices when it comes to responding to the incentives the electoral system in which they operate offer them. When it concerns tariff barriers, the incentive to favour narrower groups of the electorate seems a strong argument in support of the theory of protectionist bias in majoritarian systems. But when it concerns non-tariff barriers, this effect does not appear to offer such clear conclusions. In an analysis of the relationship between protectionist bias and electoral systems, the non-tariff barriers are an equally relevant but often not as covered subject area.

3.1. Electoral systems and tariff barriers

As described above, it is evident that the design of electoral systems can have a direct impact on policy outcomes in general. The mechanisms that are at play in this relation also function in the specific field of trade policy. In relation to electoral systems and trade policy, the Grossman and Helpman model is of relevance. Majoritarian groups tend to focus on policies benefitting a narrower group of constituents. As demonstrated, protectionist policies are an effective tool to maximize welfare of a limited (either geographically or socio-economically) number of constituents or industries, but are detrimental to the economy as a whole.25 Because of this effect, the model theorizes that majoritarian systems will often have a ‘protectionist bias’ relative to proportional systems.26

Grossman & Helpman’s model is consistent with other scholar’s findings on the representation of broader versus narrower interests in different electoral systems, and furthermore supported by other research on the relationship between electoral systems and trade policy.27

There is a broad consensus in the literature on the relationship between majoritarian systems and higher tariff levels. The literature offers a number of explanations for this relationship, and identifies various significant factors. Electoral incentives offer various explanations of this mechanism. For example, research shows that the term length of representatives is indicative of their tendency to support protectionist policies.28 A short political horizon, which includes the length of their term and how much time they have left until they again face

25

See Chapter 2.3.

26

Grossman & Helpman 2005, p. 1266.

27

E.g. Evans 2009, Willman 2005, Hatfield & Hauk 2014.

28

(12)

12

reelection, leads representatives to support protectionist stances in the form of import tariffs.29 Another example of electoral incentives influencing trade policy focuses on the election campaign itself. In electoral systems with a high degree of geographic concentration of a particular product in a given electoral district, political parties and representatives often use tariffs as means of maximizing their chance of being (re)elected into office.30 Incumbent representatives can use their voting record to indicate their willingness to employ tariffs to protect local industries, whereas political parties can signal their support by announcing protectionist trade policies. By doing so, they can induce swing voters with a dependency on the economic performance of an industry that they are dependent on to support them.31 In an electoral system that divides a country into many smaller districts, as majoritarian systems tend to do, there is a higher chance of industries being geographically concentrated in an electoral system. Consequently, majoritarian systems encourage the use of the electoral incentives that supporting protectionism offers during election campaigns.

Another example a factor that plays a role is country size: country size seems to be positively associated with higher levels of trade protection by means of tariffs.32 The number of electoral districts in a country also affects trade policy. In proportional systems, a country is divided into one or more larger multi-candidate electoral districts, whereas in majoritarian systems, a country is divided into much more and smaller electoral districts, each represented by one representative. Empirical evidence shows that in majoritarian elections, each individual district elects a representative that tends to be more protectionist than the median voter’s preference.33 Consequently, a positive relationship exists between the number of electoral districts in a country and its tariff rate.34

However, the literature also contains examples of studies that do not find a direct relationship between the amount of constituencies or the constituency size and tariff levels.35 Smaller constituencies tend to incentivize representatives to support distributive policies, but a conception of protectionist trade policies as a means to achieve distributive effects is misguided.36 Data on trade policy preferences of representatives with larger constituency sizes

29

Conconi, Facchini & Zanardi 2014, p. 104.

30

Roelfsma 2004, p. 3.

31

Roelfsma 2004, p. 1-2.

32

Hatfield & Hauk 2014, p. 529.

33

Evans 2009, p. 280.

34

Willman 2004, p. 20.

35

Hatfield & Hauk 2014, p. 519.

36

(13)

13

demonstrates that constituency size is generally not a predictive indicator of voting for protectionist policies.37

Besides a relatively larger focus on the welfare of a narrower group of constituents, representatives in majoritarian systems also tend to attach additional weight to the welfare of the constituents of their home districts.38 The geographic concentration of certain industries plays a significant role in this. If an industry is relatively concentrated in districts controlled by the majority, it has a higher chance of receiving positive protection in the form of subsidies, tariffs or non-tariff barriers, even if corrected for the effect of organized lobbying by that industry.39 This is further demonstrated by empirical evidence showing that greater geographic concentration is linked to higher average tariff levels.40 An explanation for this phenomenon could include the fact that in constituencies with a high concentration of a specific kind of industry, trade protection in the form of tariffs offers an attractive way for the representative of that constituency to benefit their local constituents.41 Local industry profits from higher tariff levels in the form of higher prices for their goods, and in most cases, the costs associated with these higher prices for the local consumer is comparatively small.42 In proportional systems, the proposition of increased tariffs to benefit local industries is much less attractive for an individual politician, as both the reward of increased political support as well as the increased cost associated with rising product prices are spread out.43

