Partisanship in fiscal policy : partisan-based differences in fiscal policy between Dutch political parties during crises

(1)

UNIVERSITEIT VAN AMSTERDAM

Partisanship in fiscal policy

Partisan-based differences in fiscal policy between Dutch political parties during crises

Nathalie Schrijver 23-6-2017

Masterthesis Political Science: Political Economy Mentor: Gijs Schumacher

(2)

1 “The partisan fights on the basis of an intense political commitment” (Werner, 2007)

(3)

2 Abstract

Partisan-based differences in the political world influence the policies politicians and their political parties propose. This sentence lies at the base of a discussion lasting for many years. The discussion finds its base in the ‘conventional wisdom’ that left wing parties tend to steer more towards higher deficits than right wing parties do: left wing parties tend to spend more, while right wing parties tend to tax more (Cowart, 1978). No consensus about the outcome has been reached yet. There are researchers that state that there is indeed a partisan-based difference between left and right wing parties and there are researchers that state this is not the case. An element that has been added to this discussion is that the partisan-based differences have diminished over the years because of a convergence between the two different sides in politics. This, however, has also become the subject of debate. This project looks into whether or not there are partisan-based differences in the election manifests of Dutch political parties during the 1970s energy crisis and the global financial crisis. The project also looks into if this presumed difference has diminished throughout the year by comparing the parties during the 1970s energy crisis to themselves during the global financial crisis and see if there is a significant difference. After analyzing proposed fiscal measures in the election manifests through content analysis and checking if the results have a statistical significant difference or not, it has been found is that there are no significant differences in the proposed fiscal measures between Dutch right wing and left wing parties during both of the crises. The results have also shown that there is no significant difference in the fiscal measures proposed in the election manifests of the political parties during the 1970s energy crisis and the global financial crisis. Therefore, it can ben stated that for the Dutch political parties during the 1970s energy crisis and the global financial crisis, there is no partisan-based difference. What has been found as well, is that there is a significant difference between the fiscal policy proposals before/after the crises and during the crisis. Therefore, it seems that crises do influence the fiscal measures proposed by the political parties during the 1970s energy crisis and the global financial crisis.

(4)

3

1. Introduction

Not too long ago, the global financial crisis started its reign of terror and, for most countries, the aftermath was felt in the following couple of years. The crisis hit national economies in the fall of 2008 and sent the world into a tailspin. While this crisis is the most recent one, the world has seen more crises affecting economies worldwide (Dhameja, 2010). In these periods, you see governments stepping in. As Bermeo and Pontusson (2012) state, crises are periods of peril, but also periods of change. Rather than sticking to the “normal” way of being, governments seem to abandon the widely accepted and implemented economic policies – trust in markets, low government deficit and debt, minimal government intervention in the economy – and instead come up with new policies with which they had hoped to help their economy best recover from the economic crisis (Armingeon, 2012). A way to respond to economic crisis is through changes in policies, one of which can be fiscal policy.

Fiscal policies tend to get a great deal of attention, as is shown in research over time - from the effect fiscal policy has on economic growth to the effect of certain factors on fiscal policy itself (Alesina, 1987; Pettersson-Lidbom, 2001; Cusack, 2001; Imbeau, Petry & Lamari, 2001; Tavares, 2004; Armingeon, 2012). As Thomas Cusack (2001) states, difference in policies is the result of a different way of thinking. Difference in ways of thinking in the political arena is often linked to partisanship. Partisanship might also have an influence on fiscal policy (Cusack, 2001). The conventional wisdom is that left wing parties tend to steer more towards higher deficits than right wing parties do (Cowart, 1978). Cusack (2001) takes it a step further and states the left-wing parties follow a countercyclical policy, and right-wing parties follow a procyclical policy. These cycles explain not only behavior during a crisis, but also outside of it. While there is a stated conventional wisdom, this wisdom has been the subject of debate. According to Alvarez, Garrett and Lange (1991) there is a consensus on the effects of partisanship on policy: left- and/or right-wing parties have certain partisan economic objectives that they want to pursue. This consensus, however, appears to be debatable.

Quite a lot of statistical research has been done into whether or not partisanship plays a role in shaping fiscal policy [Alesina, 1987; Pettersson-Lidbom, 2001; Cusack, 2001; Imbeau, Petry & Lamari, 2001; Tavares, 2004; Armingeon, 2012]. As opposed to what Alvarez, Garrett and Lange (1991) state, a general consensus about the outcome of the different research seems to be lacking. Some researchers state that partisanship does affect fiscal policy and therefore is significant, while other researchers states the effect of partisanship on fiscal policy is not significant. To contribute to this ongoing debate, this research project will look into if partisanship influences fiscal policy and especially the influence of this partisanship on fiscal policy after periods of crises. There will be a focus on the effects of partisanship on fiscal policy in one specific country, the Netherlands.

(6)

5 Furthermore, this research project narrows the scope down to two specific events, namely the 1970s energy crisis and the global financial crisis. This leads to the following research question:

“Is there a partisan-based difference in fiscal policy responses between Dutch left and right wing political parties during periods of economic crisis?”

To answer this question the next section will look into the relevant theory and explain the necessary concepts of this theory in the theoretical framework. After this has been done, a number of hypotheses will be set up. These will relate to how partisanship, crises, and fiscal policy are linked to one another. The research will be done through content analysis in which one can look at economic, political, and institutional determinants that shape the fiscal policy choices made by governments or political parties (Dellepiana-Avellaneda, 2010). This content analysis will be performed by looking through the election manifests of the relevant political parties in the Netherlands in the relevant years. With the research, this project hopes to contribute to the lack of consensus as to whether or not partisanship plays a role in fiscal policy.

2. Theoretical Framework

In this theoretical framework the relevant concepts of the theory regarding the research question will be brought to light and further explained. These concepts include partisanship, fiscal policy, the 1970s energy crisis, and the global financial crisis.

2.1 Partisanship

Partisanship is linked to the concept of party identification. According to Franklin and Jackson (1983), party identification is used by individuals to gather more information and get to know the persons within specific parties and the policies these parties (wish to) implement (Franklin & Jackson, 1983). Rather than focusing on how individuals connect to these political parties, however, this research project will look into the political parties itself and how the identity of political parties is expressed through policy. This expression of parties is an important element as they play a significant role in democratic life. As Schattschneider stated back in 1942 “political parties were created by democracy and democracy is unthinkable without them” (Schattschneider, 2009). Their goal is to run the country in a way they see fit. Political parties organize government, run the processes within this government, deal with setbacks, and in general organize the country in the best way they can and the way they see is best (Jarvis, 2014). What can be gathered from the explanation of the party identification concept given by Franklin and Jackson, is that parties adopt a certain identity (Franklin & Jackson, 1983).

