The Financial Crisis and Economic Regime Change: The Case of the IMF

(1)

The Financial Crisis and Economic Regime Change:

The Case of the IMF

Abstract

This thesis studies potential change in economic thinking towards Post Keynesianism in the IMF. It offers insights possibly interesting to national governments and economic stakeholders influenced by IMF decisions. A case study is done on the IMF by document analysis of IMF country reports.

Leon Koldijk S1580175

MSc. Public Administration - Governing Markets: Competition and Regulation Master Thesis

13-03-2018

Supervisor: Dr N.A.J. van der Zwan Second reader: Dr J. Christensen

(2)

1. Introduction

“1929. 1945. 1973. 1989. 2001. One scarcely need identify the events to which these dates refer”. One ‘‘knows them when one sees them’’ as ‘turning points’ when old orders ended and new ones began to emerge” (Widmaier et all, 2007, p. 748). Today it is obvious that 2008, and its corresponding Global Financial Crisis (GFC), completes this list.

Post-Keynesian economists accurately predicted the failure of the U.S. economy as a consequence of the dominant economic regime; i.e. Neoclassical economics (Bezemer, 2009; Keen, 2013). When the financial crisis struck following the collapse of Lehman Brothers in September 2008, the United States and other western governments adopted Keynesian policies in lieu of traditional, free market, neoclassical policies (Kaletsky, 2011, pp. 1, 234-5). This change in economic policy makes for an interesting observation, especially as Neoclassical policies were widely applauded by policy makers and economists in the 1985-2007 period for leading to ‘The Great Moderation’ (of economic volatility) (pp. 68-72). The policy change from Neoclassical to Keynesian measures shows similarity to changes after the Great Depression (1924-1929) and World War II (1939-1945), when Keynesian measures were implemented by western governments to rescue the economic system (Palley, 2009, p. 6). For the student in Public Administration, such a quick pivot in economic policy creates an opportunity to study whether such a change represents a temporary deviation from a standard set of economic policy measures, or a fundamental change in economic thinking.

There is no scholarly consensus on whether or not we have witnessed a Keynesian, or post-Keynesian challenge as the start of a policy replacement (Palley, 2009, p. 20; Jäger and Springler. 2015, p. 2). Keynesian economics might exhibit elements from Neoclassical economics, post-Keynesianism separates clearly from neoclassical strands of economics. With the risk of oversimplification, Keynesian policy can be said to refer to their shared elements. The question of what kind of challenge we have witnessed, makes for an interesting point, especially when considering that post-Keynesianism provided a fitting explanation and prediction of the 2008-crisis. At the same time mainstream economic policies neither achieved the envisaged goal of solid economic growth, nor decreased high levels of unemployment (Jäger & Springler. 2015, p. 2). This brings us to the point where we are faced with an unresolved puzzle in Public Administration: mainstream economists are unable to deliver on the prophesized growth and employment, yet early evidence suggests that fundamentally their Neoclassical policies are not changed.

(4)

4

Economic policies are mostly determined at the national level; however, they are influenced by a whole range of institutions such as universities, think tanks and international organizations (IO’s). In particular the International Monetary Fund (IMF), World Bank and Organization for Economic Cooperation and Development (OECD) play an important role in guiding international and national economic policy and expectations. The role such organizations specify for themselves is one of ‘economic guidance’, yet they especially influence countries that require their assistance, as this typically requires confirmation to their views (Bradlow, 2007, p. 7). One of the ways through which IO’s influence national economic policy is through ideas (Verbeek, 1998, p. 14). Ideas are important causal factors in regime change (Block and Somers, 2005, p. 260-1). As such, during crises a window of opportunity may open when ideas may function as important inputs in the policy process (Kingdon, 2014, pp. 16, 20).

For this reason, this thesis wishes to look whether the financial crisis opened a window for ideational regime change to occur. The IMF is a representative case of IO’s that adhere to Neoclassical economic thinking and influences policy making in the national and international arena.

1.1 Research Question

This study aims to investigate whether economic ideational change occurs in economic policy making after major economic crises, such as the financial crisis. Neoclassical economic thinking is regarded to be dominant in economic policy making (Peet, 2009, pp. 16-7, 25), while post-Keynesianism has been tipped as its main contender for future dominance. This results in the following research question:

Does financial crisis lead to economic ideational change?

To answer this question, research is conducted using a case study methodology. The IMF is chosen as the unit of analysis for its great influence on economic policy making. Studying ideational elements in policy requires close study of detailed material. Hence annual reports on IMF member country economic policy are analysed by means of document analysis. To assess the impact of the 2008 financial crisis, a before and after analysis is carried out, comparing two reports before and two reports after the 2008. Data is collected using lenses for Neoclassical and post-Keynesian thinking in categories that fit both schools.

(5)

5 1.2 Justification

The focus of this study is on Neoclassical economics and post-Keynesianism for two reasons: Firstly, Neoclassical economics can be regarded the dominant ideational school in economic policy since the 1980s (Palley, 2009, pp. 7-8). Secondly, post-Keynesianism, mainly in the form of Minsky´s work, has elements that according to a variety of economic reporters seem to diagnose the origins of the 2008 crisis (Bezemer, 2009; Keen, 2013, Palley, 2009). Moreover, in public debates on the crisis– most prominently the austerity vs. stimulus debates– main characters such as Niall Ferguson and Paul Krugman strongly relied on Neoclassical and Keynesian ideas (Skidelsky & Fraccaroli, 2017). These are steps described by Block & Somers (2005, p. 266) that lead up to a potential change in economic ideational regime.

International organizations play an important role in shaping international and national economic policy. Because international organizations see their role as ‘providing economic guidance’, it makes sense to closely study these ´guides´ for any economic ideational changes. For example, Stiglitz (2002, pp. 42-4) notes that countries in need of IMF assistance are required to align their economic policies to IMF policy prescriptions. Inconsistent or erroneous policy advice by international organizations can therefore have aversive effects on individual countries and even the world economy. Since the financial crisis a number of countries were in need of IMF assistance, creating a window of opportunity for influencing economic policies.

As such, this thesis hopes to do three things: (1) it seeks evidence that supports or disavows theories of ideational change in policy making in relation to economic crises. (2) By means of the first point, it secondly hopes resolve the puzzle of whether and in what way neoclassical economics is retaining its position in economic policy making. Finally, it hopes to provide a research structure that can be used as the basis for additional research on ideational change in economic policy making, and more in particular an elegant, yet complete overview to check for neoclassical and a seemingly non-existent overview for post-Keynesian economic policies.