3.2. Electoral systems and non-tariff barriers

Besides imposing import or export tariffs, there are other ways states attempt to shield domestic industry: by imposing protectionist measures beyond the scope of tariffs, such as non-tariff barriers. Most studies on the relationship between electoral rules and protectionism do not address this in depth, and thereby leave open the possibility that electoral rules do not have an effect on the overall level of protectionism. For example, it might simply be true that non-tariff barriers are used more often in majoritarian systems. In analyzing the relationship between electoral systems and protectionism, the issue of whether or not protectionist bias extends beyond tariff measures is thus a crucial one.

37

Karol 2007, p. 492, Conconi, Facchini & Zanardi 2014, p. 105.

38

Fredrikson, Matschke & Minier 2011, p. 2.

39

Fredrikson, Matschke & Minier 2011, p. 3.

40

Ardelean & Evans 2013, p. 725.

41

Rickard 2012-II, p. 859.

42

Ardelean & Evans 2013, p. 764.

43

(14)

14

Non-tariff measures in the form of regulations are one example of protectionist policies beyond the scope of tariffs. Non-tariff measures are regulations on products usually imposed to protect consumer and the environment. However, a side effect of non-tariff measures is often that they increase the trading costs and thus act as a barrier to trade.44 Within this category, a distinction can be made between rules imposed on import and export of goods, and internal rules imposed on the domestic economy. Over the past decade, the amount of non-tariff measures that form a de facto barrier to trade has been steadily increasing.45 The relationship between electoral systems and the level of regulatory barriers to trade is however a relatively understudied field. A clear relationship between the level of regulatory measures to protect consumers and electoral systems has been identified, however. There is a significant pro-consumer bias in majoritarian systems, whereas proportional systems tend to favor producers more – but this observation alone is not enough to make the claim that majoritarian systems tend to produce regulatory policies that pose a barrier to trade any more than proportional systems.46

Another example of non-tariff protectionism includes assistance policies such as subsidies or industry bailouts. Subsidizing domestic industries is common throughout industrialized countries, and often employed as a tool to offer protection against the threat of international competition.47 Industries have an incentive to lobby for such protection, out of a desire to maximize their profits. The more an industry stands to lose from increased international competition, for example because of a high degree of asset specificity or relatively low level of productivity, the more likely it is that such an industry will lobby politicians to enact legislation that offers them more protection.48 In an attempt to shield these industries from the pressure of increasing integration into world markets, governments will often be susceptible to these arguments.49 Collective action problems are one barrier these industries will face in their attempts to lobby for increased protection. Political institutions are a determinant for how likely industries are to overcome these barriers, and as such, electoral systems can have an impact on the level of non-tariff protectionism through subsidies. The more majoritarian political institutions are, the likelier it becomes that lobbying efforts are successful, since national representatives divided in narrow, geographically organized constituencies have a

44 Olper 2016, p. 288. 45 Olper 2016, p. 290. 46 Weinberg 2012, p. 824. 47 Zahariadis 2001, p. 603. 48 Zahariadis 2001, p. 606. 49 Aydin 2007, p. 116.

(15)

15

larger incentive to offer protection to regionally organized industries.50 This effect is further amplified during pelection periods: when politicians are faced with an upcoming re-election campaign, they will be more responsive to the demands of industries to receive aid to help them adjust to international integration.51 In majoritarian systems, those elections are being held under an electoral system which allows the electorate to directly link its political support to their representative’s willingness to offer protection to industries located in their own constituency. This mechanism functions to further pressure politicians in majoritarian systems to subsidize local industries, whereas in proportional systems, this pressure would be more indirect.52

Contrarian views are also represented in the literature. For example, a study of cross-national patterns of non-tariff barriers between 1983 and 1986 by Mansfield and Busch found that countries with proportional systems often tend to have higher non-tariff barriers.53 However, they acknowledge that it is difficult to conclude that the electoral system plays a role in this difference: their observation that larger economies generally tend to have a higher incidence of non-tariff barriers is an indication of the difficulty of drawing a conclusion that proportional systems promote non-tariff barriers.54

3.3. Sub-conclusion

Representatives decide on their preferred trade policy based on a number of different incentives. Electoral systems function to shape these incentives by changing the circumstances in which representatives operate. Different variables may encourage or discourage protectionist policies, but variables that encourage protectionism are more often found in majoritarian systems. Examples include electoral incentives such as the term length of representatives, the competitiveness of their electoral campaigns, the number of electoral districts and the level of geographic concentration of industries. These variables influence both the likelihood of tariff barriers as well as non-tariff barriers imposed by a country.