This identity does not form out of thin air, instead this party identification is the result of a natural evolution, where like-minded individuals with similar interests (with regards to politics) come

(7)

6 together to organize a party to exert political influence (Jarvis, 2014). Parties, however, do not always see eye to eye on how a country should be run and what policies should be implemented. To get back to the basics of Schattschneider: “the smaller the society, the fewer distinct parties and interest composing it (…), the more frequently will a majority be found of the same party (…) the more easily will they execute their plans (…) (Schattschneider, 2009, p. 17). The smaller the society, the smaller the opposition, the bigger the chance is a party can exert their political influence. However, the democratic society the Western world lives in, is not subject to a small society and therefore a great number of parties exist. This is especially the case in multi-party systems (Garzia, 2012). This brings us to partisanship. Many definitions of what partisanship is are given, however, the definition of the concept used in this research project is the following: partisanship refers to a political commitment: politicians who identify with their party’s political identity and support their party’s policies (Converse & Pierce, 1985).

The fact that there are multiple parties, and that these parties all have their own identity, implies that there can appear a conflict between parties when it comes to policies they want to implement to steer a country into a certain direction. While not all political systems acknowledge that conflict is fostered by political parties, no political system is nonpartisan (Jarvis, 2014). This partisan-based conflict comes forward most clearly in the left/right wing divide between political parties and the policies they want to implement. Cochrane (2011) builds on this political disagreement and states that there are no connections between opinions or policies about any two different issues (Cochrane, 2011). Few notions are as pervasive as the division between left and right wing parties in the political arena. All over the world the placement on the right/left scale is seen as an important issue. This tends to be the base for all other important issues, and thus predict political attitudes and behavior of individuals and political parties (Noël & Thérien, 2008). As Cochrane said, there seems to be a lack of connection between opinions and policies between the left and the right. Hibbs makes a similar statement and says that this stems from ideological differences (Hibbs, 1977).

Noël and Thérien, who state that the left and right wing political parties not only have different views on issues such as globalization, economics and social issues, but they also react differently (Noël & Thérien, 2008). According to Cochrane (2011) four ideological orientations between left and right can be distinguished: equality, free market materialism, religion, and out-group intolerance. These ideological differences have a big effect on how political parties tend to deal with economic, social, and immigration issues. While having different views on matters in itself is not per definition a problem, the problem emerges “when different ideologies push in opposing directions on opinions about the same issue” (Cochrane, 2011). A situation wherein this might happen is a situation of crisis. Situations of economic crises, according to political parties, are situations that require fiscal

(8)

7 adjustment. This fiscal adjustment is done through fiscal policy (Bojar, 2016). Fiscal policy is what this research project focusses on. To come back to the basis of the left/right difference, Hibbs’ explanation will be used. Hibbs states that the disagreement stems from ideological and economic differences, both in goals, and ways to achieve these goals. Thus, because the ideologies of left and right wing parties differ and policies reflect ideology, the political parties end up with different fiscal policies. Thus, there are partisan-based differences (Hibbs, 1977). There are researchers that state that this partisan-based difference is subject to change.

Coverse and Pierce (1985) state that while party identification seems a rather static phenomenon, it is not truly static. The partisanship of political parties can be subject to change throughout the years. This does not mean that a party changes overnight from left to right wing. Rather, it means that while a party at first can be outspoken over how certain policy should be implemented, but that this opinion changes. Left parties then tend to drift more towards right party stances and vice versa (Coverse & Pierce, 1985; Budge, Crewe & Farlie, 2010). This view, however, is also disputed by Budge (1995) who states that political parties are under a form of ideological constraint. This ideological constraint means that parties follow policies that are linked to the ideology of the political parties. These differences between ideologies of political parties are often the opposite of each other. This limits parties from taking over ideologies and policy plans of other political parties. This limitation is what makes it highly unlikely for a convergence between left wing and right wing parties to take place, also in fiscal policy (Budge, 1995).

2.2 Fiscal Policy

Fiscal policy is a way in which the government can influence the economy through the use of government revenue and government expenditure. The goal of fiscal policy is for the government to create a balance between public expenditure and taxes (Easterly & Rebelo, 1993).

According to many economists, fiscal policy is an important determinant when it comes to economic growth. The idea is that the elements of fiscal policy, such as taxation and public investment, can influence the economy in a positive matter and provide a solution in times of stagnation or economic downturn (Easterly & Rebelo, 1993). As such, it is often used in times of economic recessions. While the influence of monetary interventions in Western economies was agreed upon years back, the acceptance of fiscal policy as a useful intervention tool remains a topic of debate. Nevertheless, fiscal policy is used as an intervention tool in times of economic downturns (Cowart, 1978). It is often used as an intervention tool by politicians, which makes it closely interrelated to politics. Fiscal policy is mainly driven by economic conditions, but the composition of that fiscal policy is driven by other factors such as partisanship, timing of elections and the fragmentation of the decision making

(9)

8 process of the government (Mulas-Granados, 2006). As this research project is focused on differences between left and right wing parties when it comes to fiscal policy in periods of crises, timing of elections and fragmentation of the decision making process over the government will be left out of the equation.

2.2.1 Partisanship and fiscal policy

The debate on partisanship and fiscal policy is one where consensus is still not reached. While some researchers state that there is a link between fiscal policy and partisanship, others state that this is not the case. The conventional wisdom is that left wing parties tend to steer more towards higher deficits than right wing parties do: left wing parties tend to spend more, while right wing parties tend to spend less (Cowart, 1978). There are also a number of studies that state, that left-wing parties are more responsive to unemployment, whereas right wing parties react more to inflation (Hibbs, 1997). According to Cusack (2001), the conventional wisdom is a misleading view on how partisanship affects fiscal policy. One of the reasons for this is that the produced evidence to support this view fails to take different ideological hues into account (Cusack, 2001).