1.3 Structure of the Thesis

This thesis is split up in three parts namely: (1) theory, (2) research methodology and (3) the results of the document analysis. Chapter two deals with the first part on theory. First, it shares an understanding of various concepts and theories related to (a) ‘ideational regimes’ and (b) ‘ideational regime change’. It provides concepts and definitions necessary to assess a potential shift in thinking, and whether such changes truly represent an ‘economic regime change’ consistent with the respective literature. In chapter three, the research methodology is presented

(6)

6

it deals with (1) the choice for the case study design, (2) the causal mechanism between ideas and policy in times of crisis, and (3) the method of analysis. Chapter four, five and six cover the empirical part. Chapter four, firstly, presents the functioning of the IMF, and secondly sets out Neoclassical and post-Keynesian thinking. The latter provides a clear understanding of the central concepts. Chapter 6 presents the analysis itself. Finally, chapter seven draws conclusions.

(7)

7

2. Theoretical framework

2.1 Ideational regimes

Economic policies are typically rooted in economic ideas. According Fred Block and Margaret Somers the importance of ideas in regime change can hardly be overestimated (2005, p. 260-1). In various theories of policy making, ideas have prominent roles- yet each employs them in a different manner. Ideas can be seen as proxies for interests, road maps towards or focal points for goals of cooperation, as strategic constructions to achieve goals, and as narratives shaping the understanding of events, collective memories or traditions (Schmidt, 2008, p. 306).

Central to this thesis is the concept of ideational regime (cf. Somers & Block, 2005). Ideational regimes are deeper worldviews, and can be seen as sets of taken-for-granted assumptions of how the world works (Danielson & Stryker, p. 134). Danielson & Stryker (p. 134) regard them to be comparable to Peter Hall’s (1993) concept of policy paradigms and programmatic beliefs. Yet, the difference between policy paradigms and ideational regimes can be found in their embeddedness; ideas and assumptions are not only known, understood and employed by policy makers, such as with policy paradigms, but ideas and assumptions also become social institutions (Block & Somers, 2005, p. 265). Institutions can be defined as “…facts by human agreement” such as citizenship or money and yet can be regarded ‘objective reality’ (Searle, 1995, p. 1). Individuals first have to internalize ideas into their system of what represents objective truth (p. 138) to make the step from idea to latent- but present belief. When they are broadly held in society, it is possible for social life to practically and cognitively function according to the beliefs: ideas are then embedded in society and become taken-for-granted assumptions.

To explain the role of ideational regimes in policy making, I rely on Block & Somers (2005) and Marc Blyth (2002). To bring this view forth I rely on Hall’s work (1998) on policy paradigms. Hall explains that for policy makers, policy paradigms provide interpretive frameworks of ideas and standards that specify (1) the goals of policy, (2) its instruments, and (3) nature of the problems to be addressed (p. 279). This provides a framework of ideas that is comprehensible and plausible to policy actors. Hall makes clear that its framework is implicitly present in the work language used by policy actors, making it a paradigm. Hall explains that policy paradigms work as prisms through which policymakers look at the economy. These paradigms have first, second and third order levels, respectively: (a) the policy rules in place, (b) the policy instruments and (3) the paradigm- which contains the first two levels (p. 280).

(8)

8

This concept of paradigm shows similarity to the concept of intuitions. Hall’s concept is similar to the concepts of Blyth (2002), and of Block and Somers, yet it differs in depth. Ideational regimes can be seen as the level above the policy paradigm, while simultaneously still informing the all levels below. Schmidt (2008, p. 304) explains that paradigms typically operate in the foreground, and are debated regularly by policy makers and politician in public debate. Yet in the background an implicit philosophical world view present containing latent, final assumptions. This is the ideational regime. The ideational regime is discussed seldom by policy makers, politicians and, unless at times of real crisis (Schmidt, 2008, p. 304). This level includes the explicit and implicit ideas that form the philosophical basis for how knowledge is accrued, what is valued, and what society should look like (Schmidt, 2008, p. 306). Ideational regimes become embedded, when they become part of standard legal, political, and organizational structures (Somers and Block, 2005, p. 264). Hence, the taken for-granted assumptions as social institutions inform the explicit first second and third degree levels of the policy paradigm. It is this broad, all-encompassing perspective of the ideational regime- that is used in this thesis.

(9)

9

1.2 Ideational regime change and international organizations

International organizations are part of international regimes, a concept central to international relations (IR) and a separate concept from ideational regimes. The international regime is discussed here to gain an understanding of the influence of international organizations on policy making. International regimes are defined by Krasner (1983) as “…sets of implicit or explicit norms, rules & decision making procedures around which actors’ expectations converge in a given area of international relations”.

In International Relations controversy remains on why regimes exist (Young, 2012). Yet, Young explains that these entities play significant roles in administrating state provisions or administrating provisions for respective international regimes. Sometimes international organizations are seen not as mere administrators but as an independent influence on the regime as well. The relation between institutions and organization has therefore become an important focus in the analysis of international regimes.

The rationalist theories neorealism and neoliberalism are mainstream in IR and have been dominant particularly in the 1980’s and 1990s. These theories see states as unitary actors pursuing the national interest, which is assumed to be survival of the state in the anarchy of international affairs (Waltz, 1979, p. 10; Keohane, p. 52; Ashley, p. 237). The structure of international affairs is defined by major powers (ibid, p. 93). Theories that deny “… the central role of states will be needed only if non-state actors […such as international organizations…] rival[ing] or surpass[ing] the great powers…” (Waltz 1979, p. 95) (Waltz, 1979, p, 60, Ashley, 1984, p. 239, Keohane, p. 52). Hence states decide themselves whether to cooperate or seek assistance (e.g. in the form of international organizations) from others and in doing so to limit its freedom by making commitments to them (Waltz, 1979; Keohane, 1985, p. 52).

Neorealists see international regimes and hence international organizations as extensions of hegemonic states to serve their interest (Strange 1982). Some neorealist do see some role for regimes; Notably Mearsheimer (1995, p. 95) holds that states cooperate in regimes to have absolute gains as long as they do not suffer relative losses to other states. For Krasner (1982) in this view International Regimes are conceived as intervening variables in between the causal variables as power and interest, and outcomes and actor behaviour. He explains that as such, causality would look as follows:

Basic causal variables → Regimes → Related Behaviour and Outcomes.

Neoliberals represent the other side view in rationalist IR theory: states are interested in absolute gains even with a relative loss. A gain can also take the form of wealth rather than only

(10)

10

power, making absolute gains sufficient for cooperation. Hence regimes and international organizations take collaborative form to overcome the obstacles of the anarchic structure of international system (Krasner 1982; Keohane, 1984). Hegemons are prepared to carry most costs of forming and maintaining the regime and have the capabilities or power enabling them to control or constrain actor behaviour (Keohane & Martin, 1984).

While both theories see regimes as vehicles to pursue interests, there is also room for the possibility that ideas and norms can hold back states in the way they pursue them (Finnemore & Sikkink, 2001). Keohane & Nye (1996, p. 280) for example even pose learning in the cooperative setting of regimes. However, rationalist theory does not considers that norms or ideas could define states’ interest itself. This results in an essentially static rationalist world view.