50 Zahariadis 2001, p. 607. 51 Aydin 2007, p. 119. 52 Aydin 2007, p. 119-120. 53

Mansfield & Busch 1994, p. 744.

54

(16)

16

4. Actors and processes in trade policy-making in majoritarian systems

Electoral systems are responsible for the circumstances in which representatives operate, and what incentives representatives have in making trade policy. Majoritarian systems tend to create two-party systems, wherein both parties have an incentive to support narrower groups of the electorate. When these incentives are translated into policy making, it becomes apparent how the different characteristics of majoritarian and proportional electoral systems award greater rewards to representatives supporting protectionist trade policy. In order to formulate an answer to the question of protectionist bias in majoritarian systems, it is necessary to analyze how the trade policy-making process takes place, and how different actors influence this process. Finally, I will formalize this analysis in a theoretical model of trade policy-making against the background of different electoral systems, to benefit the case study.

4.1. Approaches to trade policy-making

Protectionist policies arise from a desire to shield domestic industries and protect national economic interests from the threats of greater integration of the economy into the world market. From this perspective, trade policy can often be viewed as responding to economic changes. For example, imagine an industry in a country that is struggling because its relatively low productivity or high producing costs has made it vulnerable to increased integration into the world market. This vulnerability will translate into a policy preference to see this industry protected from increased integration, for example by imposing import tariffs or subsidizing it. Such a preference will be expressed by interest groups representing this industry or others connected to it, as well as by voters with an economic dependency to this industry. Interest groups have an active role in expressing this preference, by attempting to influence the outcome in their favor through engagement with parties and representatives.55 The electorate will be less pro-active, with political parties and representatives signaling their recognition of the electorate’s trade policy preferences through votes and proposed policies.56

Depending on the constitutional and electoral framework in which the country operates, this preference will translate in certain political incentives for national representatives to translate this preference into a certain political outcome. This outcome, in turn, will affect the country’s trade policy, which will invoke a market response, which will in turn have a certain

55

Epstein, Mealem & Nitzan 2012, p. 78.

56

(17)

17

economic outcome, which will translate into changing preferences of the electorate and interest groups.57 Olper,58 in adaption from Persson and Tabellini,59 has translated this process into a stylized model of trade policy making in democracies, which is depicted in the following figure:

Demand

Side ↗

Preferences of electorate & interest

groups ↘ Economic Outcome Constitutional & electoral rules ↑ ↓ Market Response Political Outcome ↖ Trade Policy Outcome ↙ Supply Side

The main focus of this chapter lies in the translation of electorate and interest group preferences to trade policy outcome and the manner in which this outcome is shaped by electoral rules. This translation is the policy-making process and is shaped by various actors and influencing factors. Several actors play a role in this policy-making process. Firstly, there is the general electorate. In representative democracies, with regular elections, representatives are dependent on the goodwill of their electorate in order to be re-elected and continue their tenure. If one assumes that every representative as well as their direct opponent will have a strong desire to be (re)elected, it stands to reason that in this policy-making process, a trade policy that is preferred by the electorate will offer distinct advantages to politicians. The second group consists of those whose interests are not otherwise represented in the electorate, and consequently organize themselves in interest groups. Such interest groups are also able to offer incentives or disincentives. For example, such groups may offer financial or other support to a representative or his or her opponent on the basis of their voting record and policy decisions.60 57 Olper 2017, p. 295-296. 58 Olper 2016, p. 296. 59

Persson & Tabellini 2003.

60

(18)

18

Representatives are therefore continuously balancing the interests of different groups in order to continue to receive their support. In literature, two distinct approaches seeking to explain this balancing act can be identified.61 The first approach stresses the importance of political competition between candidates and parties. Opposing parties will propose different policies if they are elected, and interest groups will assess the impact of those alternate policies on the welfare of their members and the interests they seek to advance, in order to determine which party or side offers the best prospects and should therefore receive their support. Through this support, parties are able to reach more voters in an attempt to gain their support. In this approach, interest groups will weigh the cost of their support against the likelihood of their support helping their preferred candidate or party win the election and thereafter enacting their preferred policies. In this approach to the explaining the balancing act of representatives, interest groups and their preferences are the most important consideration for representatives, as their backing enables them to gain more support from the first group, the electorate.62 The second approach views policy-making as an instrument of the incumbent government, and the representatives supporting it, to maximize their political support. In this approach, “the welfare that designated interest groups derive from the chosen policies and the deadweight loss that the policies impose on the society at large”63

is the principal consideration in determining government policy. In this approach, the incentives of interest groups, by means of financial or other support, are viewed as only indirectly affecting policy choices.