Cusack then refers to countercyclical and procyclical policies, stating that the left is more likely to follow countercyclical fiscal policies and the right is more likely to follow procyclical policies (Cusack, 2001). Countercyclical policies are policies that do not follow the cycles of the economy. They pursue stimulating fiscal measures in times of economic downturn and when the economy is in upswing, these policies tend to want to implement fiscal adjustment measures that tone down the economy (Feldstein, 2002). Procyclical is the opposite of countercyclical and is more likely to be followed by right wing parties (Cusack, 2001). Procyclical policy follows the economic cycles. This means that when the economy is in an upswing, right wing parties produce fiscal stimulating measures to enhance this upswing. When the economy is in a downturn, the policy implements fiscal adjustment measures to reduce spending, therefore not stimulating the economy as would be done when a party follow countercyclical fiscal policy in that same time period (Feldstein, 2002; Alesina & Tabellini, 2005). Another difference that the divide between right wing and left wing parties brings is that left wing parties are more likely for left wing parties to propose fiscal policy measures that intervene in the economic situation than it is for right wing parties. The reason for this is that left wing parties tend to want to influence the economy more. Right wing parties on the other hand tend to pull back their hands and let the economy run its course by relying on market mechanisms (Blais, Blake & Dion, 1993; Kelly-Gagnon & Geloso, 2013).

The fiscal responses that are embedded in the procyclical or countercyclical policies are responses of fiscal stimulus and fiscal adjustment. Fiscal stimulus policy is a way for the government to influence

(10)

9 the economy. It can be a response to an economic event, such as a recession, to revive that economy or it can be a measure that tries to prevent economic disasters from happening. Fiscal stimulus policy can contain a number of measures whose main goal it is to stimulate the economy. These measures include an increase in economic spending, increasing budget deficits; an increase in public spending or state investments; the reduction of taxes, creating a bigger spending capacity for the people; and a high number of incentives for people and companies to spend and invest. Ideally, for the government implementing them, all of these measures result in economic growth (Alesina & Ardagna, 2010). Besides fiscal stimulus there is fiscal adjustment. Fiscal adjustment policy includes measures that can influence the economy. By decreasing government spending, governmental influence on the economy goes down. Mostly, fiscal adjustment measures are measures that have a reduction of the government budget deficit as a goal. These measures include economic orthodoxy; a decrease in public spending or state investments; possibly, an increase in taxes, which is extra income for the government to reduce the budget deficit; and pushing out fewer incentives for people to spend and invest. For both fiscal stimulus and fiscal adjustment several combinations of the measures can be proposed (Alesina & Ardagna, 2013).

What needs to be taken into account as well, before stating that left or right wing parties spend more or less, is the context in which the fiscal policy is implemented (Bojar, 2016). The fact that context needs to be taken into account and policy is subject to change is in line with the above mentioned statement that partisanship is not as static as one might think. As Bojar (2016) states, political parties are perceived to be located somewhere on a scale that ranges from left to right, in the way they think about certain topics and also when it comes to policies. However, that does not necessarily mean that the way they are “supposed” to react, is also the way they actually react (Bojar, 2016). Taken context into account, a lot can influence the stance of a right or left wing party on specific policies. These contexts do not only consist of crises, but also other circumstances such as globalization, or influential economic theories - such as the Keynesian economic system and the neoliberal economic system.

2.2.2 Neoliberal economics versus economics in fiscal policy

As mentioned in the introduction, this research project will look at the 1970s energy crisis and the global financial crisis, both will be elaborately discussed later on. However, a fact that will be mentioned here is the time period of both crises. The energy crisis took place from 1973 until 1979 (Barsky & Killian, 2004) and the Great Recession commenced in 2007 and lasted up until at least 2012, with the exact end date not yet known (Inotai, 2013). While both the 1970s crisis and the most recent global financial crisis are, in itself, elements that influence fiscal policy (Cowart, 1978), the economic context should also be taken into account (Cusack, 2001). The period from World War II

(11)

10 onwards, up until the 1980s was known as a period in which the Keynesian economics theories were being followed. The period after the Keynesian economics neoliberal economic theories were the main theme (Gordon, 1990). Both the Keynesian and the Neoliberal theories have different viewpoints that can influence the way in which parties form their policies. The fact that this economic context can affect the way policies are formed is also confirmed by Cusack in his research about the effects of partisanship (Cusack, 2001).

Keynes’ theory describes how economies are influenced by demands. It is a theory of the total spending in an economy, which is known as the aggregate demand. Aggregate demand comes forward in the economy’s output of goods, which according to Keynes can be summed up into four components. These components are consumption, investment, government purchases, and net exports and these components can be influenced by the government (Jahan, Mahmud & Papageorgiou, 2004). Changes in the demand have influences on employment and on output. When there are periods wherein demand is inadequate, high unemployment is the most likely result, as well as changes in output (Blinder, 1987). A perfect example of periods where demands change is a period of economic crisis. During a crisis, or an economic downturn, consumers tend to spend less, which makes the demand go down. This reduced spending in turn has an effect on the investments that are made, causing investments to go down as well. According to Keynesian economics this process requires intervention of the state. The state needs to intervene with policies to moderate upswings and downturns in the economy (Blinder, 1987; Jahan et.al. 2004; Farrell & Quiggin, 2012). This theory, according to Cusack (2001), suits left wing parties best, as it entails more spending and more government involvement.

After the Keynesian economics theory, neoliberal economics became influential. Where Keynes states that state intervention is of importance, neoliberalism has a laissez faire approach. This means that state intervention is kept to a minimum and parties in the economic system are freer to do what they deem best (Palley, 2004). Neoliberal economics entails the economic and social arrangements within a society that emphasize market relations and individuality, rather than the role of the government. This makes it quite the opposite of Keynesian economics. With regards to policy the following is being said about neoliberalism: rather than having the state at the base of intervention through policy (both in periods of upswing and downturn), it is the individual, ‘an autonomous, individualized, self-directing decision-making agent’ that stands at the heart of policy-making (Springer, Brich & MacLeavy, 2016). The main focus point in this economic theory is that there is less state intervention in economic organization and more individual intervention or decision making processes. The Keynesian idea of policy making to manage demand should be disregarded and instead transparent rules that take the discretion out of policy decisions should be implemented

(12)

11 (Palley, 2004). According to Cusack (2001), this laissez-faire approach suits right wing parties the most.

Why is it important to know this difference? There seems to be a general consensus in scientific research that ideas of academic economists play an important role in determining what policy will be implemented. The link between academic economic debates and public policies is strong. This means that the general accepted theory of a specific time period, can play a role in the ideas of politicians about what policies should be implemented (Farrell & Quiggin, 2012).

2.4 Crises

As stated earlier on, fiscal adjustment and changes in fiscal policy happen in times of crises. Two examples are the 1970s energy crisis and the global financial crisis. While both crises have different causes, they had similar macroeconomic consequences that affected different countries worldwide. Because both of the crises had such similar and such intense macroeconomic effects, they will be used in the analyses of this research project. What the crises exactly entailed and what the consequences were, will be explained in this section.