Finnemore & Sikkink (2001): “constitution is causal, since how things are put together makes possible, or even probable, certain kinds of political behaviour and effects…” because they are permissive and probabilistic, however, such explanations are necessarily contingent and partial. Making general claims can only be done by making assumptions which lack foundations. Their different prioritizations and assumptions lead to different approaches and substantive claims. Therefore constructivism makes claims about the nature of social life and social change focusing on the role of ideas, norms, knowledge, culture, and argument in politics (p. 393).

They see international regimes- and hence international organizations- as shaping identity and interest. On the one hand their dynamic world view makes it possible that ideas define state interests and hence the form international regimes and international organizations take. In this sense it acts as a strategic construction. On the other hand constructivists hold that regimes assist actors in defining interests and demonstrate that actors share a common world view- acting as a focal point (p. 296). This last part is especially important in regards to this thesis, as this would mean that cooperation in international regimes- and therefore international organizations- spreads ideation the ideational regime.

1.3 Ideational regime change

According to Milton Friedman only a crisis, actual or perceived, can produce real change in economic policy; “the role of economists is to develop alternative policies and keep them available until they become politically inevitable” (Friedman in Palley, 2009, p. 6). Indeed, scholars commonly identify crises as moments when ideas become inevitable (e.g. Schmidt, 2008, 304, Hall, 1993 p. 285, Kingdon; Blyth, p. 8-11). To understand how crises can

(11)

11

open windows of opportunity, different approaches to policy change are discussed: the paradigm model by Hall (1993), the theory of institutional change by Blyth (2012), and the mechanism of ideational regime change by Somers & Block (2005).

Policy change

Despite its shortcomings, the paradigm model by Hall (1993) can be used to understand dynamics of policy change. According to Hall (p. 276), policy change equals policy learning, defined as “a deliberate attempt to adjust goals or techniques of policy in response to past experience and new knowledge”. The process of change in policy in his theory works mainly through politicians, and includes a broad array of social actors: the media, outside interests, contending political parties (p. 276). According to Hall, learning does not occur as a result of autonomous action by the state, but in response to an evolving societal debate that quickly becomes bound up with electoral competition. The political proponents of the policy are stripped of their authority when policy failure ensues, and eventually are voted out of office. Hall thus situates paradigmatic policy change, when newly elected politicians take office, as such completing the paradigm shift (p. 288).

In this process Hall sees three variables of change, from specific to general: (1) change of the precise settings of policy instruments, (2) the sort of policy instruments specific: the overarching goals, (Hall 1993, p. 279). At the most basic level, change takes place mostly incrementally (e.g. the increase of an existing benefit). At the second level, change takes form through (strategic) action: e.g. implementing the use of an interest rate tool, to secure an existing primary goal of price-stability. Third would then be a change in the hierarchy of goals itself: e.g. increasing employment through fiscal policies, instead of inflation reductions through interest rate policy. First and second order changes can be regarded ‘normal’ policy making. Change at the third level, will lead to a change in the overarching terms and typically also involves change at normal level (p. 279). When third order goals change, it can be regarded a shift in paradigm (p. 279).

Paradigmatic changes can be expected when persisting ‘anomalies’ occur, in other words developments that cannot be comprehended fully in light of the existing paradigm. Dominant paradigms then lead to wrong forecasts and policy. As the anomalies build up, policy makers will stretch the terms of the ‘paradigm’ which undermines the intellectual coherence and precision. This policy process might also include implementing experimental policies; when these policies fail, the paradigm will gradually lose authority. The change to a new policy paradigm hence involves a build-up of anomalies (p. 283-5). This process description fits what

(12)

12

Kingdon (2014, p. 19) calls incremental change of policy proposals and change, as contrasted with sudden change. ‘Hall remarks that : “learning is indicated when policy changes as the results of such a process” (p. 278).

Hall’s model is criticized for its treatment of causality. According to Mark Blyth (2012, p. 7) change in elements of a present set of institutions are juxtaposed to the ones of a previous set. A new institutional equilibrium then seems to reconstitute itself automatically, leading to the following model:

Institutional equilibrium → Punctuation → new institutional equilibrium.

As such it follows a post hoc, ergo propter hoc reasoning. New institutions and regimes are determinate functions of the problems of the previous institutions in new conditions. Yet, what comes after cannot automatically be assumed to be caused by what comes before, according to Blyth (2012, p. 7). Uncertainty and agent interests, which influence their actions under this uncertainty, are as such omitted (p. 9). In Hall’s view uncertainty exists in the form of a puzzle: when uncertainty persists it can and will automatically be overcome through by voting in new authority.

In contrast, Blyth holds that what comes after can only be linked to what comes before, when specific causal links can be made between former and latter objects. Therefore Blyth holds that causes such as underlying, non-specified pressures of society, the political system or other useful assumptions do not provide satisfactory explanation (p. 10). The static account of interests scholars like Hall have enables them to find or assume the causes within their own theoretical framework. This is in much the same manner as classical realists found reasons for events in international politics. Blyth holds that interest-based models do not specify these links because agents’ interests are inferred from observed outcomes. As such he concludes that it is a model that is at least underspecified, but circular at worst as it only finds what it looks for (p. 8).

Because causes and solutions are not given, institutionalist approaches such as proposed by Blyth hold that interests have to be explained, particularly in a situation of great uncertainty. In these situations agents are unsure about what their actual interests are and how to realize them. Economic crisis is an example of such a situation. Then interest becomes something to be explained rather than something that has to do the explaining. Additionally, the notion of what a crisis actually is, becomes ambiguous (Blyth, 2002, p. 9). Institutional views hold that the nature of a crisis is not determined by its own “effects, dislocations or causalities” (p. 9). Instead, Blyth stipulates that institutional failure is often not self-apparent to agents on the

(13)

13

ground, who demand obvious solutions. Even though the failure of an institutional order and its regimes is destabilizing, it does not mean that this instability automatically leads to a modified or new order. Therefore institutionalists hold that how agents redesign and rebuild regimes, and the conditions under which such activities take place, needs to be analysed (p. 8).

According to Blyth, in times of crises and specifically the accompanying uncertainty are the moments at which institutional arrangements are opened up for change (Blyth, 2002, p. 8-11). Somers and Block (2005, p. 260-1) emphasize that in regime change, ideas “… exert extraordinary political influence” in a causal way. Blyth (2002, p. 11) remarks that ideas are not the only relevant aspect or that institutional change is solely an ideational matter. This leaves room for existing interests and material aspects such as power. Nonetheless, in times of great uncertainty, such as economic crises, economic ideas become very relevant and instruct agents on the actions to take and the future to construct. This does not happen directly and neither automatically, as this depends on the ideas and a public debate.