Besides the incentives that the electorate and interest groups offer, the policy-making process of representatives is also influenced by the political party to which they belong, and the ideological position of this party. This might appear to be an obvious factor, but it is nonetheless relevant to mention the relationship between electoral systems and the characteristics and ideological positions of their parties. An example of the influence of the majoritarian electoral system on the prevailing ideologies represented in the policy-making process is the number of parties active within an electoral system. A known consequence of Duverger’s law is that relative to a majoritarian system, in a proportional system more parties

61

62

63

(19)

19

tend to contest the seats in the legislature and a broader variety of parties are represented in the legislature.64

4.2. Actors and variables influencing trade policy

Within Olper’s model of trade policy, two distinct groups that affect the process can be distinguished: the electorate and interest groups. Interest groups will actively try to shape trade policy in their favour, whereas the electorate tends to be much less proactive. During the policy-making process, representatives will weigh the interest group preferences and those of the electorate, as well as taking into account the prevailing ideology of their party and the degree of party discipline.65

Regarding the first group, the impact of the electorate on trade policy-making is most significant when trade liberalization is causing negative economic effects. Free trade and the integration of domestic industries into the world market can cause harm for domestic workers through lower wages or job losses associated with more international competition. However, in order for this factor to become a relevant consideration for representatives in making trade policy, these changing circumstances must be an electorally salient issue. Where the electorate’s views on free trade feed into aversion against further trade liberalization, representatives face a strong incentive to argue in favour of imposing new trade barriers during the decision-making process.66 This is even more so the case considering that the most significant electoral consequence of free trade is that job losses as a result of free trade have a distinctive negative effect on the electorate’s support for the incumbent representative and decreases their chances of re-election.67 Noticeably, this negative effect is significantly larger for trade-related job losses relative to job losses due to other circumstances.68 This effect is even stronger in electorally competitive districts, whereas in districts that are generally considered ‘safe’ for either party, the effect is weaker. Consequently, the opportunities that majoritarian systems provide for regional differences in the economic effects of free trade to be translated into incumbent representatives losing their seats mean that the influence of the electorate on the trade policy-making process must be regarded as significant.69

64

Orellana 2010, p. 614.

65

66

Jäkel & Smolka 2017, p. 16.

67 Margalit 2011, p. 166. 68 Margalit 2011, p. 167. 69 Margalit 2011, p. 167-168.

(20)

20

Some opposing scholars hold the view that the causal relationship between the preferences of the electorate and trade policy-making is not in fact so clear. Several studies have attempted to formulate an answer to the question whether it is the electorate that impacts trade policy by expressing its policy preferences through elections. There is evidence present in the literature that seems to suggest that a party’s platform on trade policy in fact affects the electoral outcome, which would mean that the causal link between electoral preferences and electoral outcome is in fact the other way round.70 However, though at first instance this might seem illogical, it is in fact not inconsistent with Olper’s theory of trade policy-making as is a vicious cycle, constantly responding to changing circumstances. Protectionism will clearly cause economic loss with some parts of the electorate while causing gains with others. Empirical evidence supports this theory of a responsive mechanism in trade policy making.71

However, short of cases wherein voters are directly faced with economic loss due to free trade, trade policy is not always an electorally salient issue. For many voters, trade policy is not amongst the most important issues in determining their vote, and many voters in majoritarian systems do not follow their representative’s trade policy positions directly.72

Instead, the electorate draws conclusions on whether the trade policy of their representative is aligned with their views using other cues, such as the party affiliation of their representative and media statements.73 Additional factors that affect the degree to which voters can influence the policy-making process by holding their representatives accountable are divided government and a growing number of voters who identify as independent.74

Where trade policy is not salient enough for the electorate to strongly incentivize representatives to follow their preferred trade policy, interest groups will fill that void. Interest groups, representing domestic industries with an economic interest to see their trade preferences met, will weigh the costs and benefits of supporting a candidate and act accordingly.75 In order to incentivize representatives to follow their preferred policies, interest groups make use of two distinct mechanisms: selection and influence. The first must be interpreted literally: by influencing the recruitment, nomination and election of candidates,