2.4.1 The 1970s energy crisis

The 1970s were a period of dependence on imported oil. Unexpected disruptions in this global oil market disrupted the macroeconomic performance in the United States (US). This had a major impact on the worldwide macroeconomic performance. While the period is known as the 1970s energy crisis, it is not one crisis, but it actually consists of two different crises. One is the oil crisis of 1973 and the other one is the energy crisis of 1979 (Barsky & Kilian, 2004). Both the 1973 oil crisis and the 1979 energy crisis are known to have started due to political conflict between the US and the oil providing countries. The 1973 oil crisis was caused by the involvement of the US in the Yom Kippur war in Israel, causing the Organization of Arab Petroleum Exporting Countries (OAPEC) to place an oil embargo against the US, the United Kingdom, Japan, Canada and the Netherlands. The political conflict that led to the 1979 energy crisis was the Iranian revolution (Lieber, 1979; Williams & Alhajji, 2003).

Looking into how the crises started, political conflict was not the only problem. The 1973 crisis is partly consequence of the US shortage of oil due to a domestic oil production peak in 1971 (Runyon & Rocks, 1973) and partly due to oil policies implemented by the US government earlier on to stop the country from importing cheap oil from abroad, rather than using the domestic oil production (Lifset, 2014). The implemented oil policies helped to make sure that domestic oil producers would not suffer from cheaply imported oil. However, the problem came about when an oil shortage started to doom (Lifset, 2014). The period from 1971 until 1973 was a period of intensive growth in

(13)

12 the US that had led to a surge in oil usage throughout the country. The oil production had doubled up until then, but the oil consumption had tripled. This left the US with a shortage in oil. Not the shortage of oil, but the ever-growing consumption, led to a rise in import of oil. This rise of import came at the same time as the American involvement in the Yom Kippur war, causing the OPAEC cartel to raise its oil prices successfully and place an oil embargo not much later (Savage, 1976; Barsky & Kilian, 2004; Lifset, 2014).

The 1979 energy crisis of the 1970s energy crisis knows a similar start. The crisis commenced in 1979 and was the consequence of a decrease in oil output due to the Iranian Revolution. This decrease in oil output caused the OPEC countries to increase the price of oil. Later on, when the Iranian revolution had calmed down and the government had seized control of the oil exports, the problems were not at all resolved (Lieber, 1979). Instead of putting an embargo on oil, the OPEC cartel was simply unwilling to send the amount of petroleum the OECD countries wanted, stating that they are not willing to cover the international demand for petroleum because there is not enough oil. The consequence of this was a worldwide spread of panic over the (claimed) oil shortage. This led to lower amounts of oil export, but also an increase in the price due to the scarcity. This shocked the world economy once again (Lieber, 1979; Tugwell, 1988; Williams & Alhajji, 2003).

The consequences of the crises were severe. The oil crisis of 1973 brought a new global recession from 1973 until 1975, the most severe economic blow since the Great Depression in the 1930s. While the US economy suffered the worst, the Western world suffered from stagflation, high unemployment rates, balance of payments disequilibrium, and high inflation (Perron, 1988). The combination of the oil crisis and the oil production cutback that came with that had slowed down the economic growth of the OECD countries significantly, leading to recessions throughout those countries. This recession was reflected in prices of import and export, both of which were adjusted upwards due to the impact of oil prices in the price levels worldwide. How exactly the crisis caused the several factors of the recession remains a subject of debate (Jump & Wilson, 1975; Barsky & Kilian, 2004). The economic consequences of the 1979 energy crisis were a continuance of the trend that was set in the 1973 crisis, significant price increases in oil and because of that renewed problems with inflation, reduced economic growth, recession, disequilibrium in the balance of payments, and increased unemployment (Lieber, 1979; Nordhaus. 1979).

A message of relevance that Lifset (2014) gives in his research is that an understanding of crises allows us to see how policies that are adopted have worked out to influence and shape the years to come. Crises influence all sorts of policies and these policies in turn influence how the future is shaped. A more precise understanding of why and how the policies are shaped and implemented will

(14)

13 allow people to better understand the context of the crisis and train way of thought of the policy makers as to why they choose for a specific policy plan (Lifset, 2014). This message is not only of relevance for the energy crisis of the 1970s, but can also be taken into account in other crises, such as the global financial crisis.

2.4.2 The Great Recession

The global financial crisis, also known as the Great Recession, commenced in 2007-2008 and was the result of a combination of events that had happened in the global financial market from 2007 to 2008, eventually resulting in the most severe financial crisis, both in economical and geographical reach, the world had seen since the Great Depression in the 1930s. Major financial institutions collapsed, international trade failed worldwide, unemployment increased immensely and recession occurred in a lot of countries worldwide (Claessens, Dell’Ariccia, Igan & Laeven, 2010; Gourinchas & Kose, 2011). While the crisis commenced in 2007 and the worst of the crisis happened in 2008, the build-up to it had been going on several years already. Claessens et al. (2011) state that the run-up to the crisis was characterized by a number of features that could have easily predicted the financial crash that was about to come: sharp asset price increases, credit booms leading to debt, and the build-up of loans and systematic risk.

The fundamental mistake that led to the financial crisis was the housing bubble, which happened alongside the stock bubble in the 1990s (Baker, 2008). The housing bubble consisted of the belief that the housing prices would never fall, and if they would fall, they would quickly climb back up again. This had caused an uptrend in home prices. This belief was combined with the stock bubble. The stock bubble of the 1990s had led to a boom in the consumption of houses. The increased wealth people had generated from the stocks had led them to buy bigger houses, this in turn contributed to the housing bubble. Because there was the idea that the prices of houses could only go up, people bought their new houses at prices far higher than they would normally have (Baker, 2008; Claessens et. al, 2011; Shiller, 2012). This was not only the case in the US, but also in other countries across the world. The increase in housing prices was accompanied by a fast growth in credit, which was combined with a period of an upswing in economic activity (Claessens et. al, 2011). This process of the housing bubble continued until 2007, when the bubble burst (Shiller, 2012)

The bursting of the bubble had a severe effect on the global economy and led to the global financial crisis. The feature that contributed most to the severity and breath of the crisis in terms of the effects it had worldwide was the increased interconnectedness, both on the national and international level, among financial markets (Claessens et. al, 2011). Worldwide the economic growth rates experienced declines, the real economy got into a downturn, unemployment rose,

(15)

14 budget deficit situations deteriorated for nearly all affected countries, international trade collapsed, and in general the global financial crisis had negatively affected, not only the worldwide integration, but also the European integration in the Eurozone due to its financial and macroeconomic consequences (Inotai, 2013; DeBardeleben & Viju, 2013). The situation was made worse because of how households responded to the crisis, by cutting their spending. The immediate response of nearly all governments was to introduce a fiscal stimulus package, with which they hoped to counteract the consequences the crisis had brought along (Claessens et. al, 2011; Inotai, 2013). However, most of the governmental fiscal plans to fight against the consequences of the crisis were not as effective as they had hoped and were met with political and economic controversy, especially within the European Union (Leblond, 2013).