Ideas as catalysers for change

To institutionalists such as Blyth, economic ideas provide a diagnosis of what constitutes an economic or financial crisis and what has to be done to solve the situation. This is then a construction that makes the uncertainty perceived by agents explainable, manageable and “actionable” (Blyth, p. 10). Blyth concludes that economic ideas are the interpretive framework for agents: they describe and account for the workings of the economy by defining constitutive elements and proper (and therefore improper) interrelations (p. 11). As such economic ideas do not “simply reflect the world that precedes them” but are causally powerful by themselves as they construct reality and allow agents to define a situation as a crisis. Some institutionalists see the possibility that ideas and the theories even have the capacity to make themselves true by changing reality to reflect their abstract theoretical models (Bourdieu, 1998, p. 95; Somers and Block, 2005, p. 262). Therefore, Blyth (2002, p. 10) concludes that it is essential to know what economic ideas exist with agents in times of economic crisis.

Blyth describes how economic ideas give agents a scientific and a normative account of the economy and polity. He provides a vision that specifies how these elements should be constructed. He concludes that economic ideas also function as blueprints for new institutions: they allow agents to reduce uncertainty, propose particular solutions to crises, and empower agents to resolve crises by constructing new institutions in accordance with ideas (ibid, p. 11). For Somers and Block (p. 266) this is the causal power of a set of ideas to unseat the previous set of ideas: The capacity to unseat depends on the ideas’ internal capacities or “epistemic

(14)

14

privilege”: a comparative advantage of a set of ideas over another set of ideas (p. 266). Here lays the role that Friedman describes for economists: developing and maintaining alternative policies.

The process of ideational regime change typically starts with external factors in the form of a crisis (Block & Somers, p. 266). Blyth (2002, p. 11) suggests an event sequence of three steps where ideas have different effects:

1. reduction of uncertainty 2. specification of causes 3. supply of new institutions

According to Blyth, this sequences make institutional change dynamic, contingent and political. As comparative statistical models exclude these steps they are unable to explain process or contingency (p. 11). When political agents mobilize ideas with adequate internal capacities, they are able to construct the unified experience of a crisis from a multitude of individual agents’ experiences. (p. 10). Blyth (p. 12) notes that political agents must “…argue over, diagnose, proselytize and impose upon others their notion of what the crisis actually is before collective action to resolve the uncertainty facing them can take any meaningful institutional form”. Available ideas therefore interpret the environment, reduce uncertainty and make purposeful collective action possible, which is of essential importance in determining the form of new institutions.

Block and Somers see a process with similar steps as the process described by Blyth, yet their process has a focus on the public debate over the causes and solution- instead of ideas merely providing solutions. Block and Somers also hold that the success of a new ideational regime depends on the way ideas are employed in a political discourse. It is necessary that the new ideational regime challenges the existing regime in a counterfactual manner. Political proponents of the new regime have to convince the wider public that when the alternative regime would have been in place, a crisis would not have occurred (p. 274). This involves a conversion narrative that helps “teach” individuals how to see the world and convert them to this world view (p. 274). Collin Hay (1999, p 321) sees this as “the mobilization of perceptions of crisis involv[ing] the formation and triumph of a simplifying ideology […] construct[ing] points of resonance with a multitude of individuated experiences”. The conversion narrative should also provides an explanation “why intelligent well-intending people did wrongfully believe in the bad ideas of the previous ideational regime” (Somers and Block, p. 266). Identity-building aspects of a regime are important to constructivists: a successful regime takes the form

(15)

15

of a “public narrative… [retelling]…the story of the nation’s meaning, morality, and place in the flow of history” (p. 266). In short, their process looks as follows:

1. A new set of ideas has to reframe the crisis by changing the very definition of reality (and the crisis) itself.

2. Then, there should be a battle between opposing ideas in the public arena

3. The new ideas have to establish themselves as the only possible solution to the previous regime.

When applying both Blyth’s and Block and Somers’ theories to the case of ideational regime change within the IMF I expect that (1) the financial crisis of 2008 presents IMF policy-makers with a puzzle and that (2) they will identify this as such. More specifically, I expect this puzzle to present a misfit with Neoclassical theory. I further expect (3) some ideas of Keynesianism to present likely solutions to IMF policy makes. Over time, I believe (4) post-Keynesian ideas to gain more ground and that Neoclassical assumptions are challenged. This last point would mean that fundamental assumptions which distinct an ideational regime from a policy paradigm, are not omnipresent anymore or not always unconditional. Hence, I expect some change in policy ideas shortly after the outbreak of the crisis, and more change further along. My main expectation is to see a battle of ideas in the documents analysed here, which constitutes a degree of ideational change. In the next section, I will cover ideational changes in regimes we have seen in the west over the past century.

(16)

16 2.3 Economic Regimes in History

In international relations, international economic regimes support an international economic order, for example in the way the trade and monetary regimes of Bretton-Woods support and comprise the Liberal International Economic Order. I do not conflate orders with the assumptions of the underlying ideational regime, as the respective orders operate on the second level in the form of a policy program or policy paradigm.1_{A number of international economic} orders in history can be distinguished from Ruggie (1982) and Maczynska & Pysz (2015):

 Classical Liberal Order (Laissez Faire Liberalism).

 Liberal International Economic Order (Embedded Liberalism).  The Neoliberal Order (Post 1970s).

I will present these orders in the coming part of this chapter.

2.3.1 The classical liberal order.

The classical liberal order before World War I was established under the hegemony of Great-Britain. It took form in the regimes for free trade and the gold standard and depended on the special position of Great-Britain in the international monetary system (p. 385-6).

By the gold standard mechanism international the balance of payments (BoP) differences were settlement through international gold transactions (p. 389). Domestic money supply would be adjusted via gold movement. Gold receiving countries would have an expansion of credit, while gold loosing countries would have contraction so that domestic prices and income would adjust bringing trade and therefore the BoP back in equilibrium (p. 389). In practice however gold rarely moved; instead small and short term capital movements adjusted the BoP via the principle of the gold standard mechanism.

Nonetheless, it was mainly Great Britain that supported the flow of investment capital in this system, acting as lender of last resort so that the international trade system remained liquid when shortages arose (p. 390). This required other states’ attitudes on how to conduct monetary policy to be aligned with the British position. In monetary policy they believed to “follow the market” essentially meaning following the Bank of England (p. 391). Maczynska & Pysz, (2015, p. 20) trace the origin of the belief in this order to Adam Smith’s 19th century classical economic theory and the market metaphor. It provided ideational coherence for policy to focus on with the goal to facilitate maximum scope to the market. (Ruggie, 1982, p. 386).

1_{Classical economic thinking can be seen to inform the classical liberal order, neo-Keynesian economics the}

liberal international order and Neoclassical Macroeconomics the neoliberal order. Yet all these economic theories rely on the core assumptions of Neoclassical Economics.