70 Yu 2009, p. 552. 71 Yu 2009, p. 553. 72 Guisinger 2009, p. 545. 73 Guisinger 2009, p. 546. 74 Guisinger 2009, p. 554.. 75

(21)

21

interest groups can affect the trade policy-making process by introducing representatives into the process that share their policy positions into the process.76 Majoritarian systems allow for more possibilities for interest groups to play a role in this selection process: since majoritarian elections tend to develop two-party systems, they are characterized by many one-against-one election campaigns that increase the interest groups’ leverage. Besides selection, interest groups can also attempt to shape the policy-making process by influencing incumbent representative, through lobbying or bargaining.77

One of the most obvious ways in which interest groups can influence the policy-making process is through financial means. Interest groups can offer political contributions to representatives, which hold great value to representatives as these contributions can be used to increase their chances of re-election.78 Often, these contributions are part of an implicit quid pro quo: while direct concessions in return for political contributions are often illegal, interest groups still often implicitly relate prospective contributions to the trade policies chosen by representatives.79 The more discretion an electoral system allows for political contributions, the more influence interest groups will gain.

A determinant of the likelihood of interest groups and voters seeing their trade policy preferences translated into actual policy is the number of ‘access points’ an electoral system offers. In other words, the more susceptible representatives there are in an electoral system, the more opportunities there are to influence the policy-making process. More access points make lobbying cheaper, as there is more competition between representatives for the benefits that interest groups can provide.80 Majoritarian systems tend to have a higher number of access points, as a consequence of having more electoral districts and less party discipline.81 However, because interest groups have a collective action advantage relative to the electorate, they will often benefit more from more access points. This tips the balance between weighing interest group preferences and the preferences of the electorate even more so in favour of the first during the trade policy-making process.

76

Fordham & McKeown 2003, p. 522.

77

Fordham & McKeown 2003, p. 523.

78

79

80

Ehrlich 2007, p. 576.

81

(22)

22

Even when both the electoral as well as the interest group incentives are taken into account, many analyses of the trade policy making process hinge on the assumption that representatives in majoritarian systems decide upon the trade policy that offers them the most benefits in unity. In other words, many theoretical models assume complete party discipline. In reality, majoritarian systems are often characterized by lower party discipline.82 This is partly caused by the fact that majoritarian systems, by dividing the country into many electoral districts, incentivize representatives to cultivate a personal vote with their electorate.83 As a consequence of this incentive, party leaders in majoritarian systems tend to have fewer opportunities to control access of their members to media and the electorate, which leads to lower party discipline.84 A consequence of lower party discipline in majoritarian systems is that trade policy making requires a bargaining process within the majority party. This bargaining process impacts the policy-making process because representatives will seek to avoid it altogether in certain situations by delegating their control over the setting of tariffs to another authority.85

4.3. Theoretical model and sub-conclusion

Building on the observations of the chapters above, we can develop a theoretical model which captures all of the identifiable variables that affect trade policy. By using the Olper model of trade policy-making in representative democracies, we can establish that the catalyst for formulating trade policy is a change in economic circumstances. These circumstances will in some cases affect the preferences of the two main actors that influence trade policy: the electorate and interest groups. Whether changing economic circumstances affects these preferences and the degree of salience of this issue for both actors depends on factors that are different for each group. In the translation of these preferences into trade policy, electoral rules play a significant role: majoritarian electoral rules make offering incentives to representatives easier and less costly, thereby increasing the chances of these incentives successfully affecting trade policy.

82 Mcgillivray 1997, p. 584. 83 Pearson 2017, p. 199. 84 Pearson 2017, p. 200. 85

(23)

23

This process can be demonstrated by the following theoretical model:

Trade policy-making in representative democracies

Changing economic circumstances

↓

Changing trade policy preferences of interest groups and electorate, depending on: Variables regarding the electorate Variables regarding interest groups

 Electoral salience of trade policy

 Job loss of the electorate due to free trade

 Potential for influence on representative selection process

 Opportunities for financial aid to representatives

↓

To see preferences met, incentives are offered to representatives (financial and substantive aid or electoral support). The success of these efforts is dependent on variables, often created by

electoral rules:

↙

↓

↘

Variables in majoritarian systems Shared variables Variables in proportional systems

 More electoral districts incentivizes policies

benefitting smaller/narrower groups

 Geographic concentration of industries facilitates interest groups advocating tariffs

 Two-party system and

majority elections give interest groups greater bargaining power

 Representatives are easier to be held accountable for their support for protecting local industries

 More access points lower costs for interest groups and improves their chances of success