2.4.3 How did both crises influence the Netherlands in particular?

In general, a large part of the (western) world suffered the consequences of both the energy crisis and the global financial crisis. Next to the general effects of the crises, it is relevant to highlight how the crises have affected the Netherlands in particular as well, since the focus point of this research project is the Netherlands (Claessens et. al, 2011)

The energy crisis of the 1970s influenced the Netherlands in particular because of its linkage to the United States. As stated before, the crisis was partly due to US involvement in a political conflict in Israel, which caused an oil embargo against the US. The OPEC, however, did not only place this embargo against the US but against the Netherlands as well. The consequences were considered as severe. Not only did the Netherlands have to cope with an oil embargo, but like the other countries, at the same time the crisis brought a recession with high unemployment, a balance of payments disequilibrium, and high inflation (Licklider, 1988; Perron, 1988)

The consequences of the Great Recession in the Netherlands are more elaborately covered in scientific research. Like the rest of the countries affected by the Great Recession, the Netherlands experienced a recession. While the crisis began in 2007, the Netherlands did not experience it until the second quarter of 2008 (DeBardeleben & Viju, 2013). Like the other crisis affected countries, the Netherlands experienced declines in growth rates and unemployment. Surprisingly enough, rather than the budget deficits worsening, the Netherlands was one of the countries that improved their budget deficit in 2008 (Claessens et. al, 2011). This, however, all changed in 2009 when the countries with strong budget surplus positions were hit by crisis once again. Especially countries with “prudent policies, such as the Netherlands, suffered. The Dutch government, among other countries, responded with a fiscal stimulus policy, in which they supported businesses, invested in the housing market and tried to stimulate people in spending. This resulted in a budget deficit (Inotai, 2013).

(16)

15

3. Hypotheses

The theoretical framework gives us an insight in the fact that there is a lot of ongoing discussion about the effects of partisanship on fiscal policy and on the change of the partisan effect throughout the years. As mentioned in the introduction, the research question of this project is

“Is there a partisan-based difference in fiscal policy responses between Dutch left and right wing political parties during periods of economic crisis?”

The question has some assumptions in it. Firstly, the question assumes that crises influence fiscal policy. Secondly, it questions if there is a partisan-based difference. Thirdly, by looking into several years it can be assumed that a change might come forth (or no change at all). This project will be looking into all these assumptions.

In order to set up the hypotheses the theoretical framework is used. Firstly, the theory1 states that there are different policy cycles political parties follow when it comes to fiscal policy. There are countercyclical fiscal policy cycles and procyclical fiscal policy cycles. Countercyclical fiscal policies means that fiscal policy does not following the cycles of the economy but does the opposite. This results in fiscal adjustment measures that tone down the economy in times of economic upswing and stimulating measures in times of economic downturn. Procyclical fiscal policy does follow economic cycles, resulting in fiscal stimulating measures in times of economic upswing and fiscal adjustment measures in times of economic downturn. Secondly, the theory2 states that there is a difference between the cycles left wing political parties tend to follow and the cycles right wing political parties tend to follow. Left wing parties follow countercyclical policies, while right wing parties tend to follow procyclical fiscal policy. Based on this information the following hypotheses are formulated: H1 In their election manifestos the Dutch left wing parties are more likely to propose countercyclical

fiscal policy than Dutch right wing parties3

H1a In their election manifestos Dutch left wing parties propose fiscal adjustment measures to “tone

down” the economy before and after the crises4

H1b In their election manifestos Dutch left wing parties propose stimulating fiscal measures during the

crises5

1

Feldstein, 2002; Alesina & Tabellini, 2005

2_{Cusack, 2001} 3

The years taken into account during the 1970s energy crisis: 1971, 1977, 1981, & 1986 The years taken into account during the global financial crisis: 2006, 2010, 2012 & 2017

4_{1971, 1981, & 1986; 2006 & 2017} 5

(17)

16 H2 In their election manifestos the Dutch right wing parties are more likely to propose procyclical

fiscal policy than Dutch left wing parties

H2a In their election manifestos Dutch right wing parties propose stimulating measures to stimulate

the economy before and after the crises 6

H2b In their election manifestos Dutch right wing parties propose fiscal adjustment measures during

the crises7

These hypotheses cover the period before, during, as well as after the crises. The reason for this is that the cycles encompass not only what the fiscal policy plans do at a certain moment in time, but they show a change in the fiscal policy plans. When there are no periods of crises the left wing parties tend to follow countercyclical policies in the sense that they do not propose stimulating measures, whereas right wing parties follow a procyclical policy in the sense that they do stimulate the economy. In times of crises these roles reverse. Therefore, looking at periods before, during, and after the crises will tell whether or not the political parties do indeed follow the policy cycles the theory states they do (Cusack, 2001; Feldstein, 2002; Alesina & Tabellini, 2005).

There is also a theory that states that it is more likely for left wing parties to propose fiscal policy measures that intervene in the economic situation than it is for right wing parties. The reason for this is that left wing parties tend to want to influence the economy more. They do this by proposing more intervention of the government in situations of economic downturn. Right wing parties on the other hand tend to pull back their hands and let the economy run its course by relying on market mechanisms (Blais, Blake & Dion, 1993; Kelly-Gagnon & Geloso, 2013).

H3 Left wing parties are more likely to intervene in situations of economic crises than right wing

parties are.

Coverse and Pierce (1985) state that partisanship is not static but that it can be subject to change throughout the years. Left parties then tend to drift more towards right party stances and vice versa. Budge (1994) disputes this and says that parties do not tend to converge in terms of election policy because the different ideologies will not allow for a convergence in policy. To test this, the following hypothesis has been formulated:

H4 Partisanship in fiscal policy of Dutch political parties is not static but can change throughout the

years

6_{1971, 1981, 1982 & 1986; 2006 & 2017} 7

(18)

17

4. Methods

In this chapter the research method will be described more elaborately, the choices that have been made will be justified and the scheme for analysis will be presented. The research method of choice is content analysis. This research project looks into whether or not there is difference in the responses of Dutch political parties in terms of fiscal policy in times of crises. The distinction that is being made is between left wing and right wing political parties, which is be done through the following research question: “Is there a partisan-based difference in fiscal policy responses between Dutch left and right wing political parties during periods of economic crisis?”