(17)

17

The maximum scope for the market meant that domestic, adjustments in wages and prices took precedence over the use of currency reserves as cushions for economic shocks (p. 390). Consequential domestic public pressure against free markets together with deep cyclical pressure and monopolistic tendencies in the economy gradually eroded classical liberalism. The Great Depression gave a final push and interventionist ideas spread, most notably with socialist Russia and managed capitalism in Germany (Maczynska & Pysz, 2015, p. 20). The end of the gold standard in 1931 marks the regime’s definite breakup.

2.3.2 Liberal international economic order.

The Great Depression enabled a rapid expansion of Keynesianism (Maczynska & Pysz 2015, p. 20). While the United States (U.S.) as new hegemon sought to establish a liberal order with a free trade regime and the dollar as a reserve currency, other industrialized countries opposed a liberal order (Ruggie, 1982, p. 393). The U.S. aspired a laissez faire market regime with a dominant position for itself in similar fashion to the British position in earlier regimes. Other industrialized countries were divided on the forms and depth of market intervention, but were united in the legitimacy of this objective. Ruggie (1982) concludes that the institutional reconstruction that followed, was one of compromise between the U.S. and other industrialized countries. The resulting order sought to safeguard domestic stability on the one hand, but on the other had multilateral, free trade rules. This system was to be erected through the IMF, which provided the currency exchange framework, and General Agreement on Tariffs and Trade (GATT). The capital controls, overdraft allowances and parity changes proposed by Keynes, who led the British delegation, were implemented. Nonetheless, the U.S. sought and achieved to water down these measures. Ruggie (p. 395) explains that this provides the context leading to the compromise of “embedded liberalism” or the Bretton Woods system. The Keynesian system of balancing the trade account via IMF funds did stabilize trade, yet also left room for private international finance to come up as an alternative adjustment mechanism. This alternative cushioned the BoPs of countries with trade deficits, at least in the short run. As such the system enabled trade deficit and surplus positions to increase making possible to pursue domestic policies centred on full employment (Lucarelli, pp. 12, 114).

This order lead to a period of economic boom from 1950 to the 1970s. However, a new puzzle presented itself policy makers, economists and politicians. From the 1960 inflationary tendencies arose in the U.S. economy which went unaddressed at first (Blyth, 2012, p. 132). From the end of the 1960s signs of difficulty for Western economies became more concrete. In 1969 for the U.S. economy showed the new phenomenon of inflation under economic

(18)

18

stagnation: an obvious puzzle for policy makers as Neo-Keynesian theory did not account for it. This led the Federal Reserve to increase interest rates and tight monetary policy; measures not in line with the embedded liberal order (p. 134). Blyth traces the problems at least partially to the Vietnam War and budget deficits, which would mean the measures were targeting the wrong cause (p. 134). At this time Britain faced similar problems although for reasons such as decolonization (Peet, p. 79).

.

2.3.3 The neoliberal order.

When measures were not sufficient, more extreme measures were sought: dollar-gold convertibility was suspended in 1971, to stop the outflow of gold associated with trade deficits. The consequential free-floating currency regime defacto ended the Bretton-woods monetary regime (p. 134). Moreover, the Nixon administration imposed price controls, when such controls were loosened commodity market reacted very volatile (p. 135). As such the policy paradigm was further weekend.

In 1971 unemployment hit 7% and fiscal stimulus was instated by congress, a tool within the Keynesian policy framework, yet not congruent with other measures. To make matters worse, from 1971 to 1973 the Organization for Petroleum Exporting Countries (OPEC) increased the oil price. This lead to more inflation and caused panic buying of oil (Blyth, 137-8). Yet Blyth explains that policy paralysis remain as the slump was also due to various, incoherent policies being put in place (p. 138). The combination of the failure to confront inflation, price-controls, volatile commodity, currency and labour markets violated in Blyth word “… the core underpinnings of American embedded liberalism”.

and the end of the Bretton Woods regime (p. 134).

With the election of Reagan in the U.S. and Thatcher in Britain, neoliberal policies were adopted at the beginning of the 1980s and took over the Keynesian discourse (Peet, p. 13).

This cleared the ground for a new theory to step in. Particularly the Chicago school of economics dismantled the dominance of the neo-Keynesian paradigm in economic theory and policies. This lead to an order typically called ‘neoliberal’ in the 1980s.2_{The new order is also} named the Washington Consensus by its critics. This refers to the Bretton Woods-institutions situated in Washington, which agreed on the proper economic policy prescription.

However, the origin of this neoliberal order can be found much earlier: in the beginning of the 20th century the turn away from laissez faire policies lead to a counter initiative by liberal

(19)

19

thinkers at two conferences: the 1938 Paris colloquium in and 1947 Mont Pèlerin Society’ (MPS). The name neoliberalism was chosen for an alternative to laissez faire liberalism after it exhibited serious flaws (Mirowski & Plehwe, 2009, p. 14). Secondly, and more importantly, in 1947 Friedrich Von Hayek brought together the (MPS) consisting of over twenty liberals (Hampe, 2010). The society discussed ways to fight the rise of the ascendancy of the state and “Marxist or Keynesian planning sweeping the globe” (Hayek in Matthijsen, 2012. P. 101). It was attended by Milton Friedman, Frank Knight, Friedriech von Hayek, Georg Stigler, Karl Popper, Ludwig von Mises, Wilhelm Röpke, Walter Eucken, Alexander Rüstow, and others – names that already indicate later division among liberal thinkers (Maczynska & Pysz, 2011, p. 28). At the Chicago School of Economics, economists Frank Knight and Milton Friedman – but also Austrian economist Von Hayek- developed neoliberal ideas further. Their ideas captured the minds of economists and policy-makers alike. The resulting economic zeitgeist can be found in the policies of Ronald Reagan and Margaret Thatcher (Mirowski & Plehwe 2009, p. 21-2).

Nonetheless, the meaning of neoliberalism remains contested. In their analysis of the word Neoliberalism, Boas & Gans-Morse (2009, pp. 140-3), find that it “… is a term that some, but not all, scholars use to refer to a variety of concepts whose unifying characteristic is the free market. They find that it refers at least to: (a) a set of economic reform policies, (b) a development model, (c) a normative ideology, and (d) an academic paradigm. Beside this wide range of meanings, they establish that it is used asymmetrically across ideological divides, particularly on free market standpoints. Every underlying concept to which neoliberalism can refer, has a contested normative connotation. The term is found seldom in literature that makes a positive evaluation of the free markets, but is in used widely by critics of free-market policies. They argue that terms should not be applied when: (a) the normative connotation of a term is contested; and (b) such normative connotations also go by other terms (pp. 138, 155). This is the case with neoliberalism and the argument can be made that its use should be reduced to a minimum. Unfortunately, the set of policies that can be distinguished after after the fall of the neo-Keynesian inspired liberal international economic order, commonly goes by the neoliberal order. I will however focus my attention at the fourth level Neoclassical ideational regime. Instead a Neoclassical ideational regime exists at the fourth level. Hence the choice for

Hence “…neoliberalism has become a conceptual trash heap capable of accommodating multiple distasteful phenomena without much argument as to whether one or the other component really belongs” (p. 156). This circumvents meaningful academic debate.