Some variables shape trade policy but are not directly related to either electoral system:

 Short tenures

 Country size

 Fewer electoral districts

incentivizes policies benefitting broader groups

 Geographic concentration of industries does not facilitate interest groups advocating tariffs

 Multi-party system and

proportional elections do not give interest groups greater bargaining power

 Representatives cannot be held directly responsible for their support for protecting local industries, only indirectly

 Less access points creates higher lobbying costs and lowers chances of success for interest groups

↓

Higher chance of protectionist bias in trade policy

Lower chance of protectionist bias in trade policy

(24)

24

When analyzing a country’s trade policy, the theoretical model offers a guide which allows all relevant factors to be considered. Firstly, there must be a change in economic circumstances that changes the trade preferences of the two actors in the policy-making process. For both actors, there are variables that influence the degree in which these actors will attempt to see their preferences met. This first set of variables exists regardless of electoral rules and is not influenced by a country’s electoral system. Subsequently, the success of the actors’ efforts to influence the policy-making process in turn depends on a set of variables that is dependent on electoral rules. Almost all of variables dependent on electoral systems that tend to increase protectionist bias exist in majoritarian systems, whereas proportional systems do not have these characteristics.

(25)

25

5: Case Study

5.1. Design and Methodology

In the above chapters, I have defined the characteristics of the main electoral systems, their effect on protectionist bias in trade policy-making and the different actors and variables that influence this policy-making process. This has resulted in the development of a theoretical model of trade policy-making. Using this theoretical model, I will attempt to test the hypothesis that majoritarian systems tend to have a protectionist bias, by looking at the trade policy-making of the United States.

A number of different factors are relevant in justifying this choice for analyzing U.S. trade policy as a case study. Firstly, the United States trade policy, as one of the world’s largest trading countries, is very well documented in literature. The same applies to the electoral system of the United States, as one of the largest and economically most influential countries using a majoritarian system. Secondly, the presence of the variables identified in the theoretical model that incentivize the two relevant actors in the policy-making process to pressure representatives to follow their trade policy preferences is very well documented in the United States. This therefore removes the possibility that because the presence of these factors cannot be established, interferences cannot be easily drawn from the observations in the case study.

Regarding the relevant electorate’s variables, a first relevant observation includes the electoral salience of trade policy in the United States. Discussions on trade policy played a central role in the 2016 U.S. Presidential Election, with the Republican candidate taking an ideological view on trade liberalization that departed from his party’s ideological line, and the Democratic candidate expressing relative support for free trade.86 However, research shows that concerns over rising imports and lost manufacturing jobs were not amongst the main issues of the electorate.87 In fact, during the 2016 presidential election, economic hardship was an element distinctly but weakly predictive of support for the Democratic candidate.88 The theoretical model predicts that if trade policy is not an electorally salient issue, this leaves a void for interest groups to incentivize candidates to follow their preferred trade policy instead. The theoretical model identifies two variables that influence the degree of success of

86

Freud & Sidhu 2017, p. 2.

87

Freud & Sidhu 2017, p. 3.

88

(26)

26

interest groups in this effort: the potential to influence the selection process of representatives, and the opportunities to provide financial and substantive aid. There are relatively few regulations on campaign contributions by interest groups in the United States, which predicts a high chance of success of interest groups in their efforts to steer trade policy.89

In order to test the hypothesis of protectionist bias in majoritarian systems, I will look at two recent political events in the field of trade policy in the United States: the 2002 United States steel tariffs, and the 2015 renewal of trade promotion authority to the executive. I will conduct qualitative analysis of these two cases, together with an analysis of the policy-making process, to enable reaching valid interferences. I will do so by looking at how the policy-making process in these two cases has been influenced by the relevant characteristics that arise from a majoritarian electoral system.

5.2. Protectionism, trade policy-making process & electoral system in the U.S.

According to Article 1, Section 8 of the U.S. Constitution, the U.S. Congress is the primary authority over trade policy. Trade policy in the U.S. is based on statutory authorities, set up by the Congress, which include laws that authorize trade programs and govern tariff and non-tariff barriers. Congress furthermore has the power to ratify international treaties such as trade agreements.90 The Congress is a bicameral body elected using a mixed majority and plurality system, whereby legislators represent a specific area, either a congressional district (for U.S. Representatives) or states (for U.S. Senators). Individual representatives have the opportunity to affect trade policy through their votes on statutes governing trade matters, and can exercise influence by sitting in relevant Congressional committees that deal with trade policy. Furthermore, representatives can influence their colleagues in an informal way.91