Krippendorff (1983) defines content analysis as follows: “content analysis is a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use” (Krippendorff, 2004, p19). According to Krippendorff (1983) content analysis is a new way of understanding the meanings of texts. These meanings refer to the wider cultural context that they take part in. So while the analysis is not exactly new anymore, it is still relevant (Rose, 2001). Content analysis is the most useful research in this particular project; because it puts an emphasis on the role that context has in explaining the meaning of certain phenomena. This project looks specifically at the role partisanship plays in fiscal policy. This in turns is dependent of the economic situation of the moment. Therefore, context plays is of influence (Krippendorff, 2004).

In further content analysis literature, scholars have given three definitions of content analysis: definitions that take content to be inherent in a text; definitions that take content to be a property of the source of a text; and definitions that take content to emerge in the process of a researcher analysing a text relative to particular context. Each definition leads to a different way in which content can be analysed (Krippendorff, 2004). The last definition is the one that will be used in this research. The expectation is that the meaning of the content is subject to the cultural context in which they take part; partisanship, but also crises times, and the ruling economic theories.

4.1 Data

The data used in this analysis are the election manifestos of the Dutch political parties that ran for government in the times of crises8. The election manifestos are collected from the Manifesto Project Database and the documentatiecentrum Nederlandse politieke partijen. The Manifesto Project is an online database that covers over 1000 political parties’ election manifestos from over 50 countries, scattered over five continents, from 1945 up until now, and provides the policy positions of those political parties. It provides its users with coded and un-coded election manifestos, as well as

8

(19)

18 codebooks, related project and publications. It is a database best used in comparative political studies9 (Manifesto Project, 2017).

The election manifestos used in this analysis are the manifestos of the political parties that ran for government in the years prior to, during and after the crises. The reason for this is that this project looks into if there is a certain cycle that the political parties follow. This cycle encompasses periods before, during and after the crises. To see if that cycle indeed exists, a sufficient number of years need to be taken into the analysis. The counter- and procyclical policies both contain information about how a party acts in times of economic downturn, but also in times when there is no economic downturn. In order see if there is indeed a difference, the analysis needs to done over a sufficient number of years, containing both periods of normal economic activity, as well as activity of economic downturn.

Not all the parties that were active in the analysed years are analysed. The reason for this is, in order to see in the left and right wing parties indeed follow a cycle, the same parties need to be analysed prior to, during and after the crises. This gives a better picture into whether or not there is a cycle followed by the left wing and right wing parties. Therefore, a selection of political parties has been made, which can be seen in table 4.1 and 4.2.

For the analysis a number of election years have been chosen. For the 1970s energy crisis the chosen election years are the following: 1971, 1977, 1981, and 1986. The year 1971 is the election year prior to the crisis, in 1977 the energy crisis was in full swing, and in 1981 the crisis had officially ended. The year 1986 is also analysed because it takes some time for politicians to acknowledge the end of a crisis. According to Gourevitch (1986), for politicians a crisis only truly ends in terms of policy response when the next crisis hits. The late 1980s and early 1990s were periods of economic downturn (Gourevitch, 1986; Poterba, 1994), whereas the period from 1983 to the late 1980s was a relatively stable period (McMahon, 2000). Therefore, the final election year taken into account is 1986, since this was a period of relatively ‘normal’ or stable economic activity. The years 1972 and 1982 were a re-election years. These years, however, will not be taken into account because many of the parties either used their old election manifest or did not produce a (new) election manifest at all due to the election being on such short term.

For the global financial crisis the chosen election years are 2006, 2010, 2012 and 2017. The year 2006 was the election year before the crisis hit national economies, including the Dutch economy, 2010 was the year when the crisis was still in full swing and in 2012 the aftermath of the crisis was still

9

(20)

19 sensible. From 2013 onwards the economy in the Netherlands slowly started to recover and in years to follow became more stable again (Klein, 2016), which is why the election year 2017 is also part of the analysis.

The parties that will be analysed are those that formed a coalition in the government. The reason for this is that this data is convenient for follow-up research – looking into whether parties actually stick to the policies they propose when they come to power. During the 1970s energy crisis there are four major parties. The other right wing and left wing parties that made it into parliament got a very small number of seats (<3) and are therefore left out of the analysis. For the period of the global financial crisis a similar reasoning has been used. There is, however, a difference between the two time periods. During the 1970s oil crisis the parties that formed the coalition has a significant majority in parliament. This was not the case in the period of the global financial crisis. The divide has been more spread. This is taken into account in selecting the parties that are analysed. Parties that exist with the goal to express their beliefs rather than run the country, such as the Staatkundig Gereformeerde Partij (SGP) and the populist party the Partij voor de Vrijheid (PVV) are left out of the analysis.

The 1970s energy crisis:

The starting point for the analysis of the 1970s energy crisis is 1971, therefore the election manifestos of that year will be analysed. The election manifestos of 1972 will not be analysed. There are two reasons for this. The first one is because some of the election manifestos of that year do not exist. The other reason is that if they do exist, they are very similar to (or the same as) the ones from 1971. The coalition that will be looked at, however, is the one that formed in 1972. The reason for this is that the coalition formed the year before did not last longer than a year and a number of changes took place. One of the parties (DS’70) did not manage to last throughout all of the years that are taken into the analysis. Besides this three of the parties, Anti-Revolutionaire Partij (ARP), Christelijk-Historische Unie (CHU) & Katholieke Volkspartij (KVP), that were part of the cabinet in 1971 merged in the following year and formed a new party, the CDA. Because this party took part in the elections from 1972 onwards, for these parties, the 1972 election manifesto will be used.

 1972-1977: Kabinet- Den Uyl, a coalition between the PvdA, D66 and CDA  1977-1981: Kabinet- Van Agt I, a coalition between the CDA and VVD  1981-1982: Kabinet- Van Agt II, a coalition between the CDA, PvdA and D66  1982-1986: Kabinet- Lubbers I, a coalition between the CDA and VVD

(21)

20 The global financial crisis

 2006-2010: Kabinet- Balkenende IV, a coalition between the CDA, PvdA and ChristenUnie  2010-2012: Kabinet- Rutte I, a coalition between the VVD and CDA

 2012-2017: Kabinet- Rutte II, a (minority) coalition between the VVD (with PVV support) and the PvdA

 2017-?: (Potential) Cabinet with the VVD, CDA, D66 & GroenLinks.