(20)

20

2.3.4 The neoliberal order and neoclassical economics.

There is no straightforward overlap between neoliberalism and neoclassical economics. According to Mirowski (2005), conflating neoliberalism with neoclassicism is wrong because neoliberalism is an interdisciplinary exercise including Austrian and Ordoliberal thought which are “…clearly at odds with [the] neoclassical orthodoxy” (p. 1). All these schools accept to marginalist principles: agent decision making in trade or production based on the benefit of gaining or supplying one additional unit of a product. Austrians, however reject models that aggregate such individual marginal decisions in supply and demand and hence equilibrium (p. 122). Ordoliberalism, as school of thought does adhere to the principles of aggregate supply and demand- yet sees a role for the state in ordering capitalism to get optimal market results (p. 123). Lastly, monetarists took the marginalist equilibrium approach a step further and applied it to money where in equilibrium, the marginal benefit (or utility) of holding wealth in money equals the marginal benefit of holding wealth in stock, real estate or any other form (Blyth, 2012, p. 139). Increased money supply would let consumers exchange money for assets and hence prices would rise to the point of equilibrium- explaining inflation.

Nonetheless, as Austrians, Ordoliberals and the Chicago School all adhere to marginalist principles they can all be included in the neoclassical tradition. As such the distinction becomes clear between classical economists such as Adam Smith and Neoclassicals. The former saw pricing as exogenously given: as some undefined natural value of labour. The latter see prices as determined by some form of juxtaposed internal auction, between two or more agents, of the marginal unit’s utility. Their commonality is thus found in their adherence to laissez-faire, but they rely on a different theoretical framework (Verango, 2016).

On the basis of these past changes in economic orders some expectations on contemporary change can be drawn. Firstly, the process of regime change seems to take well over five years and contains a period when it is unclear where change is actually moving. In an unclear period- which seems akin to times of policy experimentation- individual policies can be turned over rather quickly, yet ideationally incoherently. Secondly, full change does seem to happen with explicit debate on the set of ideas normally in the background. Crises seem to open windows of discussion, yet the actual change, when it happens takes more time. Hence for this study I expect signs of ideational change to be present, mainly in experimental form, yet not in full ideational regime change. The way to study this will be set out in the methodology.

(21)

21

Yet in the time period this study covers (2005-2011), a full ideational regime change is not expected. A loss of authority and loss of ideational regime status does seem possible, at least temporarily.

(22)

22

3. Research Methodology

This thesis aims to answer the following research question: Does financial crisis lead to economic ideational change?” Whilst it aims to find evidence for: An increase in post-Keynesian ideas in IMF reports since the financial crisis, implying ideational change.

In order to address the above research question, I have chosen a case study methodology with the IMF as the unit of analysis. The following chapter provides: (1) the rationale for selecting the IMF as unit of analysis, (2) the causal mechanism between the variables of this study, (3) the research methodology in the form of content analysis, (4) the sampling process, (5) the coding process, and finally (6) the analysis.

3.1 Case study design

The subject of this thesis is possible economic regime change in the IMF towards post-Keynesianism in the context of the financial crisis of 2008. It is an exploratory case study seeking to investigate the causal mechanism from the financial crisis to changes in the ideational economic regime.

While there might be a correlation between crises and regime change, as is widely argued in the literature, a causal pathway would not be explained by a correlation only (Blyth, 2002, p. 13). Therefore, a case study design is chosen as it fits the goal of delving deep into this causal mechanism. To enable this deep analysis it uses a small sample size with a textual type of evidence extracted reports. Since there are only a limited amount of international economic organizations an in-depth case study may be particularly relevant.

The IMF is chosen as the unit of analysis, even though it is one of several international economic organizations including the World Bank and the WTO (Noland, 2000). This is because the WTO is only concerned with trade, and hence is less relevant for this study. Both the IMF and World Bank have influence on the economic order and share the same Bretton Woods roots with an origin in economic crisis and concern for stability. They have been inherent parts of the international economic regime since World War II and hence represent relevant cases. The World Bank focusses on development, and is concerned with third and second world countries. The IMF is chosen out of these two due to its more dominant position in the international economic order and its primary concern with financial issues also stretching out to cover the first world (Peet, 2009, pp. 63, 66, 69). This makes it highly relevant in relation

(23)

23

to the 2008 great financial crisis. The IMF is central to the international monetary system of exchange rates and international payments (Peet, 2009, p. 68). Much criticism of the neoliberal order and neoclassical economics is targeted at the IMF. The IMF therefore represents a typical case for a change in economic thinking in the context of an economic and particularly financial crises.

3.2 Causal mechanism

To establish a causal mechanism first the conception and operationalization of the variables needs to be established. From the theoretical framework, I will first establish the conception of the independent variable of the financial crisis. Then I will establish the conception of the independent variable: the ideational economic regime.

The independent variable is the financial crisis represents, providing the external influence which might induce change in ideas. By relying Kingdon (2014, pp. 16, 20) the crisis can be conceptualized as a window of opportunity that may open. According to Blyth (2012, p. 11) a crisis is an external factor that can function as a start-point for change.

The dependent variable of the ideational economic regime (cf. Somers & Block, 2005) is defined as sets of taken-for-granted assumptions of how the world works (Danielson & Stryker, p. 134). Due to their embeddedness in society they come to represent some objective reality. Following Schmidt (2008) such ideas are present as explicit and implicit ideas on four levels of policy making. The first three explicit levels: (1) the policy rules in place, (2) the policy instruments, (3) the paradigm of ideas and theory that inform the first two levels, and the implicit level (4) as the philosophical world view containing latent final assumptions which informs and hence comprises all three explicit lower levels.

Operationalizing the concept of ideational economic regime can be done by breaking it down in the various concepts it entails: assumptions and ideas of an economic school found at the four levels of policy. These can then be listed. The lists state the indicators of the various concepts and hence how to recognize them. Neoclassical economics, the dominant regime makes up one of these lists. The potential alternative regime post-Keynesian economics is listed. The presence of an economic idea can be determined either by explicit statement of an idea, or the impossibility to arrive at a statement or conclusion without a particular idea. When it is likely that an idea found can be accredited to anyone of the two schools, then this accounted for by a different subcategory.