The Congress has delegated much of its direct authority over trade policy by giving the executive far-reaching powers in dealing with trade-related issues through a mechanism called ‘trade promotion authority’.92

This authority allows the executive to negotiate trade deals on their behalf. Although the Congress has the constitutional obligation to approve of agreements

89 Ehrlich 2007, p. 581-582. 90 Cimino-Isaacs et al. 2019, p. 26-27. 91 Cimino-Isaacs et al. 2019, p. 28. 92

(27)

27

that liberalize international trade, the trade promotion authority provides for a much faster approval process by “guaranteeing an up-or-down vote with no amendments”.93

Similarly, much of the legislative authority on setting tariffs has also been delegated to the executive. Again, the U.S. constitution stipulates that the Congress is the only body empowered to set tariffs, through various laws, much of this power now falls under the executive.94 However, even though much of the authority over trade policy has been delegated to the executive, the majoritarian nature of the U.S. electoral system is still visible in the political calculations that inform trade policy. In this case study, the two examples also highlight this.

The tariff rates of the United States over the past 9 years have remained relatively stable, with the exception of a distinctive uptick in 2017:95

93 Cimino-Isaacs et al. 2019, p. 28. 94 Cimino-Isaacs et al. 2019, p. 38. 95

Source: World Bank Data on International Trade. 2,6 2,7 2,8 2,9 3 3,1 3,2 3,3 3,4 2009 2010 2011 2012 2013 2014 2015 2016 2017

Unweighted average percentage of effectively applied

rates for all products subject to tariffs

(28)

28

However, non-tariff barriers have seen a significant increase over the same time period. For example, the amount of discriminatory measures in the U.S. has been steadily rising:96

Although the United States is consequently not a very protectionist country, it has in the past enacted protectionist measures at various stages, in part fueled by the majoritarian electoral system that selects the trade policy-makers.

5.4. 2002 United States steel tariff

The United States presidential elections are conducted in a framework that is characterized by a unique framework of majoritarian and plurality electoral systems. In order to win an election, a candidate must attain an absolute majority of the 538 electoral votes, who are awarded per state. Yet, a plurality win is enough to be awarded all of the electoral votes of a state. This variant of a majoritarian system allows presidential candidates to push for policies favoring states that are generally competitive, at the expense of the electorate as a whole. Since the popular vote is not a relevant factor in attempting to win the presidential elections in this system, the costs associated with such a strategy are relatively small, whereas the potential gains of such a strategy, winning the most competitive states, are large.

Consequently, two variables of the theoretical model are at play here. Firstly, this system of electing presidents entails that the country is divided into many different electoral districts

96

Source: Global Trade Alert Data. 0 200 400 600 800 1000 1200 1400 2009 2010 2011 2012 2013 2014 2015 2016 2017

(29)

29

(the fifty states and the District of Columbia). Secondly, geographic concentration of industries in these electoral districts allows candidates to advocate policies that benefit the electorate in the most competitive states, thereby increasing his or her chances of winning the election. Following the theoretical model, this incentive will be even more pronounced if these policies are enacted in response to changing economic circumstances, for example job losses. Given these circumstances, the theoretical model would predict that presidential candidates, seeking to maximize their chances of election, will make use of these majoritarian characteristics by proposing protectionist policies to benefit a narrower group of the electorate at the expense of the entire electorate.

The first example of the case study provides a clear example of this mechanism in effect. The United States presidential election of 2000 was exceptionally close. George W. Bush did not manage to win the popular vote, but was elected nonetheless as he managed to win 50.38% of the electoral votes, to his opponent’s 49.63%. His margin of victory in several industrialized states in the Midwestern area of the U.S. was very small, and in order to maximize his chances of being re-elected, Bush used the characteristics that the majoritarian system of electing U.S. presidents provided him to increase his chances of winning the next election. The states that produced the most steel at the time, Indiana, Ohio, Pennsylvania, Michigan and West Virginia, almost all awarded their electoral votes by margins under 6%.97

During this period, the steel industry in the United States was struggling due to increased international competition. In the period leading up to 2002, over 150,000 people who were employed in the steel industry in the United States lost their jobs or benefits.98 Recognizing the potential to capitalize on this electorally salient issue, Bush turned to protectionist policies to gain support of the electorate in these crucial states. So, in March of 2002, Bush imposed an average 30 percent tariff on foreign steel entering the United States.99 The economic impact of this decision was negative, as it caused job losses and throughout the U.S. economy.100 Yet, the electoral benefits of this decision was potentially significant: both during the midterm elections that were held later that year, as well as during the 2004 presidential elections, districts with a high concentration of steel mills were critical to his electoral prospects and those of his party.101 This example highlights how the majoritarian electoral

97 Kelly 2003, p. 1. 98 Ho 2003, p. 830. 99 Ho 2003, p. 826. 100 Ho 2003, p. 827. 101 Ho 2003, p. 832.

(30)

30

system of the U.S. can create incentives for representatives to enact protectionist policies, especially when several of the variables included in the theoretical model (electoral salience of trade policy, geographic concentration of industry, many electoral districts) are present.