While D66 was not part of the coalition from 2006-2017 the party is still taken into account in the analysis. There are two reasons for this. Firstly, there is a consistency reason. This project will look into partisan-based change within parties throughout the years. During the 1970s oil crisis D66 was a government coalition party twice, which is why it is analysed in that period. During the global financial crisis the party kept its presence and kept growing. Secondly, the party is up for being part of the coalition in 2017.

In table one and two the relevant political parties that were active during the periods of crises are displayed.

Table 4.1 Political Parties during the 1970s Energy Crisis

Political parties10 Active period Left/right wing position11

Christen Democratisch Appèl (CDA) 1972-1986 (Centre) right wing

Democraten 66 (D66) 1971-1986 Left wing

Partij van de Arbeid (PvdA) 1971-1986 (Centre) left wing

Volkspartij voor Vrijheid en Democratie (VVD) 1971-1986 Right wing

Table 4.2 Political Parties during the Global Financial Crisis

Political parties Active period Left/right wing position

Christen Democratisch Appèl 2006-2017 (Centre) right wing

Democraten 66 2006-2017 Left wing

Partij van de Arbeid 2006-2017 (Centre) left wing

Volkspartij voor Vrijheid en Democratie 2006-2017 Right wing

10_{Sources: www.parlement.com; www.nlverkiezingen.com} 11

(22)

21 4.2 Using and testing the data

The codes

The data is coded in a mixture of a priori and emergent coding way. This means that the pre-existing codes will be taking into account, but after examination of the data, the codes have been adjusted to fit the analysis best (Krippendorff, 2004; Neuendorf, 2017). The pre-existing codes used are those from the Manifesto project. The more specified codes were established after examining the election manifestos. This resulted in slightly adjusted codes or newly added codes . Special attention will be paid to the parts where government expenditure and taxes are discussed. The other parts are not relevant in terms of fiscal policy and will, therefore, be left out of the analysis. The way the text is coded includes whole sentences, but also partial sentences. Partial coded sentences are sentences wherein two or more codes can be applied, for example: |taxes for the higher income groups will be increased|, so that taxes for the middle to lower income groups can be reduced.|. The text is partially hand-coded and a partially the by the Manifesto project coded texts have been used. These already coded texts were checked and were necessary recoded using the adjusted code book.

The relevant codes that the manifesto project used are the following:  (Anti) Economic Growth

 Controlled Economy

 Corporatism/ Mixed Economy  Economic Orthodoxy

 Economic Planning  Free market economy

 Incentives  Internationalism

 Keynesian Demand Management  Labour Groups

 Market Regulation  Protectionism

The codes displayed above do not include the extensive description the Manifesto Project uses. The elaborate code display can be found in appendix 1. After scanning through the selected election manifestos the codes from the manifesto project have been adjusted to fit the analysis of this research project better. This resulted in the following list of codes. Each of these codes, like the codes from the manifesto project, contains a number of elements.

 Mentions of crises  Economic growth  Economic orthodoxy  Economic spending  Free market economy

 Keynesian Demand Management  Incentives

 Internationalism

 Labour in hands of the government  Labour in hands of the companies  Taxes

(23)

22 What can be seen is that a number of codes have been left out: corporatism/mixed economy, foreign financial influence, foreign special relationships, labour groups, market regulation and protectionism. The reason for this is that after reading the election manifestos it became clear that these codes where never or hardly ever present. The exception on this is ‘corporatism/mixed economy’. This occurred in the texts to some extent, but this code has been taken into the code labour in hands of the government. The reason for this is that the mentions on this specific subject, in almost all cases, were a means to limit control of the companies. ‘Controlled economy’ is another code that has been integrated into labour in hands of the government because most mentions of this ‘controlled economy’ referred to the governmental influence on (minimum) wages.

All of the codes will have an abbreviation that will be used in the displayed tables. (1) Mentions of crises (Crises)

Includes: Reference to crises

Relation to literature: Mentions of crises relates to the literature regarding the statement that the left is more likely than the right to intervene in situations of economic crises.

Example: “We must realize that the current spending pattern of the government is based on a level of prosperity that ceased to exist with the economic crisis” (VVD, 2010)

(2) Economic growth (EG)

Includes: Mentioning economic growth as a goal

Relation to literature: Economic growth relates to the difference between right and left regarding the fiscal policy they implement. The expectation is that the more economic growth is mentioned as a goal, the more measures that lead to economic growth are proposed.

Example: “the election program if focused on increasing economic growth” (PvdA, 1986) (3) Economic orthodoxy (EO)

Includes: Mentions of reduction of budget deficit, government retrenchment in crises, government savings in time of economic downturn

Relation to literature: Economic orthodoxy refers to measures taken in economic cycles. It can be a right wing and/or a left wing measure, depending on the economic context it is taken in.

Example: “The coming years will be about moderation”,“the intention is to reduce the growing percentages of government spending every year” (CDA, 1971)

(4) Economic spending (ES)

Includes: Increasing budget deficit, government involvement and spending in times of economic downturn

(24)

23 Relation to literature: Economic spending refers to measures taken in economic cycles. It can be a right wing and a left wing measure, depending on the economic context it is taken in.

Example: “we can have higher budget deficits when the economic tide is against us and have no shortcomings” (D66, 2017)

(5) Free market economy (FME)

Includes: Laissez-faire economy, superiority of individual enterprise over state and control systems, need for unhampered individual enterprises, less government involvement in the economy. Relation to literature: In general right wing parties are more likely to adopt a free market economy approach than left wing parties. This means that right wing parties are less likely to intervene in the economy than left wing parties are.

Example: “Real economic growth comes from the hands of people.” (VVD, 2010) (6) Keynesian demand management (KDM)

Includes: Favourable mentions of demand side oriented economic policies, more government involvement in the economy, more government involvement in companies.

Relation to literature: In general left wing parties are more likely to adopt a Keynesian economic approach than left wing parties. This means that left wing parties are more likely to intervene in the economy than right wing parties are.

Example: “When we realize new growth and employment, government spending is cut off and revenue increases, which accelerates a balance in the budget deficit” (PvdA, 2012)

(7) Incentives (Inc.)

Includes: Measures that puts consumers and businesses in a favourable position to spend money or invest money.

Relation to literature: Incentivising is one of the measures in which a government can push a fiscal stimulus measure. Incentives push people to spend money (which will lead to economic growth). The higher the number of incentives, the more likely it is that the party follows a fiscal stimulus policy. Example: “establishing new jobs should be promoted by stimulating business, by (...) improving labor supply and labor supply and (also) reducing labor costs” (D66, 1986).