Operationalization of the independent variable, the financial crisis, involves change from one set of ideas and assumptions to another. I take Block & Somers (p. 266) account of

(24)

24

ideational regime change into account, which requires (1) time for the definition of the crisis itself, (2) for a discussion in the public arena, and (3) for a new ideational regime to establish itself. According to Blyth it is an external factor that can function as a start-point for change (Blyth, 2012, p. 11). Hence operationalization of a process involves time. Determining the dominance of the alleged dominant ideational regime can hence be done some time before the crisis. As seen from the theoretical framework economic regime change broadly defined may take a decade from the first signs of issues to a coherent new regime. The point of crisis may intensify the amount of policy changes in the search for solutions. These changes may take the form of policy experimentation at the first and second level described by Hall (1993, p. 285). A coherent new policy paradigm, including new assumptions that make it a regime may only come about after the previous regime lost all its legitimacy and new assumptions are used as a starting point. Yet, these changes may indicate the stages of change provided by Hall and Block & Somers.

In the transition from the liberal international order to the neoliberal order, for some time the former Keynesian assumptions were not let loose, although specific new measures and instruments were contradictive to these elements. These contradictive elements can be regarded to damage the authority off a regime. Yet, only when previous assumptions are seen as wrong and therefore new assumptions are used to inform policy, only then one can talk about change on that specific assumption and the informed ideas. Therefore these assumptions need to be explicitly debated first, in the public domain. To operationalize this would then require for core assumptions to first let go, and then be replaced. A dominant ideational economic regime may show incidental variance in policy due to experimentation; yet, overall the core assumptions should be congruent with each other. This requires differentiating between the different policy fields in and the core assumptions. The loss of a dominant ideational regime will show in the different policy fields, for example in trade or labour policy, which could no longer exhibit the rationale of the old regime, or not in all instances show rationale of the old regime.

Increasing the interest rate may show a change at the second level. A change of the rationale to having people spend less money may even show a third level change. Yet, when the inflation argument is used to spell out that inflation policy works because people are rational and will adjust expectations and behaviour accordingly only then is an assumption changed. To operationalize this, I will regard loss of policy regime when this happened for a majority of the policy fields inclining 50%>. For a pure congruent regime to come into being all field will need to be in accordance with the new assumptions. Yet, it seems likely that due to circumstances in the economy and traces left from earlier regimes. Hence, for the sake of operationalization I

(25)

25

will take a rounded number of 70% per cent to be indicative of a new ideational regime being in place.

3.3 Research method: Content analysis

Empirical evidence is collected through content analysis of IMF publications. By content analysis the meaning of qualitative data can be extracted and described in a systematic manner, while allowing for quantitative elements (Schreier, 2014, p. 170). It involves the following steps: (1) employing a coding frame, (2) generating category definitions, (3) segmenting the material into coding units, and (4) employing a pilot phase (p. 176).

To extract meaning via content analysis a coding frame is used which includes categories that can be applied to the material. These categories assist focussing the selected aspect relevant for answering the research question: for this thesis the variables in the causal mechanism (subchapter 4.2) as deduced from the theoretical framework (chapter 2). By focussing on these aspects it generates and orders the data systematically and makes it manageable. It functions by assigning codes to sequential parts of the material: the unit of analysis. This limits the extracted data to the available code per sequential part. The method is flexible as it enables assigning each code to various slightly different paragraphs (Schreier, 2014, p. 170). Codes in content analysis can be concept and data driven, and often are both (p. 171). In this thesis they are primarily concept driven, as deduced from the theory. Yet, one World Bank report was done in the pilot phase to make sure the categories matches the data. This makes the process of coding partially iterative as one goes through the process multiple times to sharpen the categories yet objective, while sticking to the deductive character of the codes. Moreover, it provides a sequence of steps to be followed to extract data. As such going over every piece of the material is required and every piece is examined that might be relevant to the research question (p. 171). In short, qualitative content analysis is chosen for its capacity to structure the research process, order and reduce the amount of data and enable qualitative and quantitative analysis.

3.4 Sampling

The sampling process was directed at getting a sample of four IMF publications. I choose the amount of four publications to have sufficient data sources to identify and compare differences between them, yet keep the required amount of analysis manageable. I consulted the online library of the IMF, the IMF eLibrary (http://www.elibrary.imf.org). This library mainly

(26)

26

contains economic reports. The sampling process took a three step approach consisting of (1) determining the list of search terms (2) periodization, (3) theme span, and (4) country selection.

Firstly, in step one the following search terms were determined to go through the online catalogue: policy, macro-economic policy, macro, micro, economic policy, policy measures, measures, government policy, growth, employment, monetary policy, interest rate, finance, trade, policy, budget, economic forecast, forecast, budgeting, trade. The choice for these terms is informed by the theoretic framework on ideational aspects of economic policy and the way economic theory subdivides the field of economics. The use of these specific terms is for their reference to either general economic policy matters, such as economic policy, or to more particular economic policy fields, such as trade.

The second step was periodization, which was determined around the outbreak of the great financial crisis in 2008. The years 2007 and 2009 come immediately before and after the outbreak of the financial crisis are selected to see possible short time changes. When, following Blyth, change however does not occur in a revolutionary manner but is incremental, two additional time points are required to be able to see such effects. The years 2005 and 2011 are therefore included. Nonetheless, it should be noted that the further a chosen year is away the from the 2008 financial crisis, the more the context changes as well. The year 2008 is left aside because arguably for proper analysis or drawing meaningful conclusions, it was too close to the crisis outbreak.

Thirdly, theme span determined whether a document was sufficiently broad to do a full analysis on, because economic world views are broadly applicable and ideally show internal consistency. Hence reports were excluded that only deal with one issue. Further, reports consisting (almost) exclusively of statistical data enable no or very little interpretation and are also not included. This left reports dealing with global (and regional) trade and reports on individual countries. While the global trade reports showed a wider span of issues than already excluded reports, these did not or very narrowly deal with (internal) non-trade policy issues. This left country reports as the ones significant for the intended research and led to the last step of choosing countries on which reports are written. Country reports provide a framework where themes are dealt with in a structured manner at multiple time points and thereby provide a framework for idea and theme development over time. Moreover, country reports are part of periodic consultations on IMF member country economic policies, which the IMF conducts as part of its obligations under article IV of its statutes support conditions of growth and stability (IMF, 2016). These consultations are influential in national policy making as they inform the IMF position on conditions that can be applied when a country seeks assistance in IMF funds.

(27)

27

The reports are written by IMF staff teams (see for example IMF, 2005, p. 1). The reports do state that views expressed do not necessarily reflect those of the IMF executive board, but of the staff (p. 1). The title selected issues simply refers to the issues taken up in the report as the staff apparently deems them relevant for the time of publication.

The fourth and last step thus became country selection. Countries were selected that (1) dealt with most themes, and (2) showed issues relevant for the financial crisis– i.e. having a monetary and financialized economy. Specifically the United States showed a lot of themes of interest to the selected themes for both schools and the financial crisis, as the fall of Lehman Brothers in the U.S. and the U.S. subprime mortgage crisis is commonly identified as the onset of the crisis. This process resulted in a set of four IMF country reports on the United States. An overview of the selected publications is provided in Table 1.

Table 1. Document list

Report Authors

Completion date

IMF Country Report No. 05/ 258.

International Monetary fund. Selected Issues. (2005). Washington.

Barrera, N., Celasun, O., Estevão, M., Keim, G., Maechler, A., Mills, P., Vir Bhatia, A.

July 1, 2005

IMF Country Report No. 07/265.

Balakrishnan, R., Bayoumi, T., Mathai, K., Mühleisen, M. Swiston, A., Tulin, V., Bhatia, A. Kiff, . J. Mills, P. Gorter, C., Rial, I.

July 11, 2007

Bayoumi, T. Mühleisen, M., Krajnyák, K., Schnure, C., Ivaschenko, I., Justiniano, Guscina, A., Swiston, A., Muir, D., Botman, D., Kumar, M., Kisinbay, T., Roger, S., Stone, M. Laxton, D., Kumhof, M.

July 13, 2009

Batini, N., Celasun, O., Dowling, T., Estevao., M., Keim, G., Sommer, M., Tsounta, E., Bin Li, G., Kiff, J., Chen, S., Felman, J., Mihet, R., Ratnovski, L.

July 7, 2011

(28)

28 3.5 Coding

Evidence collection and initial analysis were conducted at the same time by a process of coding. This study looks for Neoclassical and post-Keynesian ideas and concepts to see whether there is a change of economic reasoning. Since the economic schools are rather broad, themes are used to guide analysis and interpretation. These themes function as a lenses to find commonalities and differences in the various reports. These are established in two steps. First, the two schools provide for a preliminary coding scheme categorized in general, neutral, economic themes applicable to both schools and ensure that results are as comparative as possible. Second, in the process of coding, the reports also were used to shape the structure of the coding scheme. The table of contents and subheadings of the reports were most indicative in this process. When I encountered themes in relation to the economic theory not yet included in the coding scheme, I considered to take them up. This is visualized with the feedback loop of the conceptual framework in figure 1. In the end categories remained such as attitudes on markets, budgets and monetary policy. The full set can be found in table 2.

The coding process took two steps. First, colour-letter codes were assigned on the basis of text being in accordance with either Neoclassical/New Synthesis Economics or post-Keynesian Economics. Text was marked red accompanied by an ‘N’ in the former cases and blue accompanied by a ‘P’ in the latter ones. Paragraphs form the unit of analysis because this text level shows most coherently how a theme is treated and whether it should be assigned to either one school. However, during coding a considerable amount of information in the reports showed not to be explicit on the two categories and involves a lot of statistical data and plain information on facts or trends. Therefore the parts most relevant for the analysis are the parts of the reports that give advice or make normative statements. Thereafter the most relevant parts are those suggestive of a certain position due to wording or emphasis. Lastly, there are also paragraphs that are mainly descriptive: merely covering figures concerning themes without any theoretical statement, value or suggestion. In these instances the code DS for descriptive is

(29)

29

applied together with a sub-code for the main theme it deals with. Previous results in the following codes: N for Neoclassical/New Synthesis, P for post-Keynesian, ST for advice/statement, SG for suggestive and DS for descriptive. The two main categories of Neoclassical/New Synthesis and post-Keynesian economics is coloured from red to pink and from dark blue to light blue respectively. The lighter colours represent suggestive paragraphs, while the darker colours represent the stating and advisory paragraphs. Besides the general coding for schools and the intensity of the present theory, sub codes are used for specific themes. These can be found in table 1. During coding it became apparent that some paragraphs contained significant concepts or ideas from both schools. When such contrasts are found these are labelled C and coloured green. With stronger and weaker opposition being coloured dark and light green respectively. These were also labelled with a number for their respective themes.

Table 2. Coding scheme

Main code Sub-code Thematic sub-code

Explicit (EX) Suggestive (SG) No. Theme

Neoclassical/New Synthesis (N) Dark Red Light Red 1. Agency

Post-Keynesian (P) Dark Blue Light Blue 2. Time

Contrasting (C) Green Green 3. Markets

Descriptieve (D) Gray Gray 4. Business cycle

5. International trade 6. (International) Finance

7. Employment

8. View of money

9. Monetary policy

10. Fiscal policy, budgets and government

11. Growth

3.6 Method of analysis

The analysis took a four step process, where the first step of initial analysis is combined with coding and categorization described in the previous subchapter. The second step involves

(30)

30

the first tool: the computer program MAXQDA. It assists in processing each selected document and categorize each piece of evidence. This was done by first interpreting the evidence and making a note with the rationalization for the selected category. These aspects mostly cannot be translated clearly to notes, worksheets or figures. Therefore it is necessary to take these aspects into account when analysing and interpreting and include them in the results whenever relevant. In the third step these labelled and weighed pieces of evidence were extracted into a matrix, the second tool, with the respective categories. Lastly, evidence was plotted in time-series graphs to find patterns. When a difference in the relative amount of Neoclassical, post-Keynesian or Contrasting codes (together the ideational codes) passes a certain threshold, then it is ideational change is argued to have occurred. The definitive loss of an ideational regime is regarded to have occurred when the relative amount of a regime formerly making up at least 70 percent of ideational coding in a report is found to be below 50 percent of all ideational codes. Conversely, when an alternative set of economic policy ideas, in this thesis post-Keynesian ideas- is found to be above 70 percent of all ideational codes, then it is regarded to represent a newly established ideational regime. Thus Ideational regime change can involve replacement of the regime, yet can solely involve the loss or gain of a position.

3.7 Validity and limitations

As a case study directed at delving into the causal mechanism of ideational change in context of the crisis, the results mainly seeks internal validity. This depends on the correct measurement of the effects of the independent variable (the crisis) to the manipulated variable (ideas). The accuracy of the measurement depends much on the instrument employed analyse the manipulated variable. Hence the quality depends on the set of ideas and assumptions that are applied. These ideas were researched and written up very precisely and extensively for this study. Subsequently they were inserted in an overview of ideas capturing the core of every idea to efficiently recognize and apply it. For in-depth latent presence the extensive description of an idea was applied. Going back and forward to see if such measurements still made sense and held truth was to corroborate the code application. A cut-off point was established at the point between on the one hand ideas that are likely to belong to a school of thinking depending on the clarity of the statements or reasoning found, and on the other hand statements and reasoning that are merely describing or do not show signs that can be attributed to the beliefs of the authors or the organization.

A further limitation, inherent to content analysis is (the threat of) subjectivity. Whilst a precise and well thought trough methodology is highly beneficial to the methodology, the

The Financial Crisis and Economic Regime Change: The Case of the IMF