5.5. 2015 Renewal of Trade Promotion Authority

As explained, trade promotion authority refers to the process by which Congress enables the executive to negotiate agreements that aim to liberalize trade. Trade promotion authority provides for an expedited legislative procedure to approve trade agreements, without the possibility to amend the agreement, and thereby enables the negotiators to know which concessions they are able to give during the negotiating process.102 This authority is always granted for a limited time period, and consequently needs to be renewed when the authority expires.103 Such an instance occurred in 2015, while then-President Obama was in the midst of renegotiating a comprehensive trade agreement with Asian and Pacific countries, known as the ‘Trans-Pacific Partnership’. Consequently, in deciding upon whether to renew this authority or not, all representatives of the Congress voted on the Trade Promotion Authority Bill of 2015, that was aimed at renewing the trade promotion authority. Due to the nature of the trade promotion authority, and the liberalizing effect it has on international trade, a vote in favor of renewing the trade promotion authority is in essence a vote in favor of more integration into the world market, whereas a vote against can be regarded as a vote in favor of less integration and thereby more protectionism in trade policy.

The ideological predisposition of the different parties in 2015 was largely indicative of how individual representatives voted on the Trade Promotion Authority Bill of 2015. The Democratic Party, with a strong voter base in labor unions that are traditionally opposed to trade liberalization, was largely against the Bill.104 Representatives of the Republican Party on the other hand were largely supportive.105 Thus, accounting for party ideology, the most interesting votes on the Bill are those of Representatives that defied party discipline and went against party ideology: Democratic representatives voting in favor of the Bill, and Republican representatives voting against it.

102

Fergusson & Davis 2019, p. 3.

103

Fergusson & Davis 2019, p. 1.

104

Johnson 2018.

105

Surprisingly, these roles seem to have shifted. Many representatives of the Republican Party now follow president Trump’s critical ideological view of free trade, whereas Democratic representatives seem to have become more critical of protectionist policies. In other words, as the colloquial saying goes, “Partisanship’s a Helluva Drug.”

(31)

31

The theoretical model predicts that since the U.S. uses a majoritarian electoral system, several variables will be predictive of an individual member’s vote on renewing the trade promotion authority. The most important factors amongst these are the geographic concentration of industries and the electoral accountability of individual members. Geographic concentration of industries enables interest groups to lobby representatives more effectively in favor of their preferred trade policy. Research on individual votes on trade bills in the context of renewal of trade promotion authority demonstrate that there is a “pronounced and consistent shift towards protectionism” as a result of changing economic circumstances in electoral districts.106

The theoretical model predicts that representatives of electoral districts that will potentially be hurt by liberalizing trade will prefer protectionist policies at the expense of the entire electorate in order to gain support of a much narrower group of the electorate: the inhabitants of their own district. One of the industries especially vulnerable to trade liberalization is the manufacturing industry. Given the fact that trade liberalization makes it easier for manufacturing jobs to be shifted to other countries with lower wage levels,107 the theoretical model predicts that representatives with a higher concentration of jobs in the manufacturing industry would vote against renewing trade promotion authority. An analysis of the concentration of U.S. manufacturing jobs in comparison with the electoral districts represented by a representative who voted against the renewal Bill shows that in districts with a higher percentage of manufacturing jobs, more representatives voted against the renewal bill relative to representatives of districts with a lower percentage of manufacturing jobs.108 This demonstrates a crucial element of the theoretical model and the influence of majoritarian electoral systems on trade policy: rather than voting to increase the welfare of all of the electorate, the majoritarian system incentivizes representatives to opt for increasing the welfare of their own electorate instead.

The electoral accountability of individual representatives is a second important factor in predicting a preference for protectionist policies of individual representatives. The majoritarian nature of the U.S. electoral system allows for both interest groups and the electorate to directly hold their representatives accountable for decisions that go against their preferred trade policy and in the case of the renewal of the trade promotion authority of 2015,

106

Feigenbaum & Hall 2015, p. 4.

107

Cunningham 2012, p. 399.

108

Trade Policy Decision-Making in Majoritarian Systems;