(8) Labour in hands of the government (Labour gov.)

Includes: job creation is in hands of the government, wages are in hands of the government, less control for the companies

Relation to literature: This code refers to the difference between left and right in terms of government involvement (in economic crises). The higher the mentions of this code are, the higher the level of government involvement.

(25)

24 Example: “After a period of necessary extra mitigation, the CDA aims to maintain the wages and benefits (…) at the current purchasing power level” (CDA, 1986)

(9) Labour in hands of companies (Labour comp.)

Includes: Wages are in hands of the companies, more control for companies, job creation is the responsibility of the companies.

Relation to literature: This code refers to the difference between left and right in terms of involvement of government and companies in the economy. It is likely that a higher involvement of companies in the economy results in a lower involvement of the government.

Example: “Economic growth and, with that improvement of employment, can be achieved through an integrated approach that strengthens the structure of businesses.” (CDA, 1986)

(10) Higher taxes (Tax +)

Includes: Higher taxes are justified, higher taxes are justified as long as it does not interfere with the spending capacity of the people, higher taxes are justified but only for the higher income groups. Relation to literature: This code refers to fiscal measures. Taxes are a part of fiscal policy. It is a way to influence the economy through spending.

Example: “if the population wants us to increase the level of governmental services, this calls for an even stronger increase of financial burdens imposed upon the population” (CDA, 1971)

(11) Lower taxes (Tax -)

Includes: Lower taxes are justified, lower taxes are justified for lower-middle income groups Relation to literature: This code refers to fiscal measures. Taxes are a part of fiscal policy. It is a way to influence the economy through spending.

Example: “By lowering taxes on labor, we increase people’s purchasing power, which increases demand in the economy, which makes it more attract for employers to hire people” (D66, 2017) (12) State investments positive (State inv. +)

Includes: Investments in the welfare state need to go up, investment in capital needs to go up, investment in small and medium sized enterprises needs to go up, investment in export needs to go up.

Relation to literature: state investments, positive means that the government wants to increase its public spending. This is believed to result in economic growth. Therefore, proposing these measures is part of a fiscal stimulus policy.

Example: “Investments help to create space for entrepreneurs and recreation (…) which results in jobs and economic growth” (D66, 2012)

(26)

25 (13) State investments, negative (State inv. -)

Includes: Investments in the welfare state need to go down, investment in small and medium sized enterprises needs to go down, investment in import and export.

Relation to literature: state investments, negative means that the government wants to cut its public spending, often to reduce the budget deficit. Measures that include reductions of budget deficits are fiscal adjustment measures.

Example: “In this way we get back as much of the tax money that was previously invested in financial institutions” (VVD, 2017)

Using the codes

Per hypotheses certain codes will be used to answer whether or not the hypothesis can be rejected. For hypothesis one and two the following codes will be used: economic growth, economic orthodoxy, economic spending, incentives, higher taxes, lower taxes, state investments (positive), and state investments (negative). All of these codes fall into spectrum of fiscal adjustment or fiscal stimulus measures. Both hypotheses have been split into two hypotheses, where H1a and H2a cover the time period before and after the crisis and H1b and H2b cover the time period during the crisis. For left wing parties, the expectation is that before and after the crises, they follow fiscal adjustment measures. The codes used here are therefore those of fiscal adjustments: economic orthodoxy, higher taxes, and state investments (negative). For right wing parties the expectation is that before and after the crisis they follow fiscal stimulating measures, which lead to the codes that cover fiscal stimulus: economic growth, economic spending, incentives, lower taxes, and state investments (positive). For the times during the crisis the measures for left wing and right wing parties are reversed.

Hypothesis three will be analysed using the codes Keynesian demand management, incentives, labour in hands of the government, and state investments, positive. The reason for the use of these codes is that all these codes require active government involvement, which is exactly what this hypothesis looks at.

Hypothesis four will be analysed using all the data that has been collected and comparing that data in different years.

Displaying the data

The data will be displayed in relative rather than absolute numbers. The reason for this is the following: every political party analysed has their own election manifest. It is up to the parties themselves to put what they think is relevant in this election manifest. The detail and length they use is also their own choice. The preferences for the level of detail and the length of the election manifests differs per party. As a result not every election manifest has the same length. A good

(27)

26 example of this is the 1972 CDA election manifest compared to the VVD, PvdA and D66. The election manifest of the CDA is around 70 pages, whereas the election manifests from the other parties do not exceed 25 pages. The more pages an election manifest has, the more codes can be gathered from that manifest. Comparing an election manifest with, for example, 200 codes to an election manifest with 30 codes does not give a good representation of the comparison between the two. Therefore relative numbers will be used. The number of codes per code will be displayed in a percentage of the total amount of codes. This way you can see what amount of space a political party has allocated to a specific code. This will give a visual presentation of the difference between political parties and how important a specific code is to the parties analysed.

For example, the CDA has a total of 200 codes (100%) and 100 codes (50%) is spent on economic orthodoxy. The PvdA has a total of 400 codes (100%) and 100 (25%) of these codes is spent on economic orthodoxy, but 250 (62.5%) of these codes are spent on economic spending. Looking at the total number of economic orthodoxy codes of both parties it seems as though they care an equal amount about economic orthodoxy. However, looking at the relative numbers, you can see that this is not the case. Therefore, to avoid a skewed representation of the data, the relative numbers will be displayed.

Testing the hypotheses

The hypotheses will be analysed in two different ways. Firstly, the data will be displayed in tables. Looking at the tables shows the difference between left wing and right wing parties or between left wing and left wing parties, and right wing and right wing parties and based on the visual differences preliminary results will be discussed. This will be done, not only by looking at the differences in the tables, but also by looking at the percentage of the codes allocated to fiscal stimulus and the percentage of the code allocated to fiscal adjustment measures. Secondly, after this analysis a statistical analysis, a t-test, will be performed to test if the results are statistically significant.

The t-test will be performed based on the data that has been collected in the coding phase. By using the different codes and the amount of times (in %) these codes are mentioned, a procyclical-countercyclical scale has been created for every party per year. The positions of each party for every year are placed on certain spots on that scale. Besides this, the averages and average differences have been calculated. The average difference of the average position of the right wing and the left wing parties are taken and used to test if it is statistically different from zero by using a standard one sided t-test, assuming they have the same variance. This result will tell whether or not the difference between parties is significant. Because the dataset is relatively small, a p-value of 0,1 with a significance level of 90% will be used.

Partisanship in fiscal policy : partisan-based differences in fiscal policy between Dutch political parties during